Amplitude Aktie

Well, good afternoon. Thank you for joining us, day 2 of the Citi Global TMT Conference. My name is YC Wong. I'm part of the software analysts here at Citi. We are happy to have CFO of Amplitude, Andrew Casey with us. Welcome, Andrew.

Thank you for hosting us.

Look forward to it. So maybe just give us a quick background about yourself. You've been at the company for, what, 2-plus years and a quick background of the company for those new to the name.

I see I'm very old, and I've been in the Valley a long time. So too much background would be -- take too long. We didn't have any chance to ask questions. But I started my career in the Valley in Silicon Valley in 1995 with Sun Microsystems. I then went to Oracle, not through the acquisition. I was there for a couple of years, then went to Symantec, then I went to HP.

And then I went to a company that very few people have ever heard of called ServiceNow. I was there for a little while and really cut my teeth on building out a scalable enterprise software business.

Right. For sure. Everybody knows Mike Scarpelli over there, right?

Mike was a great mentor and a great CFO to work for. I aspire to try and live up to his standards. I probably would still be at ServiceNow if they would have made me the CFO, but the Board really wanted somebody who had already been a CFO.

And at that point in time, John and Bill were in transition and both of them asked me to stay, but I told them that, look, even if Gina leaves within 5 years, the fact of the matter is by your criteria and the Board's criteria, I still wouldn't be qualified for the job. So that's why I left and I went to a company called WalkMe, helped take them public in 2021. I then left WalkMe before the sale to SAP and went to a company called Lacework, which is in cloud security. We sold that to Fortinet.

And the funny story I tell people is that in August of 2024, I was signing the definitive agreement to sell Lacework to Fortinet at noon. And by 2:00 p.m., I was announced as the CFO and been with the company for 2 years now.

Congrats on that journey. That's definitely a long tenure in software. Yes. Maybe going to Amplitude, it's not too shabby to manage a company on this growth trajectory right now, especially in the past few quarters. Can you give us a sense of overview what happened that saw the reacceleration in the business?

Yes. And I would say it's funny going back 2 years, John leads our IR capabilities at corp dev. He had done a kind of a synopsis of all of our transcripts and our earnings releases and other things, you go back to that. And there were 2 major strategies that we were talking about when I first joined Amplitude. First was increasing the amount of business that we do with enterprises. And what we mean by enterprises is any company that has 1,000 employees or has $100 million in revenue, okay?

So increasing the business with enterprises, everything is better with enterprises, gross retention, net retention potential, you name it. And we wanted to have a greater footprint with enterprise clients. The other thing that we were looking at was really broadening out our product portfolio.

One of the things that Spenser, our CEO and the other founders had noticed was that there were all these applications that have been created around product analytics, which is what Amplitude at its core is product analytics.

Think about any software that you distribute, you need to have instrumentation and observability associated with the application, the website, the wearable. You want information and feedback to see how customers are interacting with it. Well, all these other applications have been created around that, around experimentation and session replay and guides and surveys. And the posture that was that really those didn't need to be separate. They really should be part of a broader platform. So those 2 strategies were when I first joined, we were trying to get off the ground. We're trying to get in motion.

And there were just a host of things that we did on both fronts building new products ourselves, acquiring companies to augment the capabilities, really engineering the applications together so they work very, very seamlessly in customers' workflows, changing the go-to-market profile, the process, territory design, pricing and packaging, all these things we were pulling levers on to really go drive those strategies.

And I would tell you that the acceleration you saw in revenue from when I joined in Q2 '24 of 6%, all the way up to the most recent announcement around 22%, I would say it's kind of boring, but I would go back to those things and say, this is just us doing very well against some of those core strategies, those core actions that were meant to go drive both those strategies.

Now sprinkle that, too, with we've increasingly added more agentic capabilities into our platform. We're releasing more agentic products for fee, and that's broadening our appeal and our interest from those classic enterprise clients who are asking the key question is, how do I adopt some of these new AI capabilities and get the return on investment that I'm expecting?

Yes, there's a lot of levers, a lot of products in there that are like helping drive the growth. But since we are more in a CFO discussion, maybe can you talk about what are some of the key metrics that you are seeing that give you the confidence of that 20% trajectory?

Well, certainly, as we go forward and the underpinnings of our guidance where it shows accelerating our revenue and ARR growth is based upon customer adoption, our pipeline itself, opportunities we see around moving into new product categories.

I think all of those things are giving us a lot of benefit. And I would say one of the strategies that we implemented was to go drive longer-term relationships with our clients. And when you're driving a consolidation story in the market, which is us going into an enterprise client saying, look, you've got FullStory and you've got Pendo and you've got AB Tasty.

We can consolidate all of those onto our platform, and we'll charge you less in aggregate fee and you'll be more efficient. So when you do that type of consolidation play, one of the things you realize quickly is that customers are not going to just rip out all those things at the same time.

They want consistency in the rollout of how they're replacing that and influencing workflows. And so that usually means they want a longer-term arrangement. And for us, it's exactly what we wanted as well. We want that longer-term arrangement because it not only gives us greater predictability, but when we sign up a client who is doing a multiphase replacement of applications, they're committing to that.

And they commit to it, that shows it up in RPO. And RPO has been growing over 30% for 6 consecutive quarters. And why is that important? Well, because it's greater revenue visibility, there's greater revenue predictability. -- and it reduces the actual amount of renewals that you have to do in any given period because you've got a broader set of contracts for a longer period of time. When I first joined, we had to renew 89% of the installed base.

By enacting this, we had to renew 72% the next year, 56% the next year, and we're driving that down further and further. I'd much prefer a contract duration, which isn't 22 months, which we're currently at, but more closer to 30-plus months.

That enables me to have an investment in sales and marketing, which is more focused on new logos and expansions rather than just doing renewals. So all of these strategies, I would say, are instrumental. In fact, I remember, it wasn't maybe a year ago that some -- an analyst asked me, well, isn't this just luck? And I said, no, it's not luck.

Like the COVID luck.

Driving the RPO growth was absolutely something we instrumented. We rewrote the comp plans. We had -- so the incentive structures were there. We taught sales how to do deal constructs, which aligned to the customers' value propositions.

We got them into the right framework around gives and gets around getting to a longer contract duration. So -- and pricing and packaging reinforces this as well. All these things are, as you mentioned, the levers we were pulling to generate growth.

Yes. How does this like longer duration contract would have an impact on your like net new retention? And then also, how is it impacting your net new logo versus expansion drivers?

So it's a good question. And I would tell you that just intuitively, if I have to renew less next year, even if I had the same gross retention rate, churn would be down.

Churn dollars are down, then net is up. So it's driving growth. And I would tell you that the reality is when you execute well on that front and sales has to spend less time on doing renewals, they can spend more time going after new logos. And by the way, we've slowly but surely ratcheted up the focus from the sales team on generating new logos.

Every year, we've added things to the comp plan structure or the territory design in which they have to bring in new enterprise logos in order to meet their quotas. So once again, these are all tactics and instruments that we've used to go drive the right behaviors and to achieve the core KPIs that we look at.

Obviously, core to us is ARR. And you know this because you've studied us. But the reality is, I think a very conservative perspective and definition around ARR. It's exactly what we will recognize over a 12-month period. There are no caveats.

There are no exclusions on contracts. There's no termination for convenience. There is no monthly annualization. It's committed contracts is what -- and that's what we hold our sales reps accountable to as well, creating committed contracts.

Yes. Like since you joined, I think enterprise has been a much bigger focus. Is that like enterprise driving to have this longer-term contract versus shorter term? Or it's more the lever that Amplitude is pulling?

I would say it's more Amplitude in the construct, but you lean into the customer motivations, like I mentioned before, customer who wants to replace multiple applications over a period of time, they actually want cost predictability, which means they want a longer-term contract.

So they have that cost predictability. And a champion who's doing the work, who's actually doing the replacement of those applications with Amplitude looks at their success as Amplitude being successful in that environment. And so they want to have a partner in that.

And the way they have a partnership in that is you have that longer-term contract. So it's a multitude of factors. I would say the contract duration expanding is a result of the actions we're taking to align to the customers themselves.

Okay. Yes. I guess with the contract duration, there's also a pricing impact, which has been a big topic over the past few quarters. There's a new pricing model. Could you help us refresh our memory on the old pricing and the new pricing? What are the key changes?

Sure. So when I first joined, I gave a report out to the Board, and I do a very classic 30, 60, 90-day report out. And the first report I gave had a section on it basically said our pricing is not aligned to our strategy. And it's causing churn, it's causing complexities. It's causing a lower adoption rate.

And the Board basically said, Andrew, why is this the case? -- and I went through, basically, the pricing that we had was very much a point product. Every product had its own meter, its own price curve, its own value proposition, its own -- in fact, it was a separate product, a different software development kit.

It wasn't reflecting what we were increasingly doing in the platform and bringing products together. And customers were complaining. And it was very much an optimization for a single product rather than an optimization associated with the experience we wanted our clients to have in having access to all those applications and us being able to monetize it very easily.

So what was all those things are wrong are things that we've changed. So the first thing we did is we started talking to our customers and understanding where do we think that this platform sale is going to go? Does it need a different type of meter? Ever since we've been founded, we've used event volume as the primary meter. And I would say 86% of our installed base had that.

So when we were talking about all these other products having different meters, I mean, if you were a sales rep at Amplitude prior to making the change and a customer wanted to adopt the entire platform, you'd have to ask them for 5 different estimates on how they're going to use the platform.

And if they were over one of those estimates and under in another, we'd still charge them. for the additional, not over one of the individual. So it's very much a point product rather than the customers looking at, well, I don't really understand exactly what my adoption pattern would be. This is what my best estimate is. So we asked them, is event volume still the primary way in which you believe you're getting value from Amplitude as you use more.

And we came back and it was still one that was very valid with clients. And the industry has adopted event volume as a standard metric. And so it wasn't a problem with the meter itself.

Now that doesn't mean you get the price point right, right? You have to understand that when a customer uses more, they do expect that they're going to get incremental discount associated with the unit price they're paying. That's just standard.

Yes. Yes, especially longer duration, okay, what discount are you giving me, right?

Exactly. Well, we actually didn't have that. We didn't have a volume-based curve for quoting. So the reps were quoting prices that were all over the place. And customers talk and how come I got a better price than the other one? And so having a standard framework for volume-based discount is something we had to go build and develop, and that got changed.

And now we can go and talk to a customer and like, look, if you want to have 1.5 billion events, it's this price. If you move up to 2 billion, it will be this price. Now obviously, you architect that, that there's marginal incremental discounts. But as they move from one tier to the other, they are going to pay more in ARR just at a decreasing rate. And so all that architecture had to be done.

And then the other thing that customers were very clear about was I need greater cost transparency on what I'm going to be paying Amplitude if I adopt more of your platform. I can't have all these different meters. You're just going to push your compliance over to me.

And so what we did is we -- because we indexed off of event volume in that price point and did the volume-based curve, we then index every other application as an uplift on that price. So that's very simple for a customer to compute what they're going to pay Amplitude as they adopt more of the platform. that trend.

What is that uplift that you're seeing on the new pricing versus the old pricing?

So each module, it ranges in its, I'll call it, value attribution to a client. You have experimentation, guidance surveys, activation, experimentation.

They all have a range of about 25% to 50% uplifts individually. And then you have other ones that are smaller like AI feedback, which is more like 10%, session replay, which is more like 20%, but they each add up. And if you get each one of them and you're adopting a full platform, you probably pay somewhere between 3x to 4x, what you would if you were adopting product analytics alone.

So -- but it's very simple for a customer to compute what that's going to be, and it's very easy for sales reps to quote. And a funny thing happens when sales reps actually quote more often, they have more engagements. They actually drive greater velocity, forecasting improves, pipeline improves.

So the underpinnings of driving greater efficiency on the sales side, greater adoption from customers, having them giving confidence in what -- and transparency what their cost is going to be and driving multiproduct adoption, all these things were objectives we have with the new pricing and packaging.

And I would tell you that we're doing very well against each one of them. Now we're rolling it out in a very methodical way to our sales team.

I'll tell you a funny story when I was in the boardroom when we were talking about the new pricing and packaging and changing it. One of our Board members said, "You know what, Andrew, CFOs and CEOs have been fired for getting this wrong.

And I said [indiscernible]

Yes. And I said, I guess we shouldn't get it wrong. So we did a lot of testing with clients and got a lot of good feedback. And that's why this last quarter, 70% of all transactions, including renewals, were on new pricing and packaging.

And I expect that we're going to continue to see that ratchet up to where it's 100% of every transaction every quarter. And now we've got roughly 28% of ARR that's on new pricing and packaging. And I expect that, that's going to get into -- upwards in the 60% by the time we end of the year.

The good thing is we're seeing already that average ARR is increasing. contract duration is increasing, multiproduct adoption is increasing. And all those things benefit all other key metrics like gross retention, net dollar retention and in customer adoption.

Yes. The uplift that you mentioned, overall uplift of that especially the 70% in Q2, expecting 100% going forward. Do you see that uplift going up, increasing as we progress?

It's been about 5% to 10% based on the customer transactions itself. I actually think that probably will continue in that way. I don't think it will be dramatic though. The one thing that you've got some customers who are overpriced on some things and that they're being corrected.

You had some that are underpriced that we're taking a nice progression with. I think the biggest lever that we're seeing with the pricing packages as it applies to uplift in ARR though is multiproduct, additional product they're adopting.

Just given the pricing is still at very early stages, right, in some ways, are you worried there's going to be some optimization like going forward?

There certainly will be because some of those price points that our sales reps would quote and they didn't have a volume-based discount curve outsized. There certainly will be. And we've addressed those as we've gone forward. And in some cases, you do have contraction if it was egregious.

Other cases, we found our way into selling them more product in order to absorb those amounts. I wouldn't say it's been a huge headwind, though. I think it's been a very modest one.

Okay. No, that's definitely good to track, see where that's going. Now I want to pivot a little bit just given there's 10 minutes left, and then we still haven't talked about the newest and greatest stuff -- we have, right... Statsig acquisition, right?

I love to hear your thoughts on why you acquired them? And then what do you see with Statsig in the product lineup?

Yes. So there's a couple of things you have to understand that are going on in the market right now. First is that every enterprise is probably either has or is talking about ways in which to centralize their data in their data warehouse through Snowflake or Databricks or another. And that's a major trend happening. Two, one thing we definitely know about agentic capabilities is that the ability to create code has been never more been so fast. You can now have an agent creating code for you in specific areas.

So code development, the whole build and ship process of the product development life cycle is exponentially grown, okay? As part of that, there's also an increasing effort to figure out how you fix the next part of the bottleneck in the product development life cycle process around the QA and test and learn.

And so one of the things that the industry has been struggling with is now we're creating this code, is it actually working as it's intended to? And now we're seeing the benefits associated with it.

And that's where Statsig had done a really good job about appealing to engineering use cases in the product development life cycle process, experimenting and testing how that code has actually been deployed. And they've done it in a data warehouse native environment. Amplitude had the experimentation product as well, but it was primarily in the cloud-based environments. And so we were starting to build a product to go after this opportunity because those major trends we see happening. And so when OpenAI approached us about acquiring the technology, it was very much when we looked at is, look, this is the leading technology in this space.

They've got a bevy of customers for it. And we believe this is going to be a fundamental addition to Amplitude's efforts longer term around our agent analytics, what we see happening in the product development life cycle process and the benefit of having agentic capabilities deployed in our latest product, Wave, incorporating the experimentation efforts as well.

So think about it in terms of now Wave, which we're in beta and we demoed on our last earnings call, think of it as an agent that is encapsulating what you would normally have a product engineer, a data scientist and developer, all working towards creating a product.

Well, you can have an agent actually doing that work for you recommending what should change next in that product or that process and bringing in the experimentation capabilities associated with Statsig and doing the experiments and giving the feedback on it all encapsulated with...

Yes. For sure, I want to touch on Wave, but staying with Statsig for now, like following the acquisition, I know there's only the product side because the whole team technically just remain with OpenAI.

Like what has like amplitude did to integrate the technology? How was the hiring because now you are selling to a different persona from Statsig's business? How has that transition been?

So it's been actually better than what we expected. As part of the agreement with OpenAI, we only had a 60-day transition period to learn the code and set up a Google environment to actually host it because a lot of those things didn't come over.

And so we were -- day 1, we were hiring new engineers to staff up the team. That's gone very well. So we've got a good team now that's, especially in support. The selling side, I think, is an interesting one. It's certainly more technical sale related to an engineering persona. And -- and our sellers hadn't been used to that in the past. So we have hired what I'll call some specialists, but they're not -- I wouldn't call them overlays. I would call them more augmentation.

They're more like we always had a technical group that was in our sales team that was helping out when you have these highly technical sales. And now there's just some specialists that actually handle the engineering use case in the Statsig experimentation product itself.

Do you need something like an FDE that everyone has been talking about?

We do have FDEs. I think that that's -- it's not necessarily a Statsig, although that can be the case. I think FDEs are where customers are coming to us and asking, how do I get there?

I really don't need to start. And if that's the case, then you deploy forward deployed engineers actually show customers how they could change their business process and become or integrate more agent capabilities in order to make those processes better. Maybe it's through cost efficiencies you're looking at or optimize the revenue funnels.

Okay. I guess just given, in general, like looking at capital allocation, right, you have Statsig that's being opportunistic. How do you view the market opportunity that in the next 6 to 12 months on your capital allocation needs?

Yes. I think the next 6 to 12 months, I think that there's a couple of things that are really interesting to me. I think every enterprise is starting to deploy in some fashion, agentic capabilities, whether that's customer service related, fraud detection, business process optimization, there's just a host of areas where enterprises as they digitally engage with their clients are looking to augment with agents.

That's a surface and an area where we believe we can sell agent analytics around because every customer is going to want to know, is the agent actually doing what was intended to do? If not, why not? What would you change? And so well, I think it's a huge opportunity for us to go sell off that. We're already dealing with a number of clients in the financial services sector, in the telecommunications and media sector with an exact intent.

Even if we're not selling the agent, we're simply surrounding the agent and making those interactions better. So we think it's a huge opportunity.

Yes. As you expand on -- with agent use cases, right, anything that you see on the next frontier that is interesting that...

Yes. I think this whole move towards these router optimization companies associated with the usage of models is an interesting area as well because Amplitude could add its own behavioral heuristics around it to help optimize for the outcomes.

So think about it in terms of most of these companies are approaching model usage based on the individual model capabilities and the cost structure, not necessarily engineering it for the outcome that they're trying to achieve. And so I think that's an area where we could play as well.

Yes. Maybe just guess on like the Wave product that you talked about earlier, right? That's kind of an agent that you can add to do things. Would that have any cannibalization risk as to the core analytics or experimentation product?

How do we think about it as that agent continue to improve or even the models continue to improve. We have Astra last week, like that's just an exponential improvement.

Yes. Look, I think Wave is an interesting product for us because like Statsig, it's a whole new potential land for us. You actually don't need to be an Amplitude customer to get benefit associated with Wave.

However, if you are an Amplitude customer, it's a whole set of data set that you can leverage in order to make that agent as it's deployed either against your website or your mobile application or your product development life cycle. All of that is more richer data for that process to be optimized. So it's not an either/or.

It's you deploy Wave along with your Amplitude infrastructure and data sets and connections makes it that much better as you're actually deploying it and pick your favorite use case, but it's one that makes -- it's not working against Amplitude. It's actually working with Amplitude. You can use it outside of Amplitude.

You could literally deploy Wave against one of our competitors' products as well, use that as a data source.

Yes. As Wave kind of scale usage across Amplitude, do you anticipate any like margin compression? Like how does it work versus your core analytics product and Wave...

Well, Wave will be its own price point, its own meter. We will price it certainly to optimize for value as much as possible. We'll have Agentic capabilities embedded within it. But I see it as a little different from our core analytics. Think about our core analytics products are capturing event data. Let's use a marketing use case for an e-commerce site.

It's capturing tons of event data on the e-commerce site, and you can deploy Wave to actually optimize and personalize every interaction that people are having on the e-commerce site because it's recommending and taking action based upon parameters you've actually specified for it.

So do you see it as an option to do something like outcome-based pricing where a lot more people are talking about?

So outcome-based pricing is an interesting thing. I think what you do in your pricing structures is you optimize for the customer achieving the outcome.

And the thing we think we can do for Wave is that we can give a very detailed level of feedback that as customers spend money on tokens for a specific thing that they can actually see what the return is associated with those tokens so they can get to -- that's the outcome. I don't believe that you take every customer's unique outcome and then try to embed that within a unique pricing model. If you did that, you'd have a unique contract every time because customers have different desires. We have financial services clients who are trying to use -- or testing Wave to do loan optimization.

You've got telecommunications and media that's trying to optimize our marketing use cases around promotions. And so I started pricing on both those things, suddenly I have got a lot of unique instances, and it's very difficult to scale that. The architecture in pricing is when you've got to get to a meter, which a big rule in pricing construct is you should have a meter which you anticipate is constantly going to increase.

And the perception of that meter has got to be one that customers look at it and say, I'm getting greater value as I use more. right? If those exist, then you can get scale in your pricing architecture and still achieve those value-based outcomes based on the price point you actually charge.

Okay. No, I think that's a good way to start since pricing has been one of the bigger levers for Amplitude to continue their growth. Well, Andrew, thank you so much for joining us.

Pleasure.

Good morning, and welcome to Amplitude's Second Quarter 2026 Earnings Conference Call.

I'm John Streppa, Head of Investor Relations. And joining me today are Spenser Skates, CEO and Co-Founder of Amplitude; and Andrew Casey, Chief Financial Officer.

During today's call, management will make forward-looking statements, including statements regarding our financial outlook for the third quarter and full-year 2026, the expected performance of our products, our expected quarterly and long-term growth, investments and our overall future prospects. These forward-looking statements are based on current information, assumptions and expectations and are subject to risks and uncertainties, some of which are beyond our control that could cause actual results to differ materially from those described in these statements.

Further information on the risks that could cause actual results to differ is included in our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and we assume no obligation to update these statements after today's call, except as required by law.

Certain financial measures used on today's call are expressed on a non-GAAP basis. We use these non-GAAP financial measures internally to facilitate analysis of our financial and business trends and for internal planning and forecasting purposes. These non-GAAP financial measures have limitations and should not be used in isolation from or as a substitute for financial information prepared in accordance with GAAP.

Additional information regarding these non-GAAP financial measures and a reconciliation between these GAAP and non-GAAP financial measures are included in our earnings press release and the supplemental financial information, which can be found on our Investor Relations website at investors.amplitude.com.

And with that, I'll hand the call over to Spenser.

Thanks, John, and good afternoon, everyone. Welcome to Amplitude's Second Quarter 2026 Earnings Call.

Today, I'll cover 3 things. First, our Q2 results. Second, how we transformed Amplitude into an AI company and why every company I talk to now wants to learn how they can do the same. Third, a look at our product and a spotlight on our customers.

Let me start with the numbers. Q2 revenue was $101 million, up 21% year-over-year. Total annual recurring revenue was $410 million, up 22% year-over-year and up $36 million from last quarter. That was made up of 2 parts; inorganic ARR from Statsig of $17 million and organic ARR growth of $19 million. Andrew will walk through the details. Non-GAAP operating loss was $1.5 million. Customers with more than $100,000 in ARR grew to 824, an increase of 30% year-over-year. Both AI natives and large enterprises are driving this growth.

Let me step back and tell you about our transformation and then how we're helping customers along their AI journeys. We help companies build better products. Every company wants to transform to deliver software products in an AI-native way. We've made that transformation at Amplitude over the last 2 years, and now our customers are looking to learn from us.

Becoming an AI company starts with the organization. Two years ago, we first transformed our engineering team by bringing in AI engineers who built with it for years. Then we moved into adjacent functions like product management, design and the more technical parts of go-to-market. We also brought in AI expertise through acquisition. Founders and other members of the team from these companies have taken leadership roles across Amplitude.

I have focused on bringing in leaders who are former founders and who have a technical background. Gab, our Chief Product Officer, started multiple companies, including Loom Systems, which sold to ServiceNow in 2020. In addition, Nate, our Chief Commercial Officer, has a degree in math and physics and started his career as an engineer programming in C++ and Java and building databases. Most recently, we added Angela Ferrante as SVP of Marketing. Angela founded Laudable, which went through Y Combinator Summer 2021, sold it in 2025 and is a technical marketing leader who builds apps with AI in her spare time.

In addition to all of this, we're continually reeducating everyone at Amplitude through initiatives like AI Week, unlimited token spend and a living token leaderboard. This has all resulted in 3x the number of pull requests in 6 months. We've reduced our pull request cycle from 5 hours to 44 minutes. Bug reports are down 55%. 5% of our pull requests are submitted from designers and product managers with no engineering involvement.

We've leveraged AI to shorten our closing process by a day. We built customer health dashboards that enable our sellers and leaders to track customer usage, bringing our own Amplitude data alongside Salesforce data and data from other sources. When I talk with our customers, they are all focused on how they can transform their business to be AI-native, like we have done at Amplitude. The AI landscape is changing rapidly, and they want to learn how to adapt. Our customers are on a spectrum of AI adoption. Our job is to meet them where they are and then educate them on how to take the next step.

We work with leading AI companies to learn what the bleeding edge in product development looks like. We use that knowledge to educate the rest of the market, including the largest enterprises deploying at scale. More than 40 AI-native companies now pay us over $100,000 a year. Those customers include Harvey, Midjourney, Character AI and one of the leading foundational AI model companies.

On the enterprise side, enterprises are now more than 68% of our ARR. This quarter included agreements with Paramount, Jaguar Land Rover and Domino's Pizza. We've improved our pricing and packaging. We reduced down to a single meter to make it simpler for enterprises to add additional products. We increased the amount of data on our free plan, so we're the best for those just getting started. Amplitude has the best pricing, whether you're a start-up or a large enterprises.

One of the biggest changes with building an AI-native company we're seeing at Amplitude and with our peers in private markets is in the cost structure. A lot of inference spend is required in order to deliver AI-native products, which increases the amount spent on cost of goods sold. On the other hand, you do not need to add as much operating expense to continue to grow a business at scale. We are embracing this change in cost structure as part of our transition to an AI-native company.

For now, we expect gross margins to stay in the low-70s. We will offset that with a commensurate reduction in operating expenses. That allows us to continue to show the same leverage in operating income as we have planned. I am continuing to drive Amplitude to a 20%-plus operating margin business over the long term.

We offer 3 products to meet customers wherever they are on their AI journey. Amplitude gives you the deepest understanding of how people use your product. Our agents increasingly do that discovery for you. Statsig gives you feature flagging and experimentation built on the world's most advanced SaaS engine with an engineering-first view. It's also integrated natively with data warehouses.

Wave is the future of product development, self-improving products where we automatically recommend what to build next based on signals from users. While we're early here, I'm actually excited to show you a demo today. Together, these 3 products close the product development loop, understand what's happening, measure what ships and ship what matters. That loop is how AI-native business is built.

Let me go deeper on Amplitude. Global chat is becoming the primary way our customers interact with their product data. You ask it a question in plain language and it does the analysis, no dashboard building required. It's become the de facto way many companies do product analytics. Global agent finds the root cause behind 75% of customer questions and hands you the answer. There are 1.3 million global agent interactions every week and root cause discovery rates are improving by 1 percentage point every month. As of today, over 40% of all insights come from AI agents as opposed to humans, and we expect this to continue to grow. Today, for our demo, I want to show you custom agents, Statsig and Wave.

Let's start with custom agents. Custom agents are teammates that automate recurring workflows on your product data and push that work to other tools and systems. This is our chat interface. An increasing number of users are interacting with Amplitude mostly through chat and agents. I'll ask a question. Which group of users are most likely to purchase next week? Chat can now write its own code to perform this analysis. This unlocks the ability to run deeper analysis and create powerful new graphs and artifacts, including diagrams like you see here, out-of-time decile lift, an ROC curve, segment propensity.

You can dig in by seeing the actual code used and step-by-step analysis. This type of deep analysis has never been available before in analytics tooling. We are no longer bound by the constraints of a UI. We can also create automatic and recurring agents that run in the background. I give it these instructions. I want this analysis run every Monday morning, cross reference with marketing activity and Confluence, DM me the results in Slack. Amplitude then creates the agent that you see here. This is the entire prompt, including connectors to Atlassian and Slack. It will run regularly every Monday and push the results to me.

We are building the best analytics agent across all data sources. Statsig is the leading product for experimentation and feature management. Statsig runs experiments natively on your cloud data warehouse, whether that is Snowflake, BigQuery, Databricks or Redshift. Let me show you what this looks like. Here is the results page for one of hundreds of experiments that an e-commerce customer is running. This experiment is testing a larger product image versus the default size. There's a lot of statistical machinery behind a good experiment, but the UI makes it simple for an engineer to run.

Up top, they can monitor exposure, which is saying if the experiment is healthy or not. We expect to see a 50-50 split, so we're doing good. And as you can see over here, we're getting a healthy check. We move to the scorecard that has the results. This has a confidence interval of 95%. Statsig uses advanced techniques like CUPED and sequential testing that allows engineers to speed up time to decision. We have those turned on.

In monitoring, we see specific events we're tracking for this experiment. We're seeing positive results. The checkout event is up by 27.4%, plus or minus 2.3%. Cart conversion is up. Total purchase dollars is up, while cards per session is down. For the rollout of this feature, we have a progressive rollout, starting with employees moving to early access users, then early release and a scheduled rollout for everyone else. Statsig has a variety of advanced experimentation capabilities for rollout like feature gating, dynamic configs and automatic rollbacks. Together, these are the mechanisms that a team uses to ship a change gradually, tune it while live and pull back automatically it goes wrong.

Last, I want to show you Wave, the future of product development. Wave allows for self-improving products that automatically recommend what to build next based on signals from your users. Wave is magical. Wave looks across all the different data sources you have; analytics, experimentation, session replay, Guides and Surveys, feedback and many others. It then synthesizes that data into a set of product recommendations, plans those recommendations and then helps you create those changes in your product.

I'm going to walk you through a real example Wave suggested and built for Amplitude's documentation site. On our documentation site, Wave found a spike in failed searches through looking at Session Replay and analytics data. The core problem was that search on our docs page fired on every key stroke, typing a single letter to start a search returned an empty no result state before the person finished typing their search leading to a bad experience for users. Wave explains the reach of this issue. Every user who uses search, it has an expected impact of decreasing total search failures by 80%.

Then Wave has automatically created a visual example of the problem below, so it's easy to understand. It also has a full explanation of the evidence. For the plan, Wave sketches a wire frame of the recommended update, setting a 3-character minimum and a 200-millisecond debounce to trigger the search. Wave can also drive execution. It automatically created the poll request and cursor wrote the code. Mark, our Technical Writer, was able to merge this poll request and ship this. No engineers, no designers and no product manager. Finally, Wave measures the results of the change. There is a massive decrease in total search failures. Simply amazing, simply amazing.

Now, let's talk about some of our customers. We had a great quarter with both new lands and expansions. We added or expanded our relationship with customers, including Paramount Global, Jaguar Land Rover, Teladoc Health, Chime, Disney ad platforms, F5 Networks, Coursera, Grammarly, Kraken and Crunch Fitness, among others.

I want to tell you 3 stories about how these customers are leveraging our platform. First is Coca-Cola FEMSA, which sells to hundreds of thousands of small shops across Latin America. Every shop is different, but for years, they have to run the same broad campaign to everyone because there is no way to tailor a message to that many retailers by hand. AI changed that. They began sending each retailer its own recommendation every week written by AI. Their own teams were actually skeptical.

A different message for every shop every week felt risky and no one knew if it was going to work. They used Amplitude to find out. Their AI campaigns actually had an 11% click-through rate, 4x higher than their previous approach. Our cohort analysis also showed that this lift lasted. Once a retailer engaged, its revenue stayed higher in the weeks that followed. That evidence turned skeptics at FEMSA into believers, and they scaled from a 2,500 store pilot to 690,000 retailers.

The second is Replit. Replit is an AI app builder that allows non-technical builders to turn an idea into an app using AI. Replit has a large global user base of passionate builders that provide feedback. Replit is using Amplitude AI Feedback to understand how customers are engaging with their agents. They've connected AI Feedback to Zendesk, App Store Reviews, Twitter and Reddit and surfaced and prioritize what problems should be solved to increase their retention and engagement. It changed weeks of manual work on their end into a simple click with Amplitude. This is the next generation of product development at work.

Third is The Economist. The Economist is a print magazine that's in the midst of a transition to digital delivery and subscription. Their research arm built an AI assistant called Lens that answers questions for analysts and strategists using The Economist's content. Their normal analytics could show what users did, but not whether the AI's answers were any good. The team was reading sessions by hand, but they couldn't keep up.

Amplitude Agent Analytics now scores every answer Lens gives automatically. They went from reading a handful of sample sessions to being able to see across all of them. Today, Lens holds a 96.9% task success rate and weekly failures are down 84%. That is the loop working, build with AI, measure whether it is good and fix what is not.

To wrap up, the companies on the bleeding edge are choosing Amplitude. We've transformed Amplitude to be AI-native, and we're building the future on what can be done in analytics. Self-improving products are closer than ever with Wave. Our pace of innovation continues to accelerate, and we're building in a way that can scale with leverage. I'm extraordinarily excited about what's ahead.

With that, I'll hand it over to Andrew to walk you through the financials.

Thank you, Spenser.

This was a strong quarter and a clear step forward in our execution, bringing our vision of how products will increasingly be developed and improved. We crossed $100 million in quarterly revenue. ARR reached $410 million, growing over 22% with the addition of the ARR assumed from the Statsig business and free cash flow was a record quarterly high of $23.7 million.

We also returned $69 million in capital during the quarter as part of our share repurchase program. We accomplished these milestones while integrating the Statsig technology and customers, managing through our own AI-native evolution and implementing our new pricing and packaging strategy.

AI is changing how customers use Amplitude. The more our customers build with AI, the more they need to measure. Customers that adopt our AI into their workflows run nearly 10x the number of analyses compared to those that are running things manually. This increases the value that customers receive from the data ingested into our platform and makes it more likely that they'll both ingest larger amounts of data and expand into additional products, which is the basis of our growth.

Our new pricing and packaging is working. It supports our market consolidation strategy by providing customers with a lower overall cost if they consolidate applications onto our platform. It provides customers greater cost predictability and simplifies the quoting process for our sellers.

In the second quarter, 70% of the ARR we closed was on the new model, up from 25% in the first quarter. Now, 28% of our total ARR is on the new pricing and packaging. This is leading to average ARR increasing, higher multi-product attach and longer contract duration, which all contribute to greater durability of our revenue. Our margins reflect the choice. These are investments we are making to drive future growth with increasing profitability.

Our gross margin was down over 1 point versus Q1 due to the integration of Statsig. We are working to optimize the new hosting environment and cloud structure, but it will take some time to improve from the low-50s gross margin closer to our expectation of 70-plus for the Statsig business. We're also experiencing higher customer adoption of AI capabilities and greater data ingestion into our platform, which combined has increased our costs and reduced our gross margins by an additional 2 points versus Q1.

We have long maintained that we will grow with leverage. This investment in the cost of revenue places greater emphasis on the management of our operating expenses to a lower level in order to achieve the leverage. In Q2, we've managed down our sales and marketing to below 40% of revenue and G&A to the low teens, which is contributing to an increase in operating margins. We will continue to manage both areas lower as a percentage of revenue over time, and we will continue to invest in R&D to drive innovation.

We are instrumenting our business to accelerate growth, capture market share and show leverage. One key metric we monitor is the usage of data compared to the entitlement for our customers as this is a primary monetization metric. Today, that metric is at an all-time high. This is the output from better pricing packaging and more usage driven by our AI features. We have increased the durability of our business through our RPO growth and reinvented our internal processes to capture scalability that AI offers. We are running the AI opportunity and taking share as we go.

Turning to our second quarter results. As a reminder, all financial results that I'll be discussing, with the exception of revenue, are non-GAAP. Our GAAP financial results, along with a reconciliation between GAAP and non-GAAP results can be found in our earnings press release and supplemental financials on the Investor Relations page of our website.

Second quarter revenue was $100.9 million, up 21% year-over-year and 8% quarter-over-quarter. Total ARR increased to $410 million, exiting the second quarter, an increase of 22% year-over-year and $36 million sequentially. This includes $17 million of incremental ARR from the Statsig business compared to the $16 million we expected to add when we shared our first quarter earnings.

Total remaining performance obligations grew 35% year-over-year to $483 million. Current RPO was up 30% year-over-year and long-term RPO was up 47% year-over-year. Here are more details on the key elements of the quarter. We had a strong quarter for both new and expansion deals in the enterprise and platform sales were again particularly strong. 48% of our customers now have multiple products, with 80% of our ARR coming from that cohort.

We have over 26% of our ARR from customers with 5 or more products, up 2x since the second quarter last year. In period, net dollar retention was 105% on a pro forma basis, led by cross-sell expansions across our customer base. This pro forma basis includes Statsig and Amplitude customers. Gross margin was 71% for the second quarter, down approximately 4 points from the second quarter of last year and down 4 points sequentially. This was driven by continued growth in inference costs as customer adoption of our AI tools accelerated, along with the integration of the Statsig business in its hosting environment.

Sales and marketing expenses were 39% of revenue, down from 44% in the second quarter of last year. G&A was 13% of revenue, down 1 point from the second quarter of last year. R&D was 21% of revenue, up approximately 3 points from the second quarter last year, reflecting investment to scale the Statsig opportunity and support for those customers.

Total operating expenses were $73 million, or 72% of revenue. Operating loss was $1.5 million or 1.4% of revenue. Net loss per share was negative $0.01 based on 129.4 million basic shares compared to $0.01 a year ago. Free cash flow in the quarter was $23.7 million, or 24% of revenue compared to $18.2 million, or 22% of revenue during the same period last year. We ended the quarter with approximately $162 million in cash and investments.

We have conviction in the long-term value of our platform and have used and will use our cash to minimize the impacts of dilution. Our balance sheet position remains strong and allows us the opportunity to be more aggressive in our M&A strategy to accelerate our R&D road map when appropriate.

Now turning to our outlook. As a reminder, the philosophy of how we set guidance is through the lens of execution. We are pleased with our overall progress on consolidating point solutions to our core platform and the adoption of our different AI technologies. We've instrumented our business and selling process to make it easier to use more of our platform. We believe that we are well positioned to continue to accelerate our growth in a profitable way.

For the third quarter of 2026, we expect revenue to be between $105.6 million and $108 million, representing an annual growth rate of 21% at the midpoint. We expect non-GAAP operating income to be between $2.5 million and $4.5 million. And we expect non-GAAP net income per share to be between $0.02 and $0.03, assuming a weighted average shares outstanding of approximately 133 million as measured on a fully diluted basis.

For the full-year 2026, we are raising our expectation for full-year revenue based on the performances in the second quarter to be between $407.2 million and $411.2 million, an annual growth rate of 19% at the midpoint. We are also raising our expectation for the full year non-GAAP operating income due to performances in the second quarter and actions taken in the first half to be between $6.3 million and $9.3 million. We expect non-GAAP net income per share to be between $0.06 and $0.08, assuming weighted average shares outstanding of approximately 137.1 million as measured on a fully diluted basis.

In closing, we are accelerating our pace of innovation, and we're growing the value that we can deliver to our customers. We have confidence in our ability to scale a durable and growing business while also bringing agentic analytics to the world.

With that, we'll open up for Q&A. Over to you, John.

Thank you, Andrew. Going to Q&A. [Operator Instructions]

Our first question today will come from the line of Mark Cash from Raymond James, followed by Jackson Ader at KeyBanc.

Yes. If I could start with Spenser, I really wanted to ask around Wave. I appreciate it's still limited beta, but I think you've been using internally for several months now. I guess, do you see Wave, if it could cause maybe a shift -- a company shifting away from using bespoke agents for specific use cases towards a broader AI-native product development platform from what you're seeing? And if so, how could that change your buyer, maybe the budgets you see in the addressable market over time?

When you say bespoke, like, say, more on that, like...

Yes. Instead of using particular agents to do a specific task underlying because you have like a swarm of agents doing things underneath for Wave. So...

I see what you're saying. I see. Okay. So yes, let me separate out a few different things. What we have on the Amplitude side, and I showed with custom agents is you have these agents that can look across the data and find insights for you and get to the root cause of questions and do that on a regular basis and kind of send it out. What Wave is doing, in particular, to your point, is it's kind of -- it's looking at all your data all the time and then saying, hey, here are points of friction. Here's something that's not working out should be. Here's a feature that I think you should emphasize more. Here's something that I think is a best practice that you're not doing. And so it's operating at a kind of higher level.

In terms of the persona, I think what we're seeing is a convergence between engineers, product managers and designers into this AI builder persona. It's not really like you have engineers who are thinking about what to build and you have product managers who are also just chipping code. And so the best -- if you look at where the AI-native teams that everyone is aspiring to be, these roles are melding. So it's still the same problem we're solving, which is how do we help you build a better product, but we're just automating more of it because we're saying, hey, we're going to look at all the data all the time and then suggest recommendations. That -- like I've been -- we've been talking about self-improving products here at Amplitude for about 9 years.

And so I'm actually been blown away by what is possible with the technology today where it's just -- it's the perfect problem for AI in a lot of ways. The data sets are massive and complex. So, you can't get any human to look at them. And then the synthesis of, okay, here's what I think could be better and best practices is actually like extraordinarily impressive. And so what that means is that just by the fact that someone is using your software, like it's getting better because it's just translating recommendations. You no longer need someone to go into an Amplitude or to any data system and say, oh, here's what my interpretation of these results.

So, I do think -- in terms of budget and persona, I do think, again, that means instead of having these distinct roles, you have engineering, product management and design merge. You're still doing digital product development, and that still rolls up to some leader, the same executive before. But yes, the way you do it looks different.

Did I hit it on what you are looking for?

Yes. Absolutely. And if I could follow up with Andrew real quick. If my math is correct, the guidance for the year was raised by more than 2x the beat for revenue and operating income. So, I was wondering if you could just go to the key drivers of lifting growth expectations Why you saw some pressure on pro forma expansion sequentially there in the quarter? And then what you're considered regarding margin leverage and levers while you're facing COGS pressure and ramping token spend internally?

Yes. Sure. So, a couple of things. One, that when we look at our ability to actually generate revenue in the out quarters, one, we start with the strong balances we're booking that are showing up in our RPO. When you've got commitments from customers for a longer-term duration, you start to have better and better predictability about your future revenue. So, that's the first thing. It's one of the reasons why we emphasize that metric so much.

The second thing is we look at how much our customers are actually responding to some of the initiatives we're putting out, and that comes in the form of our new product capabilities, our new pricing and packaging, areas where our sales team is running new promotions and activities, all those are bolstering our ability to see a stronger and stronger pipeline, and that pipeline progresses faster through its stages, which gives us greater and greater confidence that we'll add more and more in net new ARR.

Now from a revenue perspective, as you know, the predominance of our business is all coming from our subscription revenue. So, those key factors on understanding what's the baseline, what can you see in your pipeline, what you expect to convert is what I refer to is our ability to go execute against the plans that are in front of us. And sales teams have been doing a really good job of driving consolidation in the market, and that alone with our products is driving great conversions. So, that's the first thing.

On some of the margin areas, I would tell you, look, -- we just -- in the case of the Google environment that we got or the Statsig, we're going to be focused on driving optimizations in that environment over a period of time. It's definitely lower. We said in the low-50s from a gross margin perspective. That comes from us taking on a whole new environment. Most of Amplitude, all of it, in fact, is on AWS. So, we took on a whole new cloud and hosting environment, and you have to go through the paces of really optimizing how you run those environments for customers.

Our first objective was integrating, making sure there were no disruption in service. Now, we're moving quickly into how we can optimize those environments. So, that's one big lever on the gross margin side. And we're constantly looking at how we can make investments to go drive greater efficiencies across all of our operating expense areas.

Great. Our next question will come from the line of Jackson Ader from KeyBanc, followed by Scott Berg.

I was curious on -- I guess, Andrew, kind of sticking with you and talking about rather than on the cost side, just on the operating expense side. We've seen really nice acceleration in organic ARR from the business. But if I take kind of a longer-term view, even on a non-GAAP basis, we're still around breakeven, right? And so I'm curious, as you're thinking about like driving more leverage and more incremental margin that you've talked about before on the income statement, what kind of impact should we expect that to have on the organic growth number, if at all?

Well, I'd tell you that, one, we still expect from an organic perspective, we've got a great set of products. Spenser just walked through a number of them that are brand new to the market. We think they have an enormous total addressable market that we can go after. So, revenue growth will be the predominance where we'll see increasing operating income. As far as leverage as a percentage of what that would be a percentage of operating income, I do expect over time that we'll be able to drive better and better gross -- cost to start revenue and increase gross margins over time.

It just takes time to go do those things, especially when you're seeing such a demand inflection from customers and increasing data lines. As I mentioned, we're at an all-time high for the amount of data ingested in the platform versus entitlements. When I first joined, that was in the low-60s. We're well into the 80s now as far as percentage of what customers have ingested versus what their entitlements are, and that portends increasing expansions on upsell, which is usually where we've had a lot of problems in the past of overselling and having to right-size contracts.

For the first time, we're past those things and we're starting to see really good upsell, not just cross-sell driving growth. So, revenue growth is the predominance of the first aspect of driving improving profitability. As far as the leverage goes, I think gross margins will improve over time. It's just going to take a while. And we still have a long way to go on sales and marketing is reducing that as a percentage of revenue. I think G&A has room. And I do think that over time, we'll see greater and greater efficiencies with the R&D organization as they adopt more and more capabilities to build products at a faster rate.

Okay. And then just a quick follow-up. Can you remind us, should there be any -- now that we're on a different kind of pricing and packaging model, a little bit more variable, I guess, if you will. But should there be any difference in terms of the seasonality of your revenue ramp or recognition as we move forward with the new packaging?

So on revenue, I'd say, you get a fairly predictable pattern under which revenue is recognized because as I said, most of our revenue in the future periods is designated by our RPO, the committed contracts. But ARR will follow a very typical seasonal pattern. My expectation are bit more on in the enterprise selling basis. Q1 will always be our weakest as far as net new ARR adds as we're adding new territories, adding new reps, implementing new strategic initiatives. This year, in particular, we're educating the sales teams on not only the new pricing and packaging, but a lot of the new products we have. So every year, you're going to have that.

And so it will be a slow start and then pick up. This year, too, just to remind everybody, we also had some big changes in our sales and marketing leadership, which is predominance of what you see now flowing through and the cost benefit from a lower sales and marketing as a percentage of revenue, and that's from efficiencies we're driving.

Our next question will come from the line of Scott Berg from Needham, followed by Billy Fitzsimmons.

Spenser, Andrew, nice quarter. I wanted to follow up on sales enablement that Andrew was chatting about there. We did a couple of different customer checks in the quarter. And the one thing that we came back is, I don't think your existing customers are quite aware of all of the different modules and innovation that you've rolled out this year.

Yes. Totally.

I see Spenser smiling. I know that's a function of timing, obviously. And one customer didn't even know that you had acquired Statsig. So, I guess where are you kind of in that journey? When is the sales force properly ramped in that? I mean, the quarter's sales results were good as is, but obviously, better awareness there can be even more helpful.

Yes. To your point, I think a lot of people still bucket us in the analytics company, and it drives me absolutely crazy. Honestly, just sharing, hey, we have Statsig now, and this is bleeding edge feature experimentation. And you can use it too, and this is the same infrastructure. OpenAI runs internally, like awesome. A lot of customers don't even know that. And then same with Wave, I think they're just starting to understand Wave and then same with our other products. I think if you remember from the prepared remarks, like we do see ramping. So, we're moving customers from 1 to 2 to 3 to 4 to 5 to more products, but it's much slower and that drives me crazy.

I think there is no substitute for the work of like, hey, we built something amazing. We have to educate the hundreds of people we have in our field. And then they have to educate the thousands of customers in market. Like that's just work and that's just the whole thing. Something I'm spending a lot of time with Nate, our Chief Commercial Officer as well as the rest of the executive team on in terms of how do we get that and do that more efficiently. We just had a kickoff a few weeks ago where we showed off a lot of what you saw today with Statsig and Wave and custom agents. But that's not even to say the -- all the other products we have like Session Replay and Guides and Surveys and AI Feedback that can displace point solutions.

So anyway, that is -- I think last year, we said it was the year of the platform. I think we still have a ways to go on educating people on it. I will say that the good news on it is the main thing customers are looking for is, hey, prove to me you guys are at the bleeding edge of where this field is going. And so my view is that analytics and the whole data -- behavioral data ecosystem is going to go through the same shift that coding has in the last 2 years, like that is still going to happen. And so they want to -- we see it in like a lot of the stuff we've been demoing and our customers see it, too.

And so they want to know, hey, am I working with the company that's bleeding edge on this? And so even if they're not necessarily ready to adopt a Wave or even a Statsig, I know that, okay, you at least help me take the first step to using some of the basics on these capabilities, and then I can add more even if it's maybe too overwhelming for me right at the start or I'm not ready as a company. So anyway, that's all to say we still have a bunch of work to do to make sure our field is equipped. There's definitely areas that do it extremely well, but then there's areas we need to do a better job on this. So, I appreciate you calling that out.

And then from a -- my follow-up question is on the integration traction with Statsig. You all had a pretty aggressive goal, obviously, to move that asset into your organizations. Kind of where are you with it? Because the other customers that we spoke with were super excited about that. A couple of them were Statsig customers, et cetera. So, just kind of understand, have you hit all your goals around that? And are you kind of at that point where now you can just deliver on product and sales versus just having to integrate the organization?

Yes, yes. So as you imagine, Statsig as been around for 5 years, and there's a lot of work with getting it from a whole group of people who have never seen the code base or sold it or whatever else. I think we've kind of gotten through -- there's always stuff, but we've gotten through all of the urgent fires in running and delivering Statsig. So, that's great. Customers are very excited about how it's landing. We want to make sure to give -- the fact that it's our main focus as opposed to at OpenAI, it was a little more of a side thing for them. It's all been received positively. So, that's good.

Now we're starting to think about, okay, what's coming next for Statsig? So if you look at like statsig.com/updates, we're shipping stuff. We've been shipping stuff for the last few months. We're continuing to build on the road map. We're continuing to integrate with Amplitude much more tightly so that if you're on both, which a lot of our customers are, you get the benefits of being able to use data from one and the other.

And I think a lot of -- the other thing we're seeing with Statsig is that there's a lot of demand from AI-natives in particular. So, one of the reasons we're really excited to join forces with Statsig is that they -- like a lot of the way the future product development is being run, like people are choosing Statsig for that. So it's engineering-first teams that tend to be much more technical. They're building out whole software development harnesses. They want to manage how stuff is deployed in that harness and Statsig is set up really, really well to scale.

As I mentioned, OpenAI runs a version of that infrastructure internally for themselves. And so they've tested that in tons of different ways over there, and we're doing the same thing except with everyone outside of OpenAI. And so there's a lot for us to do in terms of how do you set Statsig up to be a core part of the software development harness for all these bleeding-edge AI customers, and it's where kind of everyone wants to go over time. So, that's what we're focused on.

Our next question will come from Billy Fitzsimmons from Piper Sandler, followed by Clark Wright from D.A. Davidson.

Good to see the results and guidance. I think one of the exciting things about Statsig is potentially the cross-sell opportunity. I know there are some things to do first. But last time I looked or last I checked, I think there were 80 of the 400 Statsig customers are on Amplitude already. So, there's a lot who aren't. Can you just help contextualize for us how we should think about the potential cross-sell opportunity of Amplitude into Statsig or potentially vice versa and how we should think about that flowing through the model long term?

I think probably the much bigger opportunity is to take Statsig to Amplitude customers. I think Statsig customers, as I mentioned earlier, tend to be much more bleeding edge from an AI innovation standpoint. And so that's where everyone is trying to get their organizations to over the long term. It is a very -- it's like a more -- Amplitude is historically focused on product management and then Statsig is much more tailored towards engineers like it has tons of customization out of the box. It has like all the statistical testing.

Now, like I said, those 2 personas are merging, but it's early days on that. So, I think the opportunity is as more of our traditional Amplitude customers look and try to build like AI-natives, introduce AI to their software development process, try to build out a harness, eventually try to get to self-improving products, all of those are opportunities for us to bring Statsig. Now, we definitely do see places where Statsig customers are also very interested in Amplitude, but there's a lot more both from a number and ARR basis that are Amplitude.

Perfect. And then if I could ask a second one. Can you just contextualize maybe how either your hiring needs have kind of changed year-to-date or where you're seeing the best ROI from AI-driven efficiencies internally within Amplitude?

There's a ton. On the hiring front -- so a few different things. One, like it's been -- I've been just very focused on transforming the entire workforce, getting leaders, getting engineers, getting people in other functions that are AI-native, both by hiring that talent, acquiring it, hiring executives that have that background. And then in addition to that, retraining the -- and re-educating the workforce that we have here, great part, like everyone wants to learn. It's like, yes, people see like, hey, the more I can learn how to use AI, the more relevant my skills are going to be both at Amplitude and at other places in the future.

So, everyone is like embracing it, which is great. The few specific areas, I think on -- so yes, that's like an always ongoing thing. Like I just -- we were just adding Angela, which we announced today in marketing. We're always looking at companies and other places to pick up talent. New Grads is another great source of very highly leveraged talent. One of the funny things I'll tell you guys is during downturns or whatever, a lot of companies pull back on university hiring because it's like the easiest thing to cut. But if you have the confidence to evaluate who is great from that talent pool, you can get some exceptional folks right out of school, which is awesome. So we've been -- had that as a big focus here at Amplitude. So, that's like the primary thing.

And then the one specific area is Statsig. As you imagine, this is a huge complex product and code base and architecture. And so our -- we've taken our existing experimentation team, and they're now running Statsig, which is awesome, but they also need a lot more help. So, we're adding lots of different roles in hiring on that data science leads for deployed engineers, other engineers who are just familiar with that architecture. We've actually hired one person who used to work at Statsig, pre the OpenAI acquisition and we're continuing to go more there. So, there's a lot we need to do there. We've kind of caught the ball, which is good, but now we have to like go maximize it.

Our next question will come from Clark Wright from D.A. Davidson, followed by Koji Ikeda from Bank of America.

Great to see the 30% year-over-year increase in customers with over $100,000 in ARR, which looks to be the highest since 2021. Can you potentially break out the adds from Statsig? And what else is helping in terms of the new logo momentum that you're seeing today?

Sure. So, about 40 customers came from the Statsig business itself that we added. And so if you kind of do the quick math on that, you're still well in almost 23%, 24% growth in customers that are in that greater than $100,000 cohort. And so it's still growing quite nicely and contributing to ARR and to revenue growth. So, that was really good.

And as Spenser mentioned earlier, what we're seeing back when we're talking to customers, especially as we've gotten introduced them for the first time with their brand-new customers to Amplitude that were formerly Statsig customers is we're finding that they're, one, very appreciative of the fact that Amplitude is shepherding and taking forward the road map and showing confidence in our ability to actually give them a future where self-improving products is a reality.

And they do that through adopting an experimentation mindset, and they're very confident then to move further with Amplitude in other areas. So, that cross-sell expansion opportunity is real. I think we talked about it at the time, there was a multi-hundred million dollar opportunity for us just in the installed base. So, we're pretty excited about it.

Got it. And then last quarter, you called out event volume growth being 21% year-over-year. What is that now as you kind of talk about the momentum that you're seeing in all-time highs? And how should we think about the ramp of that metric going forward given agentic workflows and the amount of events that they can process?

Yes. It's definitely growing faster than both ARR and revenue. And it's one of those areas that -- for us, it feels like we've gone through many, many quarters of trying to bring it up and get the entitlements rightsized and everything else. It's definitely a leading indicator for us that, one, we're not going to have the same types of churn issues like in the past. Two, sales has adopted that value-based orientation sale where they're not trying to get everything upfront. They're trying to get our customers to value quickly and show them the value of an expansion. And like I said, it's an indicator that we're going to see upsells have a larger meaningful contribution to growth, whereas before it was a tractor and the predominance of our growth with cross-sell. We're just not going to have those same instances if we've got customers who are bumping up against their entitlements and getting value from the investment they've made.

Our next question will come from Koji Ikeda from Bank of America, followed by Nick Altmann.

I wanted to ask a question on Wave. Love the demo, long-term vision. I mean, it sounds like it's going to be awesome for finding problems and finding solutions, generating code, measuring outcomes. I mean, it looks like the full deal here. And so the question really becomes, if Wave is successful in all the things I think it could be, then why would you need the other products from Amplitude like Statsig and product analytics? Seems like you could do it all from Wave.

Yes, yes, totally. Okay. So yes, this is -- I brushed over this architecturally. What Wave does is it takes data from lots of different data sources. So it takes analytics data from Amplitude, experiment data from Statsig. We're eventually -- we're planning to make it agnostic long term, so it can take data from any analytics thing. If you're using Google Analytics or Adobe or something else, it doesn't matter, and then translate that insight. So, you still need a place to get that data. Like it's not like it can just look at a product and figure out what people are doing it. It actually needs to have that data from some area.

And so it's a nice build where it's like, hey, use Amplitude, use Statsig. The more data sources you put into this thing, the better the output that you see. One of the big learnings from the AI boom is that the power of massive scale of data, it just gets you better and more accurate and more insightful results like that is just a straight -- that -- like you can -- the scaling laws look like you can grow that almost infinitely. So, Amplitude Analytics actually as well as the experimentation and everything else we have, play a really important part in being the collection points for that data. Again, though, the goal is to be agnostic, so we can just plug into whatever system, your data warehouse, your own internal thing, other tools, third-party tools and kind of build it on top of that.

I think another thing is that because we have that data, that gives us the ability to have much greater insight into the right things to build. If you're a start-up starting out for the first time and you don't have the massive multiple petabyte data set that we have, it's like, okay, how do you even know if what you're recommending is best practice or what leads to something good? And so there's a lot of feedback loops that we have because we have this data set, we know, okay, hey, here's what a great e-commerce app looks like. Here's what a great social media app looks like. Here's what a fintech app should look like. Here's the typical workflows for sign-up that work well. Here's what message customization should be so and so on.

And so because, like, we're one of the few companies out there, there's no open source equivalent data sets for it. And so having that allows us to develop a much higher quality, better version of Wave than kind of anyone else out there. So the other good part is it's not like a -- it's an alpha. So, there are customers using it. We're using it internally. There's a number of start-ups. There's a few enterprises that are using it. And so it's spitting out real things that frankly, you look at this and you're just like [indiscernible] how did AI come up with this? This is crazy.

I'm convinced that whoever wins this space, that's going to be a multibillion-dollar business, if not more. And so our thing is like let's run forward with that as fast as possible. I think we're well positioned in the opportunity because we're the leader in analytics and a few other areas. Yes. And let's go build that business as quickly as we can.

Our next question comes from Nick Altmann from BTIG, followed by YC Wong from Citi.

Awesome. Just to build off Koji's last question, I kind of wanted to ask the inverse on Wave of, like, it seems like there's more incentive to adopt the broader platform with Wave. And I know it's still very early, but how are those kind of conversations going with customers? Like are you having more sort of multi-product or platform adoption customers as they kind of look at Wave and this vision of the self-improving product?

And then the follow-up there is just how should we think about Wave being monetized more so in the near term? Is it kind of indirectly in the sense of it gives customers more incentive to adopt the broader platform, and that's how you sort of plan to monetize it? Or is it kind of a stand-alone SKU?

Yes. So you're exactly right, which is the more data sources you feed to this thing, the better. And so we've already -- I've already seen multiple customers who have gotten on Session Replay as well as one that signed up for AI Feedback specifically because, hey, the stuff fed to Wave makes it a lot better. And so you're absolutely right, where like it drives like the whole platform play where it's like, okay, you have all these individual point things and then you just -- there are more data sources.

Session Replay in particular, is very, very powerful. Like as you imagine, viewing the exact state of UI and where a user clicked has a lot of value for how it can be better. So, that's been awesome to see. And again, early, there's a handful of customers on it. But as we grow it out, I think that will drive more adoption. And I also don't think like -- to my point earlier to Koji, it's like, the goal is to be agnostic with it. We want to build the most bleeding-edge thing. And so if we plug in other sources too, all the better.

On the monetization front, we'll charge for it. We absolutely will charge for it. I mean, you think about the value that this creates. Now, you go from analytics or data tooling where it's like you have to manually go in, collect an event or look at -- ask a particular question, get a result out, think about how to apply to the business. And now you're having a whole flow that does it for you, hey, I've already seen this user is having friction here like the docs example I made is like, hey, we see most search queries are failing. Why is that? Well, they're single characters, and we're not waiting until someone types a complete word, so they get this error when they're in the middle of the typing and it feels bad. And it's like, okay, yes, you should resolve that and make that better. And it's not just that. It's like that times hundreds of things all across all surface areas of your product.

One of the lessons is that like behavioral data and product surface areas are so large, it is impossible for any team to stay on top of them. And so the fact that this thing is looking all the time for how it can be better is it's magical. Like it's crazy what it can do. So, I think whatever company goes to win that is going to be multiple billions in revenue, if not more, and we want to aggressively go after it. And yes, customers are willing to pay for that.

Now again, early days, we're in alpha. So, we haven't figured out exactly how we're going to monetize it, but we absolutely will charge for that capability. That's like -- that's one of the great -- people are talking about, hey, there's all this money going to AI, where does it actually come out? And this is one where you can draw the line really directly. It's like, look, the customer experience is getting better. They're spending more. There's more revenue. There's less friction, there's less downtime, like the whole thing is just better, like great use from an application standpoint.

Our next question will come from YC Wong from Citi, followed by Arjun Bhatia from William Blair.

Spenser and team, great to see the fast expanding AI platform, here you have like every quarter. Like, I want to touch on Agent Analytics, which now seems to measure like agent themselves, right? I mean, where the market that we see is already multiple vendors out there trying to measure prompts, measure latency, hallucination, like, all the stuff that you can see, but what is the customer problems that the Agent Analytics could solve that the current observability platform cannot? And then how do you view the market opportunity of that problem?

Yes. So, I mean, I think, first, to the extent this replaces most traditional interfaces and the market opportunity is as large, if not larger, than what's going on, on traditional user interfaces with Session Replay and analytics. In terms of our unique positioning, what we offer, which I shared a little bit in the customer story about Economist is that you can connect what's individually happening within a session to the long-term impact to your business. So, you can say, okay, hey, you got a successful answer back from the bot, did that lead to you spending more or signing up or keeping your subscription? Conversely, if you ran into a problem and you got frustrated, did that lead to some negative long-term outcome? And that loop is really, really important.

A lot of the engineering-specific observability products we've seen in this space are just kind of stand-alone. It's like, okay, they'll just show the traces and that's kind of it. And you have no idea if it's actually leading to different results down the line. And so that's why we see both traditional enterprises that are transforming their businesses like the economists as well as a lot of AI-natives. I mentioned one of the largest foundational model companies. They also are looking at, as you imagine, they'll have a lot of tooling there, but they want to know, okay, is this leading to someone to becoming a subscriber to upsell and all of that sort of stuff long term. And so being able to connect that journey end-to-end is what we uniquely offer.

That sounds like a more TAM expansion opportunity there.

Absolutely. Absolutely. I didn't cover as much today. We demoed it more on the Q1 earnings call. But yes, it's actually one of the things my Chief Commercial Officer and I are very excited about.

Yes. Definitely look forward to hearing more, including Wave. I have a quick follow-up for Andrew as well on the guidance. Like Amplitude growth has definitely been accelerating for the past year and more, right? Even adjusting for the Statsig business this quarter, I think it's still accelerated. But the implied guide that I'm looking for Q4 shows about 2- to 3-point decel. Could you kind of help us double-click on the largest step down on the Q4 guide? Is it more just seasonality or incremental conservatism?

What I would tell you is that we always take a look at what -- when we're building our guidance, what we believe is very strong occurrence to occur. And I mentioned some of the factors earlier about pipeline, how well that pipeline has developed, where we're seeing good demand from our customers. Usually, Q4 is our strongest quarter from a new ARR perspective, and it's because that's the way we built our comp plans. That's the way enterprise selling cycles run typically in a calendar-based company. So, I would just tell you that our guidance is based upon what we know is out there as far as our pipelines, our RPO, and it's what we're comfortable with.

And our last question will come from the line of Arjun Bhatia of William Blair by Willow Miller.

Can we hear your updated thoughts on the 20%-plus revenue growth target given the strong growth this quarter and the strong third quarter guide? I'm curious to hear how you're thinking about it now, considering Statsig and now Wave?

Yes. I mean, I think Statsig is an accelerant to our long-term plans, which is part of why we -- Vijaye and I agreed Amplitude would be the best home for Statsig long term. I think the -- so we put up $19 million in organic growth last quarter in Q2. And so it's just -- we're just touching on that 20%. It's like the annual number is $410 million. So if you divide that out, it's like we're just shy of that 20% growth target when you annualize the quarterly numbers.

To me, as I've always said, 20% is kind of bare minimum, frankly. We want to be making sure to continually hitting and exceeding that 20%. Long term, we're aiming a good deal higher. We want to get to 30% and then beyond that as we continue to grow the business. Obviously, a lot of work between here and there, but that's what we're very focused on doing.

Thank you, Willow.

That will conclude our second quarter earnings call. Thank you for your time and interest. We look forward to seeing you this quarter on the road as we attend conferences hosted by KeyBanc, Citi and Piper Sandler.

Thank you.

Thank you all.

Thank you.

Good afternoon, and welcome to Amplitude's First Quarter 2026 Earnings Conference Call. I'm John Streppa, Head of Investor Relations.

And joining me today are Spenser Skates, CEO and Co-Founder of Amplitude; and Andrew Casey, our Chief Financial Officer. During today's call, management will make forward-looking statements, including statements regarding our financial outlook for the second quarter and full year 2026, the expected performance of our products, our expected quarterly and long-term growth, investments and our overall future prospects.

These forward-looking statements are based on current information, assumptions and expectations and are subject to risks and uncertainties, some of which are beyond our control that could cause actual results to differ materially from those described in these statements. Further information on the risks that could cause actual results to differ is included in our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and we assume no obligation to update these statements after today's call, except as required by law.

Certain financial measures used on today's call are expressed on a non-GAAP basis. We use these non-GAAP financial measures internally to facilitate analysis of our financial and business trends and for internal planning and forecasting purposes. These non-GAAP financial measures have limitations and should not be used in isolation from or as a substitute for financial information prepared in accordance with GAAP. Additional information regarding these non-GAAP financial measures and a reconciliation between these GAAP and non-GAAP financial measures are included in our earnings press release and the supplemental financial information, which can be found on our Investor Relations website at investors.amplitude.com.

With that, I'll hand the call over to Spenser.

Good afternoon, everyone, and welcome to Amplitude's First Quarter 2026 Earnings Call. Today, I'll cover 3 things. First, our Q1 results; second, how AI is reshaping the software development life cycle; and third, a deep dive into our latest AI products and the customers putting them to work.

Let me start with the numbers. Q1 revenue was $94 million, up 17% year-over-year. Annual recurring revenue was $374 million, up 17% year-over-year and up $9 million from last quarter. Non-GAAP operating loss was $3.1 million. Customers with more than $100,000 in ARR grew to $727, an increase of 18% year-over-year.

Our progress in expanding the enterprise and growing our multiproduct footprint continued in the first quarter. Dollar-based net expansion improved sequentially to 106%. This reflects continued strength in our core business as we expand the capabilities of our platform to help the next generation of builders understand, improve and grow their digital products.

I am focused on aggressively transforming Amplitude into an AI company. In Q1, we made broader changes to the leadership within go-to-market to remove layers and become a more technical team.

Nate Crook is now our Chief Commercial Officer, overseeing sales, customer success, revenue operations and enablement. Nate and the team now own the entire path from landing a customer to ensuring they succeed long term. We restructured customer success and marketing to match customer buying trends. Customer success now has fewer handoffs and deep technical coverage with forward-deployed engineers. Marketing is now oriented around AI-native storytelling.

We welcomed Gab Menachem as Chief Product Officer last month. Gab is a serial founder who built Loom Systems, an AIOps company acquired by ServiceNow. Loom Systems analyze log data across cloud and on-prem similar to what Amplitude does for behavioral data. Gab then spent 6 years scaling ServiceNow's IT operations management business to more than $1 billion in revenue.

I'm excited about that combination of founder DNA and enterprise experience at scale. Gab is part of a growing group of founders we brought into Amplitude over the past 18 months to lead our AI transformation. A few weeks ago, we ran AI Week at Amplitude. We paused normal work across the entire company so that every function could build and shift AI-powered workflows to reimagine their daily jobs and functions.

It is much more important for Amplitude's talent to be AI native over the next year than any short-term initiative in the business. The team shipped hundreds of amazing demos, including automatically creating custom demo websites per customer, automating part of the quarter close process and automating how we create new creative assets in marketing.

Yesterday, we announced a strategic partnership with Statsig. Amplitude, as part of this partnership, Amplitude will take on Statsig's brand and customers. We will also maintain and develop the current Statsig platform across the cloud and data warehouse, including support for all existing Statsig customers. Amplitude will also begin building a more integrated road map for the future of Amplitude and Statsig platforms together. We will work closely with the Statsig team at OpenAI during this transition.

As context for the move, AI has dramatically lowered the barrier to building and shipping software, boosting productivity for experienced engineers and enabling nontraditional roles to become AI builders. While teams can generate more code than ever before, the software development life cycle remains bottlenecked in many other places. AI builders are generating code faster than they can understand its impact. The challenge is now evaluating code before it's released, tracking what's working after release, knowing when you need to roll things back, and turning behavioral signals into what to build next.

Amplitude is the market leader and is focused on giving the best behavioral insights to product managers. Statsig has reinvented experimentation and feature management and done an amazing job with data leaders with its warehouse native capabilities. Together, we can accelerate the software development life cycle. We now offer organizations access to the same capabilities that the world's most advanced AI companies use today.

Initial customer feedback has been promising. Many of our existing customers have already expressed interest in the Statsig product. The pace at which Amplitude builds and ships products continues to accelerate. Over 90% of the code our team ships today is written by AI.

I want to show you 4 quick demos today, each one reflecting a different dimension of what it means to close the product development loop.

I want to start with Agent Analytics. Everyone building agents has one big question. Are they working? With Agent Analytics, customers get complete visibility into every agent interaction, see every conversation's full thread, what the user asked, what the agent responded, which model was used, how many tokens it burned and how long it took to complete. Once the conversation completes, evaluators automatically run and judge your agent's performance across dimensions like user satisfaction, agent confusion, response quality and task completion.

This happens on every conversation and is fully customizable so you can build evaluators specific to your use case. Then you can put all your Agent Analytics data together with your customer data. You can see how your agent's performance directly connects to real customer events, like what impact an original agent interaction can have on a customer later completing a purchase.

We have been shipping faster around our Amplitude agents. We've added productivity updates to our Global Agent, including voice to text input for natural language prompting, image upload for providing deeper context, searchable chat history and conversation history across projects. In addition to that, we've also added memory, so agents now monitor when they're corrected or directed in specific ways and save that for the future. For example, a weekly active user in Amplitude is a user who saves a chart, not someone who simply logs in.

The agent, after telling the agent this, it remembers it for future analysis instead of needing to be corrected every time. 90% of these memories are automatically created as people use agents, so agents get smarter and better the more people use them. We also have MCP connectors built directly into agents. Agents are incredible at analysis, but the connection to action is broken. A human still needed to file the linear ticket to the write up a notion or read the slack channel for context, not anymore with MCP connectors. All of these actions can be triggered automatically.

Non-event data can now be paired with Amplitude's data, connect financial information from BigQuery and quantify the real cost benefit of an experiment or with GitHub and Amplitude, retroactively track how specific releases affect error rates, session length or feature adoption. Now our agents can run and connect to all your data sources to surface insights and deliver the context wherever it's needed.

This is exactly what one of our large financial institution customers experienced. They had agents running on Amplitude surfacing insights for a new interface rollout they are planning. One of those agents surfaced pages that were indexed incorrectly before they went live without being instructed to find the inaccuracies. That helped the team avoid serving incorrect data to customers without even being asked. That is what it looks like when the loop closes on the right side automatically. People don't check dashboards. The system catches the problem before it become one.

We're also building new AI products to expand our platform for customers. Our most recent launch was AI Assistant. AI Assistant is a chatbot that answers customers' questions in real time like Intercom Fin. It's tied into Amplitude so it can know who users are, where they've been previously and where they are right now. If users want to know how to accomplish a task, instead of giving text instructions, it can create a visually guided tour that walks users through the interface. Here, I'm asking how to integrate with Slack and it's triggering a guide that helps me do so. It shows me where to click on the screen and guides me through the process. This is live for customers to purchase today and is a great way to highlight how we're using AI to infuse context and understanding of the user for our customers.

The last demo I want to show with you today is our command line interface Wizard. AI builders need an automated installation of Amplitude. That is why we built the CLI Wizard. Setting up Amplitude used to be a sticking point for some users in the past. Now with our CLI Wizard, it's one line of code in the terminal. The rest is done for them. The CLI Wizard package runs against any code base, any programming language and it instruments Amplitude for you. It adds SDKs, creates the taxonomy and instruments all events and configures MCP. It will even create an initial dashboard for you. What used to take weeks now takes a few minutes, all initial setup into one action, dead simple install for humans.

After this, we're going to give the ability for agents to install Amplitude automatically in the cloud. There is now no barrier to installing Amplitude.

Let me tell you about a few customers who are putting this to work. Granola is one of the fastest-growing AI companies out there. They came to Amplitude before they had actually, before they had even launched because they wanted to understand from day one whether what they were shipping was actually working for users. Today, more than half the company uses Amplitude every day. And actually, I think everyone Amplitude is a granola user. They ship new features fast and rely on real-time behavioral signals to decide what to do next. They have grown with us horizontally and use the full platform.

Granola is what a next-generation software company looks like. No separate analytics team, no weekly reporting cycle. The loop from ship to learn runs continuously and Amplitude is the infrastructure that makes it possible.

Smartsheet is an intelligent work management platform that helps enterprises unite people, data and AI to turn strategy into results. As Smartsheet accelerated its push into AI-driven experiences, the team faced a real bottleneck. Their product managers were entirely dependent on the BI team for every single insight. A question as basic as how many people use this feature last month and what does this mean for retention could take weeks to answer.

Today, with Amplitude Analytics, feature experimentation and guides and surveys, Smartsheet's product managers, engineers, designers and researchers have that answer instantly. They've used those insights to identify and fix drop-off in their onboarding funnel with a direct measurable impact on retention. As Smartsheet invests in AI, Amplitude gives them the velocity to understand whether new experiences are working at the speed their ambitions demand.

Astra Tech chose Amplitude as its partner to support Botim & Botim Money's evolution into an AI-native fintech super app for over 150 million users across 150 countries. Botim uses insights from Amplitude to optimize fintech entry points, pinpoint critical journey drop-offs and establish clear engagement baselines for Botim AI across user segments, usage patterns and downstream actions. With cross-functional teams in growth, design, tech, using those insights to steer a completely revamped Botim to reposition itself as a fintech-led communications platform.

I'm very excited to share the business impact we had with them. Their revamp across services, including international transfer, local transfer, ad funds and gold, Astra Tech increased fintech service entries by 4%, lifted engagement from top offers and 4U by up to 3% and grew fintech transacting users by 3x. That happened all within a span of 9 months of working with Amplitude.

I want to note that the companies on the bleeding edge of the tech industry are Amplitude customers. That's because the faster you build, the more you need to know what to build next. AI natives understand that better than anyone. This underscores the long-term case for Amplitude. AI makes what we do more critical than ever. We are set up to close the right side of the product development loop, and we have the platform, the customers, the leadership and the conviction to see it through. I'm extraordinarily excited for what's next.

Now over to Andrew to walk you through the financials.

Thank you, Spenser, and good afternoon, everyone. The first quarter was solid with incremental improvement in our dollar-based net retention to 106%, multiproduct accounts for more than 77% of our ARR and our ARR growth was 17%. We beat our guidance on both top and bottom line, and we are combining the best of Statsig with Amplitude.

Reflecting on Q1, there were many changes in our go-to-market team. We've introduced a number of new AI products, and Amplitude has been implementing a host of new AI-based workflows to drive efficiency. We are in a moment of transformation. We are transforming the value our customers receive. We are transforming how we deliver value, and we are transforming our organization from the ground up. We've done this while continuing to execute on our core business.

We are leveraging AI at scale across our organization and helping customers unlock incremental value faster. No longer is a good piece of code with a friendly UI good enough. We must deliver customer valued outcomes. We are focused on becoming a true partner with our customers to understand how to apply technology in the most effective ways. We are building on a decade of understanding context of delivering this knowledge through our services, our platform and our know-how.

The speed of change is accelerating, and we're leaning into that moment. We're seeing increased usage of our AI agents along with data ingested into our platform. This has created some headwinds in our cost to serve, but it's also aligned to our monetization strategy. Adapting quickly and delivering greater value to our customers will be the advantage of the next generation of winners in software, which is why we've made changes to our products, pricing and internal operations.

Taking on the Statsig business is another great example of our ability to be flexible and act quickly. By combining Statsig's industry-leading warehouse native experimentation with Amplitude's best-in-class analytics platform, we're expanding our total addressable market and meeting customers where their data needs are. We will build this business to be incremental and accretive to our core business.

Spenser highlighted some of the changes our team has undergone, and we're instrumenting the business for long-term scale and efficiency so that driving business growth continues to result in greater leverage. That being said, our goals as a business remains steady. We want to grow our enterprise business, expand our multiproduct footprint and deliver great value for our customers.

This focus has enabled us to drive consolidation in the market through our platform approach, now having over 77% of our ARR coming from customers with more than 2 products, up 3 points from last quarter. Customers with 5 or more products now account for 24% of our ARR, up from 20% last quarter.

We believe that as customers continue to adopt our AI products, they will naturally expand their use cases into the full suite of our platform and drive incremental upsell opportunities.

Turning to our first quarter and full year results. As a reminder, all financial results I will be discussing with the exception of revenue are non-GAAP. Our GAAP financial results, along with a reconciliation between GAAP and non-GAAP can be found in our earnings press release and supplemental financials on the Investor Relations page of our website.

First quarter revenue was $93.5 million, up 17% year-over-year versus 10% in the first quarter of 2025. Total ARR increased to $374 million exiting the first quarter, an increase of 17% year-over-year and $9 million sequentially.

Total remaining performance obligations grew 31% year-over-year to $427 million compared to 30% growth in Q1 2025. Current RPO was up 20% year-over-year compared to 18% in Q1 of last year. Long-term RPO was up 60% year-over-year compared to 72% from the first quarter of last year.

We had a strong quarter for both new and expansion deals in the enterprise. Platform sales were also particularly strong. 47% of our customers now have multiple products with 77% coming from that cohort. We have made great progress on expanding our multiproduct footprint within our customer base compared to a year ago when only 30% of our customers had multiproducts and accounted for only 64% of our ARR.

The number of customers representing $100,000 or more of ARR in Q1 grew to 727, an increase of 18% year-over-year and up 29 customers since the last quarter. In-period net dollar retention increased to 106% from 105% last quarter, led by cross-sell expansions across our customer base. We expect net dollar retention to improve over the long term as we continue to see customers adopt multiproduct.

However, it may not be in a linear fashion. Gross margin was 75% for the first quarter, down 2 points from the first quarter of last year. This was largely driven by growth in inference costs as adoption of our AI tools by our customers outpaced our expectations. We now expect this adoption trend to continue given the feedback we received from our customers.

In the short term, this will cause gross margin compression, but we believe this will help us to drive greater data ingestion and monetization of our core platform over time. Sales and marketing expenses were 45% of revenue, in line with the first quarter from last year. Some of the increase in costs included severance costs related to our organizational changes and other activities like our go-to-market kickoff that occurred in the first quarter.

We have focused our entire go-to-market team on driving value for our customers, increasing adoption organization-wide and improving our internal processes, coverage and expanding the buyer personas that we can sell to. These changes will take time to manifest in net new ARR, but ultimately, they will increase the health of our customer base and drive greater opportunity to grow our net dollar-based retention.

R&D was 20% of revenue, up 1 point from the same period last year. We will be adding to the team to scale the Statsig opportunity and continue to support those customers. G&A was 13% of revenue, down 2 points from the first quarter of 2025, and we expect G&A to improve as a percentage of revenue over time.

Total operating expenses were $73 million or 78% of revenue, down 1 point from the same period a year ago. Operating loss was $3.1 million or 3.3% of revenue. Net loss per share was $0.02 based on 133.3 million basic shares compared to a net loss per share of $0.00 with 129.7 million shares a year ago.

Free cash flow in the quarter was a negative $13.2 million or negative 14% of revenue compared to a negative $9.2 million or negative 12% of revenue during the same period last year. We continue to be active in the open market last quarter, retiring shares against our open buyback. We have conviction in the long-term value of our platform and have used and will use our cash to minimize the impacts of dilution while our share price continues to not align with the value we believe we're creating.

Our balance sheet position remains strong and allows us the opportunity to be more aggressive in our M&A strategy to accelerate our R&D road map when appropriate. In Q2, we will also take into consideration bringing the Statsig customers and technology over to Amplitude as of the beginning of May.

To start, we will record an additional $16 million in incremental ARR from the Statsig customer base, aligning that business to our definition of ARR. As we take on the Statsig business, we will also be investing in the transition team as we ramp an internal team to continue to provide the best support for the Statsig customers.

Over time, we will scale our internal team to continue to develop the warehouse native and cloud aspects of Statsig. Additionally, there will be some pressure on gross margins for the remainder of the year as we integrate and optimize our hosting environment.

Now turning to our outlook. As a reminder, the philosophy of how we set guidance is through the lens of execution. We are pleased with our progression on driving adoption of our core platform, our different AI technologies and multiproduct adoption. Our new pricing and packaging rollout is progressing very well. And in the first quarter, 25% of total ARR contracted, both new business and renewals was on our new pricing and packaging.

We will continue to increase this percentage as we make it easier for our sellers to quote and make it easier for our customers to understand the path to platform adoption. We are already seeing early signs of willingness to test new features and products on the platform given the easier on-ramp from a contract view. This will also lend itself to allowing easier adoption of our AI agents as we continue to iterate and ship.

So, for the first, for the second quarter of 2026, we expect revenue to be between $96.9 million and $99.1 million, representing an annual growth rate of 18% at the midpoint. We expect non-GAAP operating income to be between negative $3.6 million and negative $1.6 million. We expect non-GAAP net income per share to be between negative $0.02 and negative $0.01, assuming basic weighted average shares outstanding of approximately 134 million.

For the full year of 2026, we expect full year revenue to be between $397 million and $403 million, an annual growth rate of 17% at the midpoint. This assumes a $5 million to $7 million contribution from the Statsig business, taking into account the assumption of the customer contracts and the impacts to deferred revenue.

We expect our full year non-GAAP operating income to be between $2.5 million and $6.5 million. This reflects incremental investment we'll need to incorporate the Statsig business. We expect non-GAAP net income per share to be between $0.03 and $0.06, assuming weighted average shares outstanding of approximately 145.1 million as measured on a fully diluted basis.

In closing, we are accelerating our pace of innovation, and we're growing the value that we can deliver to our customers. We have confidence in our ability to scale a durable and growing business while also bringing Agentic Analytics to the world. With that, we'll open it up for Q&A. Over to you, John.

Thank you, Andrew. We will now turn to Q&A. [Operator Instructions] Our first question will come from the line of Taylor McGinnis from UBS, followed by Rob Oliver from R.W. Baird.

Maybe first, Spenser, for you. Could you just maybe explain why OpenAI is foregoing the Statsig business? And if there's any parts that OpenAI is retaining in that? And then, Andrew, maybe a second one for you. Helpful color on breaking out some of the Statsig impact this year to the guide. If we strip that out, does that mean that you're taking down, I guess, the organic growth guide a little bit on revenue this year? And maybe you could just unpack that and the margin impact.

So, first, just to answer the question on the Statsig side. I mean, Vijay and I have known each other for years. Amplitude and Statsig have been competitors and kind of pushing the bleeding edge in their respective niches. I'm extraordinarily excited that we get to kind of carry a bunch of that forward with both the customers, the technology as well as the brand.

I think Vijay was looking for a home for the kind of continued support of the Statsig customer base. And after looking at a number of different places, him and I agreed that the best place that would be Amplitude. We just executed that agreement on Friday. So, we're kind of still just getting up to speed with all everything it entails and making sure it's a smooth trend, making sure those customers continue to be supported and then figuring out what the long-term plans for Amplitude and Statsig are together.

But I'm just, I'm very, very, very excited about it. OpenAI will be continuing to run the technology internally that they have from Statsig, and so they'll be continuing to use it, but that will obviously be supported by Vijay and the existing Statsig team at OpenAI.

And Taylor, part of the guidance we have is incorporating the accounting associated agreement like this, where you have to take a fair value assessment on the revenue that's aligned to the annual recurring revenue I mentioned. But by taking that fair value assessment, you actually take a haircut on the value. It actually reduces down.

And so what you're seeing in the amount I'm indicating that comes from ARR and the lower revenue is really related to that fair value assessment. And I would tell you that we had a good quarter in Q1. We beat expectations. We beat what our guidance was, and that's flowed through into our guidance for FY '26. So, for us, we think it's a huge opportunity for us to go build out a great product that a lot of customers will be very interested in.

Perfect. And just a quick follow-up, if I may. So, if I look at the net new ARR numbers, it looks like maybe it was flattish on a year-over-year basis. I know you mentioned that there were a number of changes that you guys made in the quarter from leadership changes to pricing and packaging. So, did that at all have any impact in the quarter? And maybe you could just talk about what occurred and how you guys are thinking about that metric for the remainder of the year?

Whenever you make big changes in organizational structures or you're making changes in core processes, invariably, there is going to be an impact. I would tell you that we're pretty proud of the fact that given those changes that we made, we were still able to, one, exceed the guidance we had put out with respect to ARR and turn in a pretty good quarter, especially with respect to net dollar retention increasing, the number of $100,000 customers we added. So yes, there's always going to be some impact, but we did a pretty good job of kind of executing through it.

Our next question will come from Rob Oliver at R.W. Baird, followed by Jackson Ader at KeyBanc.

I apologize for background noise. I'm out in the wind here a little bit. Yes. So, I guess first question, Andrew, for you. Really great progress on the new pricing model. I mean it feels like just yesterday, you guys were in pilot on that, and now you're at 25% of ARR. I guess a couple of questions there to start.

One, how should we think about the progression of that? I think you said it's key to the selling motion. But is that something as customers come up for renewal this year, we can expect that number to continue to move higher? And any, recognizing it's very early, any early indications on what kind of pricing uplift or impact it's having on the contracts in terms of the combinations of usage? And then I had a quick follow-up as well.

Yes. I'd say we're pretty pleased with our progress on the pricing and package as well, Rob. The response from our sales team has been tremendous. They love the simplicity in the way they can actually express value back to clients. That proxy on value from a price perspective and the methodology is one that customers really understand.

I can tell you there are a number of deals in Q1 that customers added more product associated with our platform because of the simplicity and the way they get cost predictability on that new strategy. So I do expect that the percentage of our ARR that's going to be on the new pricing and packaging will increase. We're not going to force customers through hard migrations. We're going to give them the carrot and show them the value, and we expect that customers are going to really want to adopt the new pricing and packaging.

Great. Helpful. And then my follow-up, Spenser, in your prepared remarks, you made it clear that being an AI company right now is the most important thing. And I guess that creates a ton of exciting opportunity like around Statsig. It also creates some, a fair amount of uncertainty around both the gross and operating margin line.

So just wondering, I know we've got updated numbers for you guys and thoughts around that. But recognizing you just closed Statsig on Friday, can we expect at some point, perhaps this year, we'll get updated thoughts from you guys around cost to integrate go-to-market and potential further impacts both on the gross cost to serve side as well as on the operating margin side.

For sure, for sure. So yes, I mean, again, it's a few days old, so we did our best with the guidance that we put out, but we'll absolutely have a much better picture as we get into next quarter and subsequent quarters. Let me talk about Statsig specifically, and then I'll talk about more generally on the AI transformation of Amplitude.

On Statsig specifically, I think it will be long term, very accretive to the business. A lot of Amplitude customers are very interested in their product. A lot of Amplitude customers are also Statsig customers, like we just talked to one yesterday, Atlassian, who is a big user on the Statsig experimentation side while being a big user on the Amplitude Analytics. And they're actually really excited because now the data from the two products will talk to each other, and that will drive a whole bunch more value and usage and good things for both Atlassian and Amplitude. And we expect to see similar things across the entire customer base.

Now in terms of exactly quantifying them, again, it's very rough in the air because it's only a few days old, but we'll have a much better picture into it come next earnings call. In terms of AI generally, it's absolutely going to be, so I think on operating margin leverage, it absolutely will be accretive. People get more efficient. We'll be able to get a lot more done with the same number of people.

We'll be able to have 2x or 3x the impact without having to grow the team. And so, I'm extraordinarily excited about that. The mistake I see a lot of companies making is there, like we just said, hey, just go and be aggressive on your internal spend. I see a lot of companies that's like, oh, let's only have like a $200 budget per person for AI spend. And it's like that's nowhere near unleashes the full capabilities. I mean you see some of our top engineers that are shipping 5x the amount of pull requests, but they're also spending thousands of dollars a month or more on tokens.

So, we're figuring out exactly how to budget and price it out, but we absolutely expect that will translate to operating leverage long term. The last thing to call out is as part of using these products, there is inference spend. So right, if you're using Global Agent and you're asking Amplitude, hey, find me what's the cause of this drop in this conversion funnel, like that's going to cost us a bunch of tokens and all of that.

Now for the here and now, we've said, hey, let's just support it, and we're going to bundle it in with our core stuff because that just means more Amplitude usage. And you can see that in a little bit with what Andrew shared with the gross margin numbers. And so that has, that does put short-term pressure. But we, again, expect that to be accretive long term, most importantly to revenue growth, but then also to operating margin as it requires less people on our end to support more customers.

Just one clarification to Rob. I'll tell you, we did take into consideration the operating expenses associated with the Statsig business into our guide.

Our next question will come from Jackson Ader at KeyBanc, followed by Clark Wright at D.A. Davidson.

The first question I had, Spenser, on the command line interface and the MCP server that you're kind of turning live, making it frictionless, right, to actually adopt and use Amplitude. But if I think about the other side of maybe it's enterprise customers where you're having forward deployed engineers, right, who are ostensibly, I would think, trying to like make sure you go hand-in-hand with customers and make sure that they are adopting things. So those two things, like the frictionless and the forward deployed engineers just seem not an odds, but just like.

Yes, a little bit different. You need a one-line thing. "Why do you need someone to teach you? " Okay. So I think a few things. One is that a lot of the, by the way, I love the question. I love that you pay attention during the demos because not everyone does. So, thank you.

So, first on the frictionless, it's so much of the process before to get set up with any data system, including Amplitude was extraordinarily manual. You have to define your objectives. You have to define a taxonomy, you put in the SDK, you put one line of code wherever you do it. You have to then create a bunch of charts and dashboards on the basis of that. And so I there's tremendous opportunities to automate that with AI. And so that's what we did with Amplitude CLI with all of that installation process. It's actually pretty wild. I didn't, 3 months ago, that wouldn't have been possible. And so it's really cool to see it possible today.

Now the flip side is what we see with customers is every single one of them is looking for education on "how do I adopt AI". Arguably, you could say that all the growth in these AI natives, if you look at the private markets, comes from companies that are just doing a really good job of educating customers. It's actually no longer a technology bottleneck; in that the models are getting so fast, software is like there's, it's very easy to build. And so, the customer is choosing on, "okay, who do I trust to actually kind of get me through this. "

So, if you're an AI-native bleeding edge, hey, it's just like, "Give me the one-line CLI and I'm off to the races. " But if you're a traditional company, like, say, Fox Broadcasting or Walmart, you're going to want a lot of hands-on help from an Amplitude to make sure you educate your team. It's one thing to just have a bunch of software running and you can get that from anywhere, but it's another to say, "Hey, educate me on how to use analytics from a bleeding edge AI standpoint, what the future is going to be and help me reskill the hundreds or thousands of people I have with my organization. " And that you just need a human touch to do.

So it's, the forward deployed engineers are much like, yes, there is a, "okay, well, why do we even need that if we have the one line of code, " but actually getting adoption of Amplitude or any software product within the enterprise is much less about like do you have the widget or the specific feature and it's much more about, "hey, are you going to be the best person out there to educate my organization on what the future of this technology is. "

Okay. All right. That makes sense. The follow-up question I had is really, I guess, for both of you. I'm just thinking like you're shifting to an AI-first company, right, which has come from a lot of personnel, which has manifested itself in a lot of personnel changes, leadership changes, personnel, we're changing pricing and packaging, right? Now doing like an acquisition, right, like this integration of another product. So there's a lot going on. What is your plan to make sure that execution does, like execution risk doesn't bubble up with so many balls here.

I mean, look, just to be very candid and direct, I think vast majority of SaaS companies are being way too conservative with the change. And I've taken the opposite approach where it's like, look, market has spoken about its opinion of what the future is going to look like. We know from talking to customers what they want. We see the innovations from a technological standpoint. And so we want to run as fast as possible to where the puck is going on all of this stuff. As part of that, acknowledging it is going to be bumpy and it is going to be chaotic and there will be things that we don't expect or can't perfectly plan for in advance.

It is much more important to get there with a lot of speed for a lot of different reasons than it is to say, "Hey, let's try to protect some existing thing we have. " The existing thing we have, frankly, isn't valued much. And so what's much more interesting to me is can we generate billions of dollars in revenue in this new world. And so whether that is changes on the organization in terms of leadership, whether that's changes on functions and roles, whether it's changing on product, on pricing, on working with partnering with OpenAI and Statsig through the future of Statsig, like we're just going to be really aggressive on making sure we reinvent the whole category.

In my mind, the same thing that happened in the coding space over the last two years, where it just looks fundamentally different today than it did two years ago, that is going to happen in our category with analytics, experimentation, session replay and the whole thing. And so, it's a race to see who can do that the fastest. And so that's what I'm really focused on is not the close to $400 million in ARR that we have. I'm much more focused on the billions and potential in the future that are going to be created.

Our next question will be from the line of Clark Wright, followed by Scott Berg at Needham.

Any update on the ramp of events in the pricing curve that you've implemented to help enterprises scale usage previously and ongoing?

Yes. So, one of the things that we were talking about, Clark, is that our new pricing and packaging that we rolled out, we did a lot of testing on. We had kind of a soft rollout this quarter. It was still one that was handheld because we hadn't implemented many of our systems related to doing the quoting or letting reps actually do the quoting themselves. That's all been now implemented, and we're seeing great responses back from clients. I think that they're appreciative of the changes we've made.

They see that as they add more events that they're getting a marginal incremental cost reduction from their perspective. But for us, it's always going to be increasing the ARR as events increase. And I think they like the simplicity of how they can quickly adopt the modules that are surrounding analytics. Enterprises want cost predictability that they can align back to what their value propositions are. And as our sales team becomes more adept at showing and delivering what customers will get in value from Amplitude, I think that the pricing and packaging changes we've made will really reinforce their ability to move at pace.

That's helpful. And then one of the other things that you noted during the prepared remarks was the TAM expansion with Statsig. Can you explain what budgets you're going after? I think the other piece that was consolidation that's unlocked with this partnership? And what could you do with that, that you couldn't as a stand-alone entity?

So the thing that, I mean, it's all kind of, there's overlapping. So, it's not like we don't have any of it. But they've done two things extraordinarily well. One, experimentation. The bleeding edge teams in AI are using them for experimentation. Like I don't know if you ever use ChatGPT, but if you ever get those like, hey, do you prefer prompt A or prompt B, that is stats internally at OpenAI powering that. And so, we're really excited that we get the opportunity to offer that out there just broadly to everyone.

The other thing that they've done extraordinarily well is work with the data leader and specifically their data warehouse architecture. While we obviously, we've done that a bunch of Amplitude, I mean, they are definitely the bleeding edge on that, where they actively both allow you to query on warehouses directly as well as run experiments and a whole bunch of other infrastructure. And so that's really exciting for us because especially at some of these larger, at the largest customers, when you start getting into the multimillion-dollar range, we often see this category of functionality owned by the data leader. And as part of that, we're excited to get much closer to them and unlock a lot of data warehouse and data warehouse adjacent budget.

Our next question will come from the line of Scott Berg at Needham, followed by Nick Altmann from BTIG.

Spenser, I want to talk kind of an architectural type question. With the pressure on gross margins, how have you thought about things like Open Source models or some small language models being used within the broader Amplitude platform versus maybe some of the frontier models that you're using with reference and such today.

There's a very large private software company that kind of announced a large, what we'll call, Open Source model and their new platform. It was really intriguing in terms of what they're doing with it. I'd love to hear what you have all considered through that process.

Yes. We're early on this to be clear. Inference spend is growing quite a bit, and you see that reflected in a few places, both the operating margin guide and the gross margin guide. Now ultimately, what we see from customers is, in most cases, they want the bleeding edge thing. So they want the latest Sonnet release or the latest Codex release from OpenAI or Anthropic or one of the other providers.

And that also leads to higher scores on our benchmark. Like last quarter, we published a benchmark where we got a 76% accuracy rate. And we could downgrade that in some places, but you'd probably be looking at maybe a 40% accuracy rate, and that's a pretty significant difference. Now over the long term, we'll obviously find places to use cheaper, more effective models where it makes sense. But in general, and we'll sort out the path on gross margin as part of it. In general, right now, we're just in the, "hey, like let's make sure we win the market first and foremost, " and then there's optimization down the line that comes with that. So, we do expect our ability either with the Open Source models or some of the cheaper models like if you look at Anthropic's Haiku model, that's a really great one that for actually a good chunk of cases actually works decently well.

But again, when you're doing some of the complex data reporting, especially with Amplitude, we see a lot of customer demand for high accuracy, and that means the most bleeding edge ones, and there's always a new one release. Now the good news in all of this is this, the curve on this is crazy. I mean you're seeing a factor of 10 or more improvements in the price performance of these models year-on-year. And so, it's hard to say exactly what it's going to be in 12 months from now. But I know we're, it's going to be a lot better, and that means we can choose, okay, exactly where it makes sense on the price performance so we have reasonable gross margins.

Understood. Very helpful. And then, Andrew, I wanted to dig into the Statsig acquisition a little bit more. I think when OpenAI acquired that business, they're doing about $40 million worth of ARR. Are you, I guess, saying or implying that the balance of the $16 million that you're bringing over versus that $40 million is effectively staying with OpenAI? And then I guess, did you happen to pay for any part of this business? Didn't know if there's a purchase price cash or stock, some sort of allocation that's committed to this that any of that any of that information we can?

Now a couple of things you need to understand about Statsig's former business prior to OpenAI acquiring them. OpenAI was a fairly large customer for them. And that was a substantial portion of their ARR, right? So, then your question is what is OpenAI's intention with Statsig? It's back to what Spenser mentioned earlier. They're intending to use it for internal noncommercial reasons, and they're continuing to use it to support their core products.

So, as we go forward, what we've taken on is the customer contracts, all of them. And we're taking on all of the brand assets, and we are increasingly going to be developing on the product itself. So, delivering great solutions for those clients and future clients, frankly, of Amplitude.

One minor technical point, we're also talking about Statsig as a partnership, not an acquisition. So it's nuanced, but yes, it's an important one as well.

Our next question will come from the line of Nick Altmann from BTIG, followed by Arjun Bhatia from William Blair.

Awesome. Andrew, we appreciate the color on the Statsig contribution. But you guys, you've kind of talked about this before, but there's overlapping customers. There's overlapping product sets. At the same time, you guys have kind of also made an effort to consolidate your customers onto more of the Amplitude products. And so in terms of that revenue contribution framework that you outlined, what does that sort of imply for those customers where there is overlap working with both you guys and Statsig and on the product side and on the customer side? Just any other details you can kind of unpack in the assumptions would be helpful.

We don't think, we don't actually see that there's huge overlap in the products that customers are using from Statsig and us. As Spenser mentioned, they're really good on experimentation. And they may, you may find customers where we did have overlaps if they were using analytics from Amplitude, but experimentation from Statsig. So there really isn't like overlapping revenue. It's an opportunity for us to actually add more and more to a consolidated platform.

So if they didn't have such a replay or guides and surveys. In fact, we see a huge opportunity for us to go sell into the customer base that is overlapping. Not to mention, there's a whole new group of customers that Amplitude now has access to.

Yes. It's not, Nick, I'll say it's not the cleanest like, okay, yes, theoretically, we both have experimentation. There's has been developed in a little bit of a different way. So it's a little bit of a different customer set. So there is a lot of great opportunity across both customer bases. Again, we're kind of three days into this thing. So we're still sizing that, and we'll have more detail when we go through on the Q2 earnings call.

Understood. And then the NRR continues to accelerate. The ARR growth remained at 17%. So Andrew, can you just maybe unpack why we're seeing those two metrics disconnect? Is it something on the new logo ACV side of the equation? Is it gross retention dynamic? Maybe just talk to us about the disconnect between those 2 metrics.

Yes. So in any given quarter, Nick, you have an amount of new ARR we're adding in new logo versus expansion. In fact, a year ago, you probably remember, we talked about having a really strong new logo quarter. and not making as much progression on net dollar retention. I would say in Q1 this quarter, you saw the opposite effect. You saw really good expansions happen. That may have been partially due to the really strong new logo quarter we saw in Q4.

And in Q1, it shifted more to an expansion quarter than we anticipated. And every quarter, we try to take a look at what is the real balance of our pipeline between new logo and expansions and do our best to estimate what the impact is going to be. It wouldn't surprise me given what we've seen in Q1 that you might see a shift into Q2 where there is more new logo versus expansion. So that's why I was commenting that in some cases, we're seeing quarter-to-quarter progressions on our long-term plan.

In other quarters, you might see it not improve as much, and it's really related to that balance. Is it overweighted in new logos in any given quarter? Is it overweighted in expansion? But the long-term view is that we're continuing to ship new products and add to the value that customers can purchase from Amplitude, which sets us up for additional expansions.

Our next question will be from Arjun Bhatia from William Blair, followed by Billy Fitzsimmons from Piper Sandler.

Perfect. I want to go back to the sort of the inferencing costs increasing. And Spenser, I'm just curious where the AI usage is coming in strong. And you've made a lot of sort of enhancements to your MCP server. And I'm curious if that's also driving a meaningful change in how your customers are using the Amplitude platform.

Yes. A lot of customers are using MCP. We're seeing huge uses of that, huge uses of Global Agent, a good chunk on specialized agents. That's kind of the bulk. Those all kind of hook up to the same underlying services. So you can say, hey, find the root cause of an issue in this chart and that can either come in through MCP, come in through Global Agent, which is a chat interface or come through specialized agents, which is purposely designed. All of those combined, they're what is the vast majority of the usage of those is the vast majority of what the inference costs are.

Got you. And then just your, I mean, your comment, and I think we see it broadly as well that the software development life cycle is changing very quickly. Obviously, it's starting in just code gen and code writing. And what is your, what is your perspective on the steps that will be required before you start to see it in your category? Like does the fundamental composition of the software team or the ops team need to change? And is there more change management ahead of us, I guess, before we see sort of this hockey stick in analytics and monitoring and experimentation. Like it's just more of a philosophical question, but I'm curious where we are in this cycle in your mind?

It's early, yes, to be clear. So it's hard to prognosticate on exactly when. I'll tell you the best, so I'd say let me talk about tech and then I'll talk about non-tech. Within tech, they're already embracing this in terms of the automated instrumentation, in terms of automated analysis, automating the product development life cycle. You might have seen blog posts or Twitter posts about how companies are having an agent harness and then automatically building software. And so that's awesome.

They're kind of already living in this future. And a lot of the companies we work with Granola is a great example I called out, are pushing us on that being like, hey, here's what we want to be relevant. Agent Analytics was one of the outputs of that. And we're in the early days with it. We just announced it, I think, 6 weeks ago, something like that. We had the first post about it 6 weeks ago. And so we're early days in the adoption, but very exciting. A lot of these companies are living in the future, and now it's on us to like, okay, let's make sure we go capture it.

On the non-tech company side, it's much, much earlier. They're looking to get educated on the basics and hey, show me a maturity model, okay, if I'm not step 1, how do I get to step 2? And then, okay, I get like eventually, you guys will help me get to step 4 on having an agent harness and automatically building software, but that might be a few years away. And instead, they're just saying, okay, well, at least let me take some of my existing workflows and speed them up a whole bunch so that I don't need to do all this instrumentation. I can just use your CLI wizard or I can use the chat Global Agent in the chat interface and have it generated a dashboard for me and now I'm saving time.

So they're more focused on making a bunch of existing stuff more efficient than going straight to the bleeding edge. So anyway, early days, I do expect in the next few years, it will look substantially different. But hard to say exactly which quarter and when.

Our next question comes from the line of Billy Fitzsimmons from Piper Sandler, followed by Koji Ikeda from Bank of America.

I'll keep to one because I know we're getting close to the end of the call. Spenser, for you. I appreciated the commentary in the prepared remarks on some of the new members of the C-suite. With Nate coming in as the Chief Commercial Officer, I want to double-click there because it seems like you're seeing some solid go-to-market progress in some of the initiatives you've already been doing, multiproduct adoption. Any additional color you can provide on his plan or top priorities going in and expected changes to sales motions, sales incentives, partner strategies and just general changes relative to what you've kind of already done?

So yes, let me talk about the transition, and then I'll talk about going forward. So Thomas, our prior President, ran all go-to-market, did a phenomenal job over the last 3.5 years, really upgraded us to an enterprise company. Before, we weren't even engaging with executives consistently. Now we are, and we have all the services across the board to support them, and that's been fantastic. And you've seen that show up in terms of $100,000-plus customers, platform adoption, a whole bunch of other metrics.

Nate actually, as our Chief Commercial Officer, was previously our Chief Revenue Officer and was in place reporting to Thomas for the last 3 years. So he's not actually, so that part is not actually new when he was running sales. The change in his role now is he's now running the post-sales motion as well as revenue operations and enablement. And so those, we're thinking about, okay, how do we streamline to make sure that is seamless between the AEs and then the technical success managers and all the parts of the post-sales motion.

In terms of going forward, I had a slide on this briefly, but, so a few different things. One, just more technical talent in post sales generally. We've renamed it from customer success managers, gotten rid of a bunch of the extraneous roles and just called them technical success managers. We expect them to be able to educate prospects on the Amplitude product and platform, how to implement it, how to integrate it with their code base and be the go-to on AI when it comes to technical queries. We're also added 4 deployed engineers that can do a lot of the actual coding work. Now it's early in the function. We just spun up that group about a month ago, but that can do like actually code and hook up like a lot of the problem with AI adoption internally within companies is not just turn on the software, it's actually hook it up to the right systems.

So, make sure it's plugged into the back end, make sure the SDK is in, make sure the events are instrumented, make sure that the right people are getting reports. We now introduce the MCP clients to make sure it can hook up the Slack and Linear or Atlassian or Jira or what have you. And so that's what the forward deployed engineers is, actually do the coding work on behalf or alongside the customer. So that change has been received. It's early, but that change has been received very well by customers.

Really, what they're looking for is expertise on how AI is going to change analytics as well as the associated functions within product management, data and everything related. And so yes, like they're really excited to learn. Actually, one of the funny ones is like we've even gotten a lot of questions from customers about our AI week. It's like, "oh, I want to do the same thing. How do I do it? " And so even if it's not necessarily one-to-one amplitude related, just having people there that are able to speak about here is how you can upskill yourself with AI and transform your organization, that's the value. They see a lot of value in it, and so we want to make sure to provide that in the post-sales motion.

Our next question will come from the line of Koji Ikeda from Bank of America, followed by Elizabeth Porter from Morgan Stanley.

This is George McGreen on for Koji Ikeda from BofA. I kind of wanted to ask as a follow-up to that last question on go-to-market changes. And forgive me if I didn't hear it, are there any sort of tweaks being made as it relates to sales incentive comp?

And then kind of as an unrelated follow-up, I'd be interested to hear if there's any update. I believe last quarter, 25% of queries on the platform were coming from AI agents. And if there's any update to that number, I'm sure it's growing healthily. And yes, just kind of like what's kind of the outlook and trends there?

For sure. In terms of sales incentive comp, we've had incentives in place both for platform adoption as well as for multiyear, and those are continuing. I mean we'll look at those, tweak those every quarter based on what it is we're trying to do with the business. Right now, like right this second, a lot of the priority is making sure that Statsig customers coming over are very successful, and we're able to continue to serve them and help them.

So we're, Nate and the broader sales team are very focused on that, and we've put in a few specific incentives around that. But we're always tweaking that stuff. It's not like there's a major massive shift in strategy there. And then in terms of the Agent adoption, it continues to grow, and there has been a significant increase. We're not sharing numbers on this particular call, but we will have an update next quarter on agent adoption relative to human usage of data analytics.

Our next question will come from the line of Elizabeth Porter from Morgan Stanley, followed by YC Wong from Citi.

I'm on for Elizabeth Porter. Spenser, you talked about seeing the puck as fast as possible early in the call. I just wanted to ask you to help us frame where your incremental AI-related investments are going this year, whether it's infrastructure, experimentation tools, workflow automation, like where is the puck going for you guys?

Specifically on expenses, the big one is inference costs. So to be able to support Global Agent, MCP, specialized agents as well as some of the other AI products, we've been spending quite a bit there. Those show up under cost of goods sold. So it's growing a lot. We'll monitor it and we'll figure out what the right long-term place for that to be and how do we value capture and get paid for it, too. But right now, we just want to drive adoption as the priority.

The other big place is on internal tooling for the team. The big one we're using a ton is Quad. So that was what we standard on an AI Week. I think Anthropic has done an amazing job with Claude and both the Chat Claude.ai as well as Claude Code in terms of integrating it with existing business systems. So we actually, you hook it up to Slack, e-mail calendar. We actually built Salesforce hasn't gotten their act together yet, I'm building MCP connectors. We built their own, built a few others. And so now you can access all Amplitude data through that interface. And so that's a significant spend internally. And then we also spend some on the team on Cursor as well.

There's a tail of them. Obviously, Granola is another one I mentioned on the call that is both a customer of ours and vice versa. But yes, those are the inference cost is the biggest one and then Claude for the team and then Cursor.

And our last question will come from YC Wong from Citi.

Just a quick one for Andrew, we just close in with AI. Last quarter, we talked about 25 AI customers like crossing over the 100 ARR mark, like if you're improving deep integration with the foundational models, the Cursor including, you are positioning Amplitude to capture a larger share of the AI market. Are these AI customers exhibiting like structurally different Net Retention Rate, consumption pattern that you're seeing compared to your traditional enterprise SaaS customer? And how do you view the opportunity longer term here?

So certainly, we believe that AI companies as they standardize on Amplitude will continue to see greater and greater value from using Amplitude and they embed and use it to drive marketing purchase, build better products, get better insights on how users are acting. And we're seeing, in some cases, some of the larger AI customers we have increase their data ingestion into our platform.

And one of the things we talked about on the gross margin headwinds was related to greater data ingestion. I think there was a confluence of that's all related to the AI agents. Well, there's also very classic cases where customers are using more of our capabilities and they're expanding their data usage and some of the AI companies are certainly exhibiting that. And one of the things that Statsig had in their customer base was a strong AI customer component. So we look forward to updating you all in Q2 on what that looks like with the combined product set and customers.

Thanks, YC. And that will conclude our first quarter earnings call. Thank you for your time and interest. We look forward to seeing you on the road this quarter as we attend conferences hosted by Needham, Jefferies, Bank of America and D.A. Davidson. Take care.

Awesome. Thank you, everyone.

Thank you.

All right. Well, let's get started. Good afternoon, everybody, and welcome to the Morgan Stanley TMT Conference. My name is Lucas Cerisola. I sit on the U.S. software team here at Morgan Stanley. And today, I have the pleasure of hosting Amplitude's CFO, Andrew Casey.

Andrew, welcome.

Thank you.

And before we get started, just a few disclosures. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative.

So Andrew, welcome. Amplitude has been an incredible company over the last couple of years evolving over the course of that time. And you've described 2026 as transformative with AI agents and new pricing. So at a high level, what business is Amplitude becoming over the next couple of years?

Well, I think it's -- increasingly Amplitudes becoming that infrastructure layer, that data layer and decision point layer that's bringing together behavioral heuristics, such you continue to see evolution in software and products and then increasingly, we're becoming more and more relevant into the marketing use cases. And I would say the customer experience use cases.

What I say all the time is that any customer who has a digital interaction with their client, they're going to want to know how that interaction is happening, how they can provide better services, provide better products and make those interactions more meaningful. The use cases around that are how do you drive greater -- free to paid? How do you drive better conversion associated with the loyalty programs? How do you drive better interactions and repeat purchases with your existing customers? And we have a pretty broad base of customer industries, everything from financial services to retail to fast casual restaurants to highly, highly technology companies like even one of the largest foundational AI models uses Amplitude.

Got it. And when you think about the next 12 to 18 months, what are the two or three key variables that you're most excited about across the entire business?

Well, when I started, we had aspirations to be a platform. And a lot of people gave us not a lot of credit for the ability to go build it. In fact, a lot of people said, "Look, everybody says that." And we also talk about major strategy moving more and more into the enterprise. And a lot of people said, "Well, of course. That's what a lot of companies that start off in SMB and mid-market want to do." But I would say over the last 1.5 years, that's exactly what we've done through both this vision of a platform where all these applications we saw in the ecosystem that we're leveraging analytics to drive outcomes for their clients, whether you think of that in terms of experimentation or -- which is like AB testing or session replays or guidance and surveys. These are all applications and businesses that are $100 million plus that are out in the ecosystem that are surrounding analytics. And customers are having lots of pain points about how they actually administer workflows by ETLing data from one place to another, dealing with UI issues, dealing with data taxonomy, and it wasn't very efficient.

And so the vision was to bring all those applications together in one environment, such that you can easily draw those workflows between those applications. And it wasn't easy. I think that if you would have asked me when I first joined, you would have seen us with a bunch of applications that have different software development kits that looked like they were separate. But the engineer team did a great job of bringing that together.

And for those people who are not very technological savvy, I tell them, think about a time when we had browsers. And if you click in a browser link and then another window would come up. And then you click on another link and another window would come out. And suddenly, you had a sprawl all over your desktop all these windows, right? And then FireFox came out with this great idea. Let's take and make those tabs into the one environment you're seeing and you get greater organization. Well, that was a great way to think about how you better manage your interaction in those environments.

And I think about, foundationally, that's what Amplitude did. They brought all these disparate applications into one environment such that you're not ETLing data back and forth. If you get an insight and you want to run an experiment on it, it's that easy to go just run an experiment. When that experiment comes back, if you want to drive a workflow into targeting a cohort of customers based upon that experiment, you can go enact a guide or a survey on it real quickly.

Now Spenser Skates, our CEO, has had this vision that product analytics is going to evolve to the point where you have self-improving products. This notion that products are evolving based upon user interactions real time and that, that becomes increasingly more personalized. Well, the advent of agents into our platform is enabling that to occur because now agents are the ones who are picking up on insights and making recommendations, running experiments and if you allow them, to autonomously start making changes to the environment, whether that's a web application, a mobile application or even a business process. So this notion of self-improving products is not that far away, given some of the changes in the architecture that we drove and the agentic capabilities.

And so when you ask what I'm most excited about, it's this notion that we can go drive consolidation into a very fragmented market associated with product analytics, marketing analytics and data workflows is becoming more real. And because we did both the technology and executed well and on the sales side, we did well to modernize our go-to-market and create a more enterprise motion than we had before, those things are helping to reinforce themselves.

And so even if you told me the agentic capabilities and all the things that we've introduced into our platform, we're not happening. I'd still tell you, I was excited about the two areas that are driving our growth around platform consolidation and the sales go-to-market team moving more and more towards enterprise.

Got it. Yes, it's really exciting. And the AI agents are a new functionality that you've developed. Could you talk about the early customer feedback and the behavior that you're seeing early on with how they're interacting with analytics and guides and surveys and all the other offerings that you talked about?

So the things that we rolled out a few weeks ago was -- you got to think of it as agentic capabilities that drive optimization and broader use within the platform itself. So our global agent, for instance, if you look at the Amplitude UI now, you see a prompt, not a set of tools and widgets and things you have to go to navigate. It's a natural language prompt and you may say, "Show me the conversion rates in the funneling campaign that I had running last night." You'll get out the information, maybe get a few charts. So it enables the barrier to adoption to be lower and starts enabling the business analysts as opposed to the technology analyst to use Amplitude.

Plus you have basic agents around automated insights. And I can turn to an agent to say, "Hey, these are the things I want you to focus on and the data sets that are coming back from the mobile application." And so the agent can run continuously about how those interactions are happening and give you insights on to what the user behavior is.

They can also start running experiments on that data continuously based on the insights of the agent. So you have an agent to agent pass and the experiments can start running. And we have an agent that does summarization of session replay information. It's one thing agents do really well, is take large data sets and they summarize them for you. And think about the engineers that had to go watch lots of different session replays to get qualitative feedback on the interaction of the product. Now the agent can summarize it and say, "These are the ones -- don't go look at 1,000 of them. Here are the three that exhibit this behavior that you should look at." And potentially, you might want to run an experiment to understand if you change that environment, what would be that outcome.

So those things working together really will exemplify the power of the platform. What our customers have been exposed in the beta have learned is that it's much easier to drive those workflows and much faster to run some of the campaign analysis or promotion analysis or product feedback than they ever could before. So when we started talking about the agents and the optimization, that was the first thing that customers are like, "I want to learn how to do that." And I'll talk about pricing and packaging in a minute, but the first thing they say after that was like, "Wait a minute, this is going to cost me a lot, too. So help me out with how we adopting that is how much it's going to cost me." And we said, "Well, we have a response -- we have a solution for you."

The other thing that was interesting, we started talking about MCP server and being able to connect to more data in a more seamless way. And one of the biggest impediments to customer adoption of Amplitude has always been the underlying data taxonomy. Most enterprise customers you go to them and you say, "Show me what your data architectures are and how you want to use that within Amplitude." they'll show you a mess. And the reality is MCP starts to help obviate some of that mess because now you're not having to do hard connections and integrations, you're actually able to use API calls to get through to relevant information. It doesn't help necessarily quality, but the quality can be addressed through an agent actually working on the relevant data.

I'll give you a use case where a customer started to bring in their CRM data into Amplitude. And they had 5 different definitions of ARR, which one is right? Well, the agent dissevered that the one that's used in most other places in adjunct applications was top line, and that became the basis of the integration.

So there's ways in which we can use the technology to really add more and more value that customers can implement and more and more data that customers can implement in the platform and a lot of them are very interested in that. Now we knew that as we were going through this platform journey, and that -- and this move into enterprise selling, that our pricing and packaging was not right. It was absolutely inhibiting adoption. It was a very complex, it created a friction in the selling process. Customers didn't have transparency on what their usage was going to be. So they are very worried about the cost associated with adopting more of Amplitude.

And we had really poor selling practices, too, that had penalties associated with customers going above and beyond their entitlement. So you think about a customer who has a million events as part of their contract. And maybe we're charging them $20 per million of events ingested. We'd have a -- if you look at one of our order forms, it used to have a clause that basically said, "I'll charge you 2x or 3x if you go over that." So talk about dissenting adoption. You're putting the fear that the customer is going to get a bill at the moment that they're not managing their data ingesting correctly.

So we flipped that. And we said, "Look, if we're going to be a platform, you've got to -- the first principle is you got an incent adoption" and you've got to lean into the customer value proposition, help them understand how they're going to adopt, be more transparent on what those costs are going to be as they adopt more, give them an incentive to displace other applications that they may have around this ecosystem. So we had to have an approach of the philosophy and our pricing strategy was not a premium pricing. It was one that was value for the money. But if I go into any customer and said, "I'm going to replace three applications", but they pay us is less than what they're paying those other three.

So those things had to be abolished. And actually, when I joined Amplitude, I gave the Board a report out after 30 days. And this is one of the things I focused on. I said our pricing strategy is broken. It's actually causing churn and it's not allowing us to go drive our platform strategy nor our enterprise strategy. So we've been working on it for a while to simplify the architecture. And we tested a lot the primary billing meter we have. Unlike other companies, Amplitude was already using a volumetric billing meter with about -- based on the number of events that were ingested into the platform. So it's not seat-based, it's already on a volumetric. We didn't know if that was going to be -- should continue it. But customers came back and told us, no, actually, that data ingestion is a great proxy and value that we're getting from Amplitude.

So we kept it. And it's good because it was 86% of our installed base was on event-based volumes. What a lot of customers didn't like was that when they adopted more of the platform, they got additional billing meters. So they adopted experimentation. They had a meter on experiments. They adopted session replay, they had sessions. They adopted web analytics, they had web sessions. And so they said, "Look, that's too complex, and that's pushing more license administration to our clients." So what we need to do is simplify it as much as possible.

So now rather than having specific meters, we have a value-based uplift based on the data ingestion. So if they're paying is $20, like the example I gave you before, they may pay us $26 by adopting experimentation or they may pay -- if they go another one, they may pay us $32 per. So it's an uplift based on the data they're using and how they're using across the different modules themselves. And why that's important is because it gives implicity and cost predictability, transparency to our clients, especially if we're going to use this over a number of years and allows us to go right back in and show how customers have that cost predictability when we quote against the installed base competitors.

Got it. And could you talk about the breakout between those within your existing customer base, who've adopted this pricing versus those who haven't? And what do you expect that to grow this year? And then maybe, over the next two?

Yes. I'd tell you that we were testing it through Q4. There were -- one of the things you do when you introduce a new pricing package, you don't try to automate it right away. You try to make sure that it's going to be the right one, you test it.

So we did that with a number of clients. And what I can tell you as an example, and I expect more of this will be the case that the customer is talking to you first had aspirations to replace incumbent with a session replay and their analytics. That was -- it was a 2-product solution they we're looking at. We started talking about pricing and packaging, what we are -- and they were expressing interest in experimentation and guide and surveys. We showed them what the pricing and packaging could look like. And they said, "Well, I'd like to do that now." So what happened was we increased the size of that deal. We also increased the contract duration of that deal. And I think they're going to be a great customer for us longer term. So a lot of times, when you simplify things and give cost predictability, they look at it very favorably.

Got it. And then kind of combining both the new pricing and packaging with AI agents, one of the things that we're really excited about is seeing your core customer base expand into more nontechnical users. Have you seen that more recently with new updates? Just talk about...

Yes. I mean one of the -- another one of the customers I was talking to was a pure marketing case, and they were having a lot of pain between what you consider IT or data, the data environment versus what they wanted to run on their campaigns on a regular basis. And they looked at Amplitude is breaking down that barrier. And then quickly, they could manage their campaigns on a regular basis and not have to worry about the data transfer from the IT department. So there's definitely that aspect happening as we move into more and more use cases and the frameworks around agents just enable that business process to work more smoothly.

And the thing we learned, too, is I'll go back to pricing packaging, the other thing we learned is we added agentic capabilities. The customers are tending to adjust more data into the platform because they're seeing more use cases. And they have a high propensity to use a broader set of suite of products. So both the cross-sell and the upsell mechanisms are -- and the pricing framework we're applying to it allow us to monetize as people use more of the agentic capabilities.

Got it. So talking about net retention here. You're currently around 105% with long-term goal of 115%. Can you help walk us through how we get to that 115% goal and maybe the time line that you forecast?

So I think you're going to see us continue to be innovative on adding more and more modules. So that cross-sell. The cross-sell is really what's driven us from that 96% to 105% improvement because throughout the last year, 1.5 years, we've had to overcome some overcapacity sales. So even if data ingestion rates were increasing in our core customer base, we had to overcome these contractions and churn that have been caused by that poor selling.

I'll give you an example. I had -- one of our largest contracts was renewing in early 2025, and data volume hadn't changed, they were using. But contractually, they had the ability to reduce their rate by 33% upon renewal. So why would we have ever taken the full value of the ARR at the time? It's not something I would have done, but because there's a high propensity for them to renew if you're giving them a 33% reduction and they're using. But that's some of the bad constructs that were out there that we had to overcome.

We're past most of those. And so what's good is data rates continuing to increase in the platform, and we don't have those contractions, suddenly, upsell has become a meaningful aspect of net dollar retention, okay? So let's call that 5 points, okay?

Now cross-sell continues to get more and more robust. And we still have -- in our investor pack, we showed how -- the percentage of our ARR that's resulting from customers who have 2, 3, 4 and 5 products. We have 74% of our customers are actually on 2 products now. But if you go down to 5 products, there's only 20% of our customers. So there's still a great opportunity just within that product set for us to expand. And I would tell you, we're going to be introducing some new products in the very near future in the marketing analytics space, in the agentic search space that we'll charge for.

Right. And there's a ton of new app development, obviously, with AI facilitating that process. Could you maybe talk about the breakout between -- and you kind of just touched on it a little bit, but the opportunity within your existing installed base versus acquiring net new logos on a go-forward basis?

So I'm one who preaches, especially at this level, we're -- call it, $350 million to $400 million in size. We still should be generating a good proportion of our ARR that's coming from the new logos. And I was worried when I first joined, we were over 30% new logos and 70% expansion. And so we started re-architecting territories, driving prospecting aspects more, our marketing demand gen was targeted more at new enterprises. But enterprise new logos take longer. You got to set up a master services agreement. You got to do a data protection agreement, you got to do security reviews. It just takes longer, especially in enterprise. So that was a little worry too coming in, like how do we start shifting this percentage a little bit more?

Well, in Q4, and we saw it in the beginning of last year, too, there was better balance. You saw more new logo ARR and I would love to get it more like 50-50 or 40-60 as opposed to 30-70 in both first quarters, you saw more new logos and that's last Q4. It was a really great balance. We had a record number of new logos in the greater 100,000 new lands, and we had 18 new G2Ks. So I think you still have -- we're still in a place where we have plenty of room to go after new logos. We only have 160 G2Ks in our contract. So there's -- we're way underpenetrated in the Global 2000.

Great. And then as you continue to make more investments going further upmarket and adjusting to this new pricing model, could you talk about long-term gross margin profile and more of the unit economics as you make these transitions?

Yes. So gross margin, we're actually a pretty simple business from a cost to serve perspective. The biggest cost we have is our hosting costs right now. And we're primarily in AWS over time, we'll probably get into multi-cloud. But with multi-cloud, we'll open up more retail, as you can imagine. The other costs are associated with software applications used to help run the environment. Services costs where we're actually doing professional services and that was a straight cost for a long time, we weren't even charging for services. And then you have cost associated inference, which is relatively new, still relatively small. But all those costs are built into our cost to serve.

Now one of the first things we did when I joined was we put a plan together, and that plan said that engineering had to take some effort and focus on driving the marginal improvement cost of data into the platform down so that we're constantly re-architecting and reengineering how our application runs in hosted environments allow us to -- as we add more customers, add more data that it's not linear increases for us, right? So we're driving that down.

I talked about the services business. We're never going to have a huge services business. It's 1% today, but it could get to 3% or 4% over time as we get more and more complex. And frankly, more and more customers are asking us to help them with their genetic implementations of workflows. But we're also going to build a partner ecosystem to do a lot of that for us. And then we got to make sure that as we're increasingly seeing larger and larger customers move into leveraging our agent capabilities and we're managing those expenses and monetizing appropriately.

There was a quarter, I think it was Q3, where we saw very much an increase in the data ingestion in the platform. Well, we've already started making changes in our go-to-market. We weren't going to drop an invoice on our client, we're going to have a conversation with them first. And a lot of times, those conversations came back. Yes, but maybe I need to align it to my budget a little bit better. So the monetization took a little longer but it also is more durable.

So some of those things, I think is a push and pull on the gross margins. We made improvement. We improved gross margins year-over-year. And our long-term aspiration certainly being in the low 80s. But there are going to be points in time where we may see spikes in demand in one area. And I would encourage everybody to think about when that happens, we've instrumented it such that we're going to monetize it and keep driving improvements.

The same goes for sales and marketing. Sales and marketing is too high as a percentage of revenue. It needs to be in the low 30s. We're in the low 40s. G&A when I joined was 17%. Now it's at 12%. So those two areas are continually going to drive scale. And the R&D is the source of our future innovation. So I want to try and keep that right around 18% to 20% regularly. And I'll come up when we do small acquisitions and go down when we see greater scale.

Got it. And one thing you mentioned in that intro was partnerships. Could you talk about how those partnerships, particularly with the data platforms change the growth algorithm going forward?

Yes. It's funny. It's another -- I told the Board in that first 30 days, like here is an untapped area for us as far as future demand goes. We were only getting 2% of our pipeline from partners. We didn't have a great relationship with AWS, so we had no go-to-market relationship. So we renegotiated that whole contract now we do, and it's much, much more positive. You see us to have stronger relationships with technology partners. They're in different parts of the, I'll call it, path to which you're delivering value to clients, like Klaviyo and Braze would use an analytics layer to go drive what they're doing in their last mile marketing. There were good progress with HubSpot. And there's a number of areas that I think we're increasing to see with like [indiscernible] and Merkle and others in the value-added reseller space.

Ultimately, if we create a framework under which they can create value on top of Amplitude, maybe they develop their own agents for logistics or for retail or for health care, and they start monetizing that. That provides a basis under which increasingly more and more global systems integrators would want to do something similar. And so over time, I think that channel has to be something that is north of 30% of our overall pipeline. And as we get more and more embedded with enterprises, that's the way they've historically consumed.

Right. You've talked about in the past winning simple and winning the enterprise. So when I think about app development going forward, there's a lot of new entrants into the space that may not be your key focus area right now as you move upmarket. How do you balance the investments needed to maintain that demand profile from the lower end, but also move up market?

It's a push and pull, you have all the time -- and just yesterday, our executive team was talking exactly that with some programs we're rolling out to some of our VC friends for their portfolio companies who are start-ups. And we're not going to make a lot of money on that group, but we'll make good money once they grow and start developing their own application and value sets. I mean at one point in time, there was a number of our customers who were very small, and they graduated into paid plans. And we see that continually.

Got it. So let's shift gears to competition and positioning. You've highlighted an accuracy advantage in complex agenetic queries versus the warehouse native platforms. Can you speak to the durability of that differentiation? And where do you see that going in the future?

I think that's one of the areas where you just have to continually innovate and provide the context back on the use cases that customers are valuing. If you're going to pin down our engineers, they'd tell you there was a lot of work to give those queries context and understanding how to get to the right level of a query. Which queries the query that follows on when you get the answer, and how best to instrument that.

It is, in many respects, the underpinnings of what we call our behavioral graph which is we spent years and years developing. I'm not suggesting that, that gap doesn't close over a period of time or that it doesn't change in its dynamism and complexity, it will. But we'll continue to recognize that that's the case, too and innovate around that.

Got it. And as product analytics and marketing analytics converge, how does your simplified pricing model and AI native architecture meaningfully improve win rates versus the legacy vendors...

One, the pricing and packaging does a great job of that because we can go right into any customer pretty much and say, "Look, if you've got 3 or 5 applications that are competing, we can show you a value for the money where you'll pay less in licensing" by standard on Amplitude. And we can -- we've also done, as I mentioned earlier, a great job of showing how you can optimize the workflows. They used to be very cumbersome across those application environments. And I would tell you from a product and marketing analytics space, every business is trying to figure out how they better digitally engage with their clients, whether it's a service or a product, what have you. They are trying to figure out how they do that in a more efficient and more meaningful way and get better outcomes. And that means that more of the classic marketing cases are moving towards consider treating their campaign as a product or cheating -- creating their promotion as a product. And that lends itself to the architecture that we've built.

Right. Let's shift gears to financials. RPO grew 35% year-over-year and looking at contract duration that's nearing 22 months now. How is the structure of these deals changed for the past year? And how do you expect them to continue to evolve?

Yes. There was -- there used to be a philosophy in our sales team that customers didn't want more than 12-month contracts. And I said that's [indiscernible] especially in the enterprise. Enterprise is -- there's always this push and pull of like, "Do you want to pay for something you're not using so you get software, cloudware or whatever you want to call it?" versus I want cost predictability for the term of the agreement and so that I'm not going to be surprised on cost as I start to adopt.

And in the SMB mid-market space, they're very concerned about their cash flow and getting unit economics. And that group, I would agree for the most part, they'll want to have 12-month contracts and they want to make sure that they're getting the right technology and instrumentation, especially if they're trying to improve product market fit in their own product.

But the enterprises think differently. Enterprises oftentimes either have a homegrown application or they have a competing application and they're planning for when they're going to transition over, they want cost predictability. And so I've increasingly taught our sales team how to have the right deal construct conversations to meet the customer needs.

A lot of times, they used to talk about our technology. What I want them to talk about is what the customer wants to achieve. So the customer tells me, "Look, I really want to move over from the [indiscernible] vendor to you, but I have a problem and that contract goes for the next 5 months." Well, "Let's do a construct that says, you're not going to sign up for just -- in 5 months because that's not realistic. You're going to have a transition where you need to use us as a staging period. You need to test this out." Let's set up a construct that's -- let's call it 3 years. And I won't charge you as much for the first 5 months, and we'll not charge you more in years 2 and 3, right? So that's a simplified construct that we've taught our reps, how to do. And suddenly, they -- wait a week, I just got them into a 3-year contract. Yes, you did because you listen to what was most important to them.

A simple one, I always tell our reps is if a customer tells you, you have a budget problem, you know what the first question you should be asking? Are you talking in terms of expense or cash? Because how you respond to that will determine how you should construct. So if they told me I have a budget problem, an expense problem. My budget period doesn't start until January 1. Well giving them better payment terms doesn't help with that problem, right? So what you really want to do is figure out how we can get aligned to their budget as much as possible.

That makes sense. Let's talk about IT budgets. Where does the budget for Amplitude typically originate today, and could you speak to the defensibility of that during an environment where IT budgets may get compressed?

So a lot of times, our initial core use cases are product analytics. So you got the CTO, the product manager, the product developer, those are the classic percentage we're going after. I'd say increasingly, we find ourselves with more broadened marketing analytics, you're going to CMO, a business analyst or you're going into potentially even the CIO environment. And when we dealt with the economist, they were moving their business from pure print to subscription, and they're using Amplitude to understand how users are interacting with content and that influenced what content they created, meaning they want to see when customers are browsing versus actually engaging. They use that information then to go build out what their advertising campaign was and they also use those interactions to understand what customers would value in the subscription packaging.

So that transition was one that enabled them to move into the digital world, but then broaden out the discussions on workflows and how Amplitude was being used. And the people that we were selling to were the CIO, the CTO and the Chief Marketing Officer, they were together in that discussion because it was so essential to their transformation.

So it kind of depends on what the use case is and who you're talking to. But our view over time is there aren't these silos of operations and data that they're increasingly converging. And so it kind of depends on the business and how they're organized and who's going to be the one who's trying to solve the core problem for the business.

Got it. Let's shift back to product for a second. So if you think about the future investments in the business and the current segments, versus investing in new segments, how do you split forward investment between your existing products and the demand you're seeing there versus new areas you want to explore?

Yes. It's a hard one because -- I mean, that's the level of maturity we've had to adopt even in our planning process. I can tell you, we go set a plan, and that plan has a limited amount of funding for R&D. And then they work with sales and our customers to try to prioritize as best they can. But invariably, you're going to draw a line and say this is where we're going with this. This is as far as we're going to go.

And there are some things that you would say are enhancements and improvements in existing applications. Like when we did session replay, we had, had a specific gap versus the incumbents out there around mobile. Mobile, we had to have especially with mobile. So that was a high priority. Well, there was others that on the marketing analytics side around orchestration and merchandising and e-commerce that we had to deprioritize. That didn't mean they fall off completely and you're just having to work through the priority list first.

And the thing that we often do is we have reviews, we question that. Our customer feedback, we're seeing demand in areas that are more important than others. And that's the job of management to give back the right choices, right? And I'm not saying it's easy. It's not. I shared today a couple of different times that I'm increasingly believing there's a great opportunity for Amplitude in this whole shift from search engine optimization to agentic optimization. We rolled out a product called AI visibility. We had the most views of Amplitude ever. And for us, it was a way to demonstrate how you can quickly create an application, but if you don't give a context and you don't give it value that customers are actually seeking that -- it doesn't really matter. And there's a lot of companies that got very mad at Spencer essentially by saying, "Oh, look, we created this really quickly. But if you don't have -- you don't have user behavior context, you're really not providing the right level of accuracy back."

Now the debate we have internally was, was there really a profit opportunity for us? Who's going to pay for this? If everything is through the search is through the LLM is free? I think, well, I think there's going to be a market in which it's created, how Andrew? Advertising. Well, guess what happened? OpenAI opened up an advertising business. And now every brand is asking us, how can you help me service my brand versus my competitors through agentic search. That's an opportunity for us. And guess what? We're working on a product that does that.

So all these things, I think that -- the one thing I would say, maybe for me close on this is tougher companies have evolved over the last 30 years, every time there's a new technology that makes developers more productive. They've sought out new ways to drive software development and value to their clients that automate and improve workflows. We're going to continue to do that, too. In fact, we think that in order to survive in the agentic world, you have to adopt fast, you have to move quickly. You have to deliver products at breakneck pace that delivers value to our clients because the world is increasingly becoming more and more personalized through agentic optimization. If you can't move quickly and drive value, and those are the companies that are going to find themselves disintermediated.

Really, really interesting. And that's all the time we have. Andrew, thank you so much for the time. Appreciate it.

Yes. Thank you.

Thank you, everybody.

Good afternoon, everyone, and welcome to Amplitude's Fourth Quarter and Full Year 2025 Earnings Call. I'm John Streppa, Head of Investor Relations. And joining me today are Spenser Skates, CEO and Co-Founder of Amplitude; and Andrew Casey, Chief Financial Officer. During today's call, management will make forward-looking statements, including statements regarding our financial outlook for the first quarter and full year 2026, the expected performance of our products, our expected quarterly and long-term growth, investments and our overall future prospects. These forward-looking statements are based on current information, assumptions and expectations and are subject to risks and uncertainties, some of which are beyond our control that could cause actual results to differ materially from those described in these statements. Further information on the risks that could cause actual results to differ is included in our filings with the Securities and Exchange Commission.

You are cautioned not to place undue reliance on these forward-looking statements, and we assume no obligation to update these statements after today's call, except as required by law. Certain financial measures used on today's call are expressed on a non-GAAP basis. We use these non-GAAP financial measures internally to facilitate analysis of our financial and business trends and for internal planning and forecasting purposes. These non-GAAP financial measures have limitations and should not be used in isolation from or as a substitute for financial information prepared in accordance with GAAP. Additional information regarding these non-GAAP financial measures and a reconciliation between these GAAP and non-GAAP financial measures are included in our earnings press release and the supplemental financial information, which can be found on our Investor Relations website at investors.amplitude.com.

With that, I'll hand the call over to Spenser.

Good afternoon, everyone, and welcome to Amplitude's Fourth Quarter and Full Year 2025 Earnings Call. Today, I'm going to cover 3 things: First, our strong Q4 results and progress in the enterprise. Second, how AI is driving demand for analytics and our strategy to deliver. Third, a look at our new AI agents in action and a spotlight on customer stories. Q4 represents one of these strongest quarters in Amplitude history. Our fourth quarter revenue was $91.4 million, up 17% year-over-year and exceeding the high end of our revenue guidance. Our annual recurring revenue was $366 million, up 17% year-over-year and up $18 million from last quarter.

This was our highest net new ARR quarter since 2021. Non-GAAP operating income was $4.2 million or 4.6% of revenue. Customers with more than $100,000 in ARR grew to $698, an increase of 18% year-over-year. Over 25 AI companies are now included in that $100,000 cohort as well. This quarter was marked by balanced execution. No single deal was over $1 million, yet we had our highest ever number of multiproduct and $100,000 ARR lands. I want to talk more about AI and our strategy. Over the past year, AI coding assistance from Anthropic, OpenAI, Cursor and others have compressed development cycles dramatically.

The velocity at which companies are shipping new products has accelerated. When software is this easy to build, it creates a gap between how fast teams can ship features and how fast they can learn if they are working. This shifts the pressure to the right side of the product development loop that you see here, the use and learn side. Understanding how users behave, what works and what doesn't and what actions to take next becomes the bottleneck. The constraint is no longer knowing how to build, it is knowing what to build instead.

This is the hardest problem in software today. I say that because builders and their AI assistants need a system of context that combines multiple data streams. They need structured behavioral data. They need the correct retention and funnel logic, and they need the right analytical tools exposed in a way that enables AI to reason effectively. The AI then needs to be able to iterate with that system, test hypotheses, refine queries, identify root causes and recommend actions accurately and repeatedly. This is not something that can be bidecoded over a weekend or replicated accurately with an LLM on a data warehouse. However, it is exactly what Amplitude is purpose-built to do. We have worked with thousands of companies over the past 13 years and amass the world's largest database of user behavior. Our AI can explore patterns, explain changes and guide teams on what to do next more accurately and reliably than any other system. Over the past 6 months, our Agentic analytics platform has reached a 76% success rate on complex production-grade quarries, that is 7x better than a straight text to equal approach. With the new agents we launched yesterday, teams can now move from insight to action in minutes, not weeks using analytics, cohorts, experiments and messaging in 1 continuous agenetic workflow. Through our MCP integrations with Anthropic, Figma, OpenAI, GitHub, [indiscernible] and Slack, we are bringing behavioral intelligence to teams where they already work. Understanding user behavior now becomes as simple as asking a question in a chat window. This puts Amplitude in a unique position. The Frontier labs are pushing the boundaries of AI models and they recognize the complexity of analytics experimentation and behavioral understanding so they turn to Amplitude. As I mentioned earlier, more than 25 of the leading AI native companies, including some of the names you see here, our customers with over $100,000 in ARR with Amplitude.

In addition, one of the world's largest frontier AI labs is a 7-figure customer as well. They came to us to replace a manual system built from fragmented internal tools and raw warehouse data. Using Amplitude enterprise analytics and session replay they can now understand activation, engagement, retention and monetization end to end. With Amplitude MCP, they can offer those insights directly within the AI environments, their teams already use dramatically improving the ability for them to automate development. And it's not just AI companies, companies of all sizes need a system that gives them trusted data, insights and action to successfully deploy AI in the real world. So they turn to Amplitude as well. This momentum, combined to one of our strongest quarters across gross bookings and new ARR alongside meaningful improvement in churn.

Our go-to-market motion has matured. There is a tighter focus on various use cases in the enterprise and on expanding with multiproduct deployments. We continue to consolidate the fragmented market. Platform win rates are increasing against point solutions and our newer products are gaining traction. Guides and surveys launched less than a year ago, is our fastest-growing product to date. We are also seeing a large increase in AI native usage as agents connect directly to amplitude. Over the past few months, the total number of quarries triggered by AI agents has increased dramatically.

In October last year, there were almost none, and today, it is 25%. Agents also drove the vast majority of overall incremental quarry growth. This tells us that customers are trusting agents with analytics work. It also indicates that our platform offers the accuracy and the context needed in production environments. Taken together, this creates a powerful tailwind for Amplitude as we continue building a durable, scalable company that can unlock the next frontier and software. Over the years, we have intentionally expanded beyond core product analytics and into adjacent workflows. We have continued that work and acquired Infinigro, an AI-native marketing analytics start-ups that can expend behavior and revenue impact. [indiscernible] brings strong AI native engineering talent to amplitude. This strengthens our platform as a system of context and expands our ability to bring acquisition, activation and retention into oned continuous feedback loop.

Yesterday, we launched our global AI agents, specialized agents and MCP. This represents the start of a fundamental shift in how teams work with their analytics data. Historically, analytics has required humans to do most of the heavy lifting, writing quarries, building dashboards, monitoring changes, interpreting results and then figuring out what to do next. That process does not scale in the world where teams are shipping faster and faster. AI agents change that model.

Instead of asking questions one at a time, teams can now delegate work to agents that continuously analyze behavior, surface insights and guide action. Our agents understand events, funnels, cohorts, experiments, session replay and outcomes because they operate inside a context system specifically designed for them. Agents make life easier by doing the work that slows teams down today. That is very, very different from bolt-on AI tools from SaaS companies that sit outside the data and try to infer meaning after the fact.

The best way to see this and understand this is to look at it in action. I want to show you a quick teaser video, and then I'm going to show you a demo of what we've released. Let's go ahead and roll the video.

[Presentation]

It's a great question. All product builders should ask themselves and what will yield. I want to now walk you through what we've launched in AI analytics yesterday. I'm really excited about the future, and I want to show you global agent. Global agent radically changes how our customers interact with their data. starting your day with a dashboard is dead. Take a look at this interface, no dashboard, no grass, no charts, just a chatbox and a few simple prompts if customers need help getting started. I can talk to global agents like I talked to a colleague. I'm going to go ahead and ask it how's our loyalty program doing?

In seconds, it comes back with a summary. Notice I did use any target about event totals or taxonomy, just a regular question. It's calling out some pretty concerning numbers. Only 5% of users who view our welcome page actually go on to join the loyalty brand. That is low, so I'm going to click in and investigate more. The global agent has followed me to a deep dive on this chart. I can keep investigating with another simple question. break this down by traffic source. Here's the breakdown. Facebook and Instagram are driving loyalty sign-ups at 5.6% and 5.2%, while Google and direct traffic like [indiscernible].

The global agent summarizes it perfectly. Social media converts 10% to 15% better. Since social media outperforms Google, I might shift ad spend, but looking overall, all the rates are low. So before reallocating budget, I'm going to go deeper. Is this a channel problem or an audience problem? Let me ask, do new users convert differently than existing customers? Without AI, this kind of analysis takes a lot of time segmenting users comparing funnels, pulling in sites together, the pool agent does it in seconds. And here it is 14% conversion for [indiscernible] repeat purchases, 5.4% overall. That's 2.6x higher. That answers my question. It's an audience issue, not a channel issue. I should reallocate my budget towards repeat purchases. Again, simple language, fasteners answers, deep learning that anyone can use. Analytics is the perfect use case for agents. So I want to show you specialized agents. Our specialized agents work continuously specific jobs that would usually take dozens, if not hundreds of hours, monitoring dashboards, analyzing session replays, processing feedback, running conversion experiments. Legwork now done automatically.

We're going to be eating our own dog food on this one. I already have a session replay a set up to monitor our own session replay tool and I have it set in addition to seminal when it has a strong finding. This specialized agent has been watching hundreds of replays and sent me some summarized findings. Users with multiple saved filters type search terms, but cannot find filters without scrolling through the full list. Power users cannot preview filter criteria before applying forcing trial and error selection.

These are all things we should improve. We could have had someone watch all those replays. We could have talked to customers from hours an end or we could have let these continue to be issues. Instead, I get these findings serve to me on a daily basis with a full report and a detailed breakdown with key findings, suggestions on what to explore next and even a highlighted set of replays of these issues. Okay. We're going to save the best for last. Finally, I want to show you what I'm most excited about, which is Amplitude MCP. We're releasing a fast library of expert level workflows that customers can trigger an AI clients like Claude with a simple command. I'm going to go ahead and use amplitude and Claude by typing use create dashboard and create a dashboard that tracks our growth conversion performance, had [indiscernible] head ender and it goes to work.

Instead of me manually creating 15 charts, running the segmentations myself, and piecing together an explanation in a dock, this scale handles it in 1 click. With MSC apps, cloud is opening and building amplitude charts right inside itself. It's done it. So I've now gone to the link it gave me in a perfectly built dashboard with top-level metro inversion funnels and segment downs. Amazing. Moving on to customers. We had a great quarter for new and expansion deals with enterprise companies, including one of the largest music streaming apps, the Cheesecake Factory Asana, PGA of America Cost [indiscernible], Stewart Tell Guaranty Company, Crunch Fitness, WOP, Once Upon Publishing and NTT DOCOMO.

I'm going to highlight 3 examples that demonstrate the power of the platform in different ways. Japanese telecom NTT DOCOMO is using amplitude across more than 1,000 active users to drive efficiencies at scale. As an early design partner for our AI agents, their data platform team uses agents to streamline analysis across existing dashboards. In 1 project, an agent reduced campaign analysis time by over 90%. Our AI-powered session replay summaries automatically localized into Japanese help UX teams identify issues faster and improve the digital journey for millions of customers. We are now working closely with NTT DOCOMO to shape our agents road map with feedback on collaboration features and AI-powered insights. Siemens, the $70 billion global technology leader partnered with Amplitude over 3 years ago to power analytics across its website presence and broader digital ecosystem.

By consolidating onto our AI analytics platform from a series of point solutions, Siemens gain a unified real-time view of user behavior. Recently, the team organizing their annual user conference use Amplitude to identify their overreliance on direct e-mail and organic channels. They experimented by reallocating spend into targeted web promos plus played and organic social. This delivered a 90% year-over-year increase in web traffic and a projected 50% increase in registrations in attendance to their conference. Lastly, we landed one of the largest music streaming apps in the world. We are working with the teams that lead checkout optimization, upgrades, churn prevention and recovery as they seek to understand the revenue drivers for hundreds of millions of monthly active users. They will use Amplitude analytics combined with session replay to get a holistic view on these monetization drivers. These stories all point to a common theme from AI start-ups to global enterprises, customers are betting on Amplitude as the AI analytics platform that will help them thrive in this new era.

Before I hand it over to Andrew, I want to be clear on how AI is shaping our opportunity. There is a common misconception in public markets that AI makes analytics either irrelevant or easy to replicate. The exact opposite is true. AI has made software easier to create, but creation is no longer the moat. The real advantage is how quickly a team can learn, iterate, improve and automate. [indiscernible] analytics is the key. It unlocks the bottleneck on the right side of the product development loop and enables teams to learn as fast as they ship. AI is a structural tailwind for Amplitude. It is why I believe the opportunity ahead is massive and why I'm excited about what's to come. Now over to Andrew to walk you through the financials.

Thank you, Spencer, and good afternoon, everyone. 2025 was a year of innovation, execution, and we delivered a solid base for our future long-term growth strategy. When we met at our Investor Day last March, we laid out a deliberate road map to capture the enterprise and accelerate multiproduct adoption, while leading the industry in innovation. Today's results demonstrate that we haven't just met those goals. We've established a new baseline for durable growth. The enterprise is now our core growth engine. ARR from our enterprise customer cohort is up 20% year-over-year, with higher retention and expansion rates than the rest of our business. This was not by accident or luck, our [indiscernible] platform has been designed to be enterprise-grade with trust and safety of our customers at the center. Our go-to-market team has worked for the past 3 years to orient our go-to-market motion to focus on the enterprise, increasing customer value through selling our platform and engaging in longer-term contracts. 2025 was the coalescence of this work to focus on our customers' value and creating durable base for future growth.

We sustained growth of current RPO greater than 20% throughout the year. And in Q4, total RPO grew 35% year-over-year. Our average contract duration is now above 22 months. In addition to our success in the enterprise, we have also formulated our product and our go-to-market team to embrace our AI platform strategy. By combining niche point product solutions surrounding analytics into our comprehensive platform, we are able to deliver greater value than stitching together point solutions.

We also believe that having a platform is essential to the hearing capabilities of AI to reduce friction in our customers' workflows. In 2025, we did a great job expanding our multiproduct attach rate for our customers. 74% of our ARR is from customers with more than 1 product, up 15 percentage points from last year. We still have a great opportunity to expand our multiproduct customers as well.

Only 51% of our ARR comes from [indiscernible] with greater than 3 products. Looking at a full platform deployment of 5-plus products that's sending just 20%, doubling year-over-year. We have a [indiscernible] opportunity to expand with our customer base. We believe our market opportunity expands dramatically with the inclusion of our new AI products that promise to expand adoption and use cases. The progress in selling our platform is best exemplified through improvement of our retention and expansion motion with dollar-based net retention now above 105% after exiting 2024 at 100%. However, our work is not done. At the beginning of this year, we introduced a new pricing and packaging strategy to our sellers. Let's start with what's not changing. We are not changing our core billing metric of events. We believe this is a great representation of the value our customers receive from our platform, and it is also an appropriate monetization strategy as we center AI engagement on our platform.

What has changed is we are centralizing the monetization of our other products, such as experimentation, such as replay, kind of surveys to be a percentage uplift on the core platform charge, which has been spaced. This reduces the friction of adoption of those products by making it easier to understand for our customers and reduces the need to estimate how many sessions or experiments they want to run in the near term. Long term, this will also encourage greater consumption in our platform as comes no longer fear over using certain parts of the contract were underutilizing others.

It's a radical simplification of our pricing that acknowledges our customers' needs for greater cost transparency and certainty and our cost is that the volume of data ingested into our platform experience. It also supports our focus of integrating AI into all of our product offerings and expanding customer usage, which can be a tailwind longer term on easier lands and faster platform expansions. In summary, as a transition to an AI analytics company, we have created a more durable base of our business focused on the enterprise. We've driven expansion of our platform through innovation, and we're making it easier for customers to get value quickly and encourage expansion. We do all this while being disciplined in our spending and driving to non-GAAP profitability with record free cash flow.

Looking at the rule of 40, which we measure based on free cash flow yield and ARR growth, we've now improved from a rule of 15 in 2024 to over 24 in 2025. We'll continue to focus on driving top line growth through a disciplined manner in 2026. Now turning to our fourth quarter and full year results. And as [indiscernible] results that I will be discussing with the exception of revenue, are non-GAAP. Our GAAP financial results, along with a reconciliation between GAAP and non-GAAP results can be found in our earnings press release and supplemental financials on the Investor Relations page of our website. Fourth quarter revenue was $91.4 million, up 17% year-over-year versus 9% in fiscal 2024.

Fiscal year 2025 revenue was $343.2 million, up 50% year-over-year versus 8% in the fiscal year 2024. Total ARR increased to $366 million exiting the fourth quarter, an increase of 17% year-over-year and $18 million sequentially. Here are more details on the key elements of the quarter. We had a strong quarter for both new and expansion deals in the enterprise. Platform sales were also particularly strong. of our customers now have multiple products with 74% ARR coming from that cohort.

The number of customers representing 100,000 or more of ARR in Q4 grew to $698, an increase of 18% year-over-year and up 45 customers since the last quarter, representing the largest sequential increase in this cohort in company history. Additionally, the number of customers representing 1 million or more ARR due to [ grew ] in Q4 to 56%, up 33% year-over-year, demonstrating our ability to land significant accounts and grow them over time. In period net dollar retention progressed to 105%, and led by cross-sell expansions across our customer base. 58% of Q4 gross ARR was driven by expansions across a broad range of customers with no individual expansion exceeding $1 million.

It's still driving meaningful progress in that dollar retention. We will continue to focus on driving net dollar retention higher through our platform strategy. Gross margin was 77% for the fourth quarter, flat to fourth quarter of 2024 and up 1 point since last quarter. We continue to make progress on optimizing our hosting, driving multiproduct contracts and monetizing our services engagements. We will continue to look for opportunities to incrementally improve its margin over time. Sales and marketing expenses were 42% of revenue, a decrease of 1 point from the third quarter.

We continue to focus on improving sales efficiencies, driving improvements through our changes in processes, coverage and expansion of enterprise customers. At the same time, we are investing in future growth while balancing those incremental investments with efficiency gains. In Q1 FY '26, we will have higher sales and marketing events as a percentage of revenue reflecting timing of events and our annual company kickoff.

R&D was 18% of revenue, flat to the fourth quarter of 2024. We expect to continue to invest in the talent and capabilities of our team to drive greater innovation in the future. G&A was 12% of revenue, down 4 points for the fourth quarter of 2024. We expect G&A to improve as a percentage of revenue over time. Total operating expenses were $66 million, of revenue, down 3 points sequentially.

Operating income was $4.2 million or 4.6% of revenue. Net income per share was $0.04 based on 141.5 million diluted shares compared to net income per share of $0.02 with 135.7 million diluted shares a year ago. Free cash flow in the quarter was $11.2 million or 12% of revenue compared to $1.5 million or 2% of revenue during the same period last year. In the fourth quarter, we managed our cash collections and made meaningful progress on shifting contracts with annual payments in advance.

For the full year, we had a record free cash flow of nearly $24 million or free cash flow margin of 7%. We have conviction in the long-term value of our platform and have used and will use our cash to minimize the impacts of dilution. We have already purchased in the open market under our current buyback. And given the strength in our balance sheet and the underlying business, our Board has approved an additional reserve of $100 million to be used for buybacks. Our balance sheet position remains strong and allows us the opportunity to be more aggressive in our M&A strategy to accelerate our R&D road map when appropriate.

Now turning to our outlook. As a reminder, the philosophy of how we set guidance is through the lens of execution. We are confident we have the right strategy and the right platform to continue to consolidate the fragmented market. We continue to improve our go-to-market motion and are accelerating our pace of innovation. We have the right monetization strategy to encourage the adoption of our AI tools, and we believe those tools will reduce the varied adoption our full platform, leading to greater monetization opportunities. Our strategy remains consistent with our go-to-market is being aided by our simplification of our pricing and packaging. We will continue to focus on gaining new enterprise customers and driving cross-platform sales with our existing customer base. We also believe that with the release of our AI capabilities, our monetization of data ingested in our platform and the cross-sell opportunities of new products gives us the right strategy to align the value of our customers received with our growth opportunities and grow our business in a profitable way.

For the first quarter of 2026, we expect revenue to be between $91.7 million and $93.7 million, representing an annual growth rate of 16% at the midpoint. We expect non-GAAP operating income to be between negative $4.5 million and negative $2.5 million. And we expect non-GAAP net income per share to be between a negative $0.02 and at a negative $0.01 assuming basic weighted average shares outstanding of approximately $135.1 million. For the full year 2026, we expect full year revenue to be between $390 million and $398 million, an annual growth rate of 15% at the midpoint. We expect our full year non-GAAP operating income to be between $7 million and $13 million.

We expect non-GAAP net income per share to be between $0.08 and $0.15 [indiscernible] weighted average shares outstanding of approximately $145.9 million as measured on a fully diluted basis. In closing, we are accelerating our pace of innovation, and we're growing value that we can deliver to our customers. We have confidence in our ability to scale a durable and growing business while also bringing a genetic analytics to the world.

With that, we'll open up for Q&A. Over to you, John.

Thank you, Andrew. [Operator Instructions] Our first question is going to come from the line of Taylor McGinnis from UBS, followed by Billy Fitzsimmons from Piper Sandler.

Maybe just first on, you announced a number of exciting agent offerings this week. And at the same time, you've also seen good traction with third-party agents connecting into amplitude's platform. So -- and then Spencer, you showed a really good example of being able to extract insight anthropic cloud. So I guess how do you see Amplitude agents and these third-party agents evolving, maybe you just talk about the differentiation that you anticipate with Amplitude agents versus what's being done with the third-party agents today. .

Yes. So to be clear, they both use the same underlying infrastructure. What will happen with either MCP model context protocol is a way for external products like Cloud or open AIs, Chat GPT or cursor to connect into amplitude and request a set of calls. But that is the same infrastructure that both our global agents and our specialized agents use. And so the way to think about it is there's a whole set of tool calls that are available to these agents. You can say, get me a list of events get your attention, get me the list of tools you have like retention and funnels, get me the possible properties for this event. And what we'll do is we'll expose that to an orchestrator that we have that basically interprets a quarry, whether it's in the chat with global agent or whether it's external from MCP.

And then I'll kind of pull in all the different contexts I talked about and then spin out the answers that you see. So it's the same underlying infrastructure because the nature and type of questions are the same whether you're asking it from Cloud or Slack or whether you're in Amplitudes UI.

Perfect. Awesome. And then Andrew, maybe just on a follow-up for you. If we look at the 4Q numbers, it looked like the upside in the quarter was a little bit lighter than what we've seen in the past. Now ARR continued to accelerate. So was that just a function of the quarter being back-end loaded? Or anything to flag in terms of the quarter may be in any areas coming a little bit below as expected? .

So first, I'd say, Q4 was a great quarter for new logo ARR. We had a lot of new customers that are getting value from Amplitude and they're starting their journey with us. Those tend to be ones in which you're working throughout the quarter, and there was a large proportion of ARR that was booked later than we've seen in prior quarters. And as we mentioned before, we didn't see a lot of really big expansion during the quarter. So it was one of those areas where you're building a lot of opportunity for future growth with these new customers. And it's always one you're competing for when you're able to do a new logo, you have to really compete and show value and sometimes it takes a little longer as well. But we're really pleased with all the new customers that have become amplitude customers, and we think that, that sets us up very well for expansions in the future.

Our next question will come from Billy Fitzsimmons from Piper Sandler, followed by Rob Oliver from RW Baird. .

I guess maybe to start, can you help us think through the NRR improvement? And how much would you contribute or attribute, I should say, to greater upsells and cross-sells or maybe better more success in kind of mitigating some of the churn in the business.

Sure. So throughout the year, we've seen our customers increasingly adopting more and more of our applications in the platform. When we started off 2 we specifically were training our sales team, how to sell our platform and we're introducing new capabilities. We acquired new capabilities, and we put those into our platform as well. So predominantly throughout 2025, the improvement in net dollar retention was related to our cross-sell capabilities. And as you were alluding to in the past, we had situations where we were overselling capacity against analytics. And even with some customers increasing data, it wasn't enough really to offset and contribute materially towards net retention. Now that we're past most of those capacity-related issues that we create for ourselves. We're starting to see customers and their data ingesting platform contribute towards net dollar retention improvements as well. And so now as we think forward, and I've said in the past that we have full intention to continue to set up our customers and expand with our customers, introducing new innovation. We think that both vectors, both data ingested into the platform, meaning upsells as well as cross-sells will contribute to further improvements.

Makes sense. And I guess on that note, if I could sneak in 1 more. Can you give us a sense of the role volume upsells will play in the FY '26 growth algorithm, especially as you start lapping some of the contract rightsizing from the first half of last year. .

So one of the things we talked about in the call was introducing our new pricing and packaging that is aligning not only to our enterprise motion but also towards the implementation of our new AI products. And in the past, I would say there were certain times where customers felt very leery about the amount they have to pay based on increasing data rates that we're ingesting in the platform that those rates were so high that they wouldn't be able to see the benefits associated with marginal incremental reductions in the cost of that data. Well, our new pricing and packaging structure rewards our customers now for adding more and more data into the platform so that they're paying marginally incrementally less. Now that doesn't mean that it's not going to contribute growth to amplitude. Our customers are getting greater value by ingesting more data in the platform. We believe it's fair for us to have some of that fair exchange of value. But if you were going to ask where we are really focused on driving NRR and where that -- the biggest benefit it will be continue to show from those cross-sell opportunities that expansion of our products because we want our customers to not fear adding more data. We want them to take advantage of implementing more data into our platform, and we want that to scale, especially as they look at longer-term contracts with us.

Our next question will come from Rob Oliver Auer from RW Baird followed by Clark Wright from D.A. Davidson.

A follow-up there, Andrew, on the pricing and packaging question. So obviously, enterprise is really like predictability. You guys have never been a seat-based model. So if you can just help us understand in the context of the new pricing model, clearly, it sounds like it's driving more engagement, a cross-sell opportunity, less of a friction experience. But how does the buyer manage that predictability? And I guess the inverse of that would be how do you get comfortable on the cost side with AI embedded in.

Yes. It's a great question. We spend a lot of time working with our sales team and our customers and showing how, one, the instrumentation in the platform can have -- give the great visibility into the data they're ingesting within it. And we work with our sellers to help them better understand as the marginal incremental data into the platform, grows, how that then translates into the cost that we're going to be charging to our customer. We're encouraging to have that conversation as part of the sales process. It's a kind [indiscernible] way of showing and working with the customer on how they are going to adopt Amplitude over a period of time rather than guessing what their data implementation of the platform is going to be. We're working with them very closely on it and showing how the instrumentation works. Now the piece that I think is really important, and you touched on it, but I think it's -- we did a lot of work with customers to understand whether we had the right billing metric where it's something that they aligned to the value proposition. We've been testing for quite a while. In fact, nearly 20% of new ARR that we booked in the quarter was actually using our new pricing packaging pilot stage. So we already know that customers like this. We already know that customers look at it as more transparent. They look at it as less friction as you were saying, we also believe it positions us very, very well given that our focus on implementing AI products into our platform is, one, it's reducing the barriers to adoption. Meaning that customers walk away thinking they're getting great value of what they've already invested in amplitude and are less fearful knowing that they have greater cost predictability and transparency and how that use is going to trend over a period of time.

Great. Super helpful. And then and then Spencer, 1 quick one for you. Just on Infinigro, you guys were very early to the AI acquisitions among our coverage, I think in very aggressive on it. And in particular, it looks like to us like this gives you guys a further opportunity to sort of go for that. consolidation play that you guys have talked about.

But if you could help us maybe understand what in particular what area or what response to what customers need in cider is going to help address and how that might accelerate that platform opportunity.

There were 2 big things that stood out to us on the Infinera team. I mean, so first, we're just always looking for great talent out there. And so when the right company and the right opportunity comes along and they are aligned with our vision and excited about it, we're going to act. With Infinigro in particular, there were 2 big things that stood about the team. So Daniel, the CEO there as well as the rest of the group, they've been in it on AI analytics and automating workflows for the last few years and have a ton of perspective on how the future of the category is going to be shaped. And I mean we're on charter territory, like we're inventing something new here, AI analytics. And so whenever you get a chance to partner with someone else who's thought about that so deeply, it's a huge deal, and we're going to -- yes, we want to figure out how we can set up a way to work with them. So that's the first one that really stood out about InfiniGo. The other piece that stood out is they have a lot of familiarity with analysts more on the marketing side versus product management. and particularly as those personas merge over long term and more customers from legacy Martec tools want to come off and use something bleeding edge like an amplitude we want to make sure that we're ready to meet them and serve all their needs and help with that transition. And again, they know a lot of those buyers better than almost any other company that we've seen in the analytics space out there.

Our next question will come from Clark Wright from D.A. Davidson, followed by Koji Ikeda from Bank of America.

You noted the cross-selling opportunities continue to be an area of strength, what is the natural pathway you're seeing in terms of product adoption? And what is the role that agents are going to play going forward to help drive additional cross-selling motions?

I mean it's great on both fronts. So analytics is the core. We're an analytics platform, something we've been very consistent about. You want to be able to track the core -- the base -- that foundation of the user journey. And that makes every single other part of the platform more valuable. So it makes it easier to do experiments because you can target users as well as measure those more effectively, it makes a better do session replay because you can understand, hey, for a group of users that ran into this error, let me see what they did by looking at the session replays and makes guidance surveys better because you can target guides to specific users based on if you see them confused. So analytics is the core and all of these -- they become more valuable with analytics and vice versa.

In terms of agents, I think the big opportunity there, and I just showed the session replay one is that these other products are -- while we launched AI analytics yesterday, and that was the main focus. These other products are actually capable of being leveraged by AI. So the session replay specialized agent demo that I shared earlier is a great example where you can watch 1, 2, 3, maybe 10 session replace, but watch 100.

I'll take you a few hours to get through them. And so to have an agent speed up that analysis still get all the valuable data, summarize it up and kind of put it back to you. I mean that's you're talking a 100x fold increase in productivity versus what you might other do. experiment is the same thing. I mean one of the things that people ask us a lot is like coal, do you have a library of best practices for what sort of web pages or what sort of interactions work and what don't. And our experiment conversion agent will actually suggest those based on best practices of what we know from all the companies that we work with. And so it makes experimentation a lot more powerful, too.

So -- and then the really cool moment is when these tie together. So you can start out in analytics and say, okay, cool. Give me my unhappiest users and suggest ideas for that I could do to improve them. And then it says, "Wow, okay, all of these users, they're unhappy because they ran into a page that wasn't working, and then you could have a replay agent come in and say, okay, well, let's look at what was it at Page. It's like, "Oh, okay, hey, this button isn't formatted correctly and wasn't labeled and so that's probably confusing to users. And then you can -- and then go even further and say, okay, great, let's run. Can you propose a variant, an experiment variant to me that would actually fix it? And then it will propose it and propose another web page and you can run the test. And so you can not know anything about analytics, not know anything about your data taxonomy, not know anything about how you think [indiscernible] how to do experimentation AB testing and do all the work of all of those projects products from the global agents or specialized agents interface. So I just -- it's going to be a massive unlock in terms of the usage. We're obviously most focused on analytics right now, but I'm really excited about some of the other things [indiscernible] funny. We already got some comments on Twitter that are like, hey, why does it only watch 100 sessions at a time? Why can't you watch 1,000 or 10,000 like, "All right, we're working on it. We're working on it.

Appreciate that. And then, Andrew, there's a reference to increasing win rates versus point solutions. Is that an output of the go-to-market changes as well as the pricing and packaging updates? Or are there any other factors that is helping driving improvements in that metric? .

I'd say the pricing and packaging is relatively new, so I wouldn't contribute that necessarily increasing win rates. I think that the biggest thing is, one, our sales team has just worked really hard at demonstrating value of our platform to our clients. and that's really resonating. And the other is you really have to credit our product team for creating just really great products that work well together. A lot of people claim they have a platform, but the reality is that's a bunch of products that's stitched together, doesn't look really well. When you have a platform, you have workflows that are instrumented well and it's easy to interact with the different modules of the product. And that's the way I would characterize our platform today. And every time that customers are adopting more than 1 product, it's because the -- that integration, those workflows seamlessly across our platform are coming through as real value. I mean I talked to a number of customers myself with the sales team and they always come back and say, we're just so far ahead of what everybody else is even representing an analytic platform to be.

On that, like if I just go through the last 30 buyers I've talked to in the last month, all they want to do is be educated about analytics. And sorry, how AI is going to transform analytics and the whole platform. And they see it coming. They see tons of automation ahead and they're like, teach me how I can be relevant. And so when we can offer that to them by, one, providing a view on how the future unfolds and then two, offering them the products, tools and services that actually enable them to be successful and relevant, they want to spend a ton of time with us. And so the competitive question especially against the smaller point solutions is kind of going away. It really is just is now the right time. And can you help me get to this future fast enough and teach me.

Our next question will come from the line of Koji Ikeda from Bank of America, followed by Jackson Ader from KeyBank.

This is George [indiscernible] on for Koji. Taking a big step back, 1 from me on kind of the big picture, where can we expect identic queries to grow to become in the mix from 25% today, maybe over the next 12 to 24 months? .

Yes. I mean I -- none of us have a full crystal ball, but my expectation is the vast majority are going to be done agentically, where you're just going to have agents that run over your data all the time. They're looking at dashboards. They're looking at KPIs, they're trying to find underlying root causes of why things are changing. They're creating suggestions for your product. They're reviewing session replays. They're constantly trying out and tweaking new experiments. So I mean, yes, like I don't want to put a number out there, but I think the vast majority. I think what we're seeing generally is if you look at query growth from direct usage of the [indiscernible] dashboards, it's increasing in line roughly with the size of our business. If you look at a genetic core use, it's skyrocketing you saw on that chart in the last few months.

And so I -- the amount of leverage, I think the same thing happened to coding in the last 2 years, where if you look at it, the best software engineering teams, the majority is produced by the majority of lines of codes are produced by agents. It's mostly humans editing interpreting them stitching them together. And kind of giving high-level direction and that's where the best software engineers are. I think the same thing is going to happen in analytics and data analysts, where the vast majority of the data munging of the tool and figuring out what quarry means what thing? And how do you get to how do you do a segmentation, understand the root cause, like that's all going to be automated to agents. And our goal is to be the first company to do that in a big way.

Our next question will come from the line of Jackson Ader from KeyBanc, followed by the line from Scott Berg from Needham.

This is Nate Ross on for Jackson Ader. So implied non-GAAP operating margin for 2026 is roughly 2.5%. And I guess we were wondering what specific possible sources of upside do you guys see for that number? .

Well, I'll tell you, first and foremost, we've been on this path where we're increasingly driving revenue growth faster than we're driving expense growth. And rooted within that is efforts on changing our go-to-market, changing our processes, modernizing our own application architectures, doing the basics of running the business in a very efficient way, so we continue to go drive growth with leverage. Those same things are not just a onetime event. You continue to focus and learn and understand how you can drive and deliver services more effectively. And so we look at the path we have in front of us with respect to our growth opportunities, what our pipelines look like, how we're managing our cost to serve with the expectation that sales and marketing will continue to improve on their efficiencies, G&A will continue to drive efficiencies as well. So it all kind of culminates in the plan that we put together for 2026.

Great. Perfect. And I guess, one more follow-up. So regarding customers' analytics budgets. Have you guys noticed any trends or changes specifically with AI affecting these budgets the current AI landscape affected customers' propensity to invest in analytics in any way? .

Yes. So I'd call it two things. I mean one, it becomes -- it's the bottleneck, right? So you remember that [indiscernible] showed at the start, right, it's like, okay, you're shipping all the software. Is it good? Are we even going in the right direction? So the comparative value of the analytics piece becomes a lot greater and more higher urgency. When you have like a year-long road map, it's okay if it takes a while to measure the success of it. But when your iteration cycle is measured in weeks or days like it is with the best of the best companies now, it's like, yes, you need to know if you're going in the right direction all the time.

And then I think the other thing is the buyers in addition to that being the pinch point for and the big need in terms of the next big step in product development, they know -- they intuitively all know that this whole space is going to get reformulated with AI. And so they -- again, they're just desperate for education and someone to show them the way. This is not a case of like -- I think one of the differences between selling SaaS and selling AI is in the SaaS world, it's very much like, okay, talk to your customers, get a list to prioritize features from them and build it and you go back and sell it to them. In this AI world, they don't know. Like they are like, is the model cable of this? Can it automatically look at a session replay for me? Can it analyze the root cause of a breakage in my funnel. And how -- what's the best way to make that happen. And so they're looking at us for all those questions. And this is where sharing the vision of what the future is as well as being close to the leading edge of the technology is super important.

Our next question will come from the line of Scott Berg from Needham followed by Nick Altman from BTIG.

This is Sean Black on for Scott Berg. With the new pricing and packaging, are you planning on separately monetizing your AI agents?

So most of our AI agents are embedded within our core platform. And so what you see there is we're getting access to customers to utilize more of the platform that exemplifies all the power of our modules together. So there's a high propensity that customers who are utilizing our AI agents are both ingesting more data into the platform as well as expanding into other modules. Now we're also going to introduce new products with continuing -- we've done really well at innovating. And some of those products will come out with a fee charges as well. So we're not worried about the ability for us to monetize our AI capabilities. We're actually very excited about the opportunities as we expand the use cases and usage of our platform.

Congratulations on the good quarter.

Our next question will come from the line of Nick Altman from BTIG, followed by Elizabeth Porter from Morgan Stanley.

This is John Gomez on for Nick Altman. With the shift to Gentek democratizing the end user product analytics, can you just talk about whether you're seeing new users or new lines of business leverage amplitude and how that's shifting to go to market. So just any commentary on new end users and how that's impacting how you think about the go-to-market strategy would be helpful.

It's not really new end users. I mean it's the same. You're talking about product teams. You're talking about marketing teams, engineering and data teams. And so it's the same people trying to leverage the data. What they're really -- again, what they're really desperate for is education. And so when we can show them global agent specialized agents, MCP, AI feedback, AI visibility what we're doing with our next products and assistant and all analytics. There's always a whole bunch they want to grab on to and say, okay, great, teach me, use this, make it successful, everything else. So that's the biggest difference. It's one where our go-to-market is -- it's about training, getting them to be able to educate, to be able to share the vision, to be able to demo these products and make customers successful.

Our next question will come from the line of Elizabeth Porter from Morgan Stanley, followed by YC Wang at Citi.

I'm Lucas [indiscernible] here for Elizabeth Porter tonight. So with the uptick in new app development that we've seen over the past few months, could you walk through your expectations for balancing this potential new demand from smaller customers with your move-up market as you evolve your go-to-market strategy?

Yes, we're doing both. I think one of the things we see on the start-ups and newer customers is they're very bleeding edge, and so they're trying to push the capabilities of us. And so we've always had a motion where we've taken innovation that we've done with them and bring it to the enterprise in a deliberate way. So we're going to continue to do that. I think there's actually massive opportunities, particularly with the rise of 5 coated apps. There's going to be 5 coated analytics, too, that needs to go along with all those applications.

So it's early, but there's a big opportunity. for us there, too. Again, the core thing you want in terms of understanding your customers and knowing if you're going in the right direction and building the product is the same, whether you're -- the newest start-up that was just founded yesterday or year-old long-lasting business. And so for us, it's like we're here to serve all of them. And again, they're very keen on learning the bleeding edge of what's happening in AI analytics. And so if you're able to teach them that, then it doesn't matter [indiscernible].

Got it. That's super helpful. And then could you speak to the 7-figure deal pipeline in 2026? And then are there any specific verticals in which you see outsized growth already?

I mean we're seeing -- like as I said on the call, we're seeing a lot of AI companies use us. We have 25 over 100 and then we have a 7-figure contract with one of the largest foundational model labs out there, who's been a customer starting last year. And so that's very, very exciting because they obviously know what's going on when it comes to what's possible, and they see a future world where we're a really big part in that.

Our next question will come from YC Wang from Citi, followed by our last question from Arjun Bhatia at William Blair.

Congratulations on a pretty strong close to the year here. I guess maybe I just want to touch on expense. You talk about MPT being one of the largest days of user behavior. Just given the rapid progress of agent capability across our data platform players caused [indiscernible], where we are seeing also customer consolidation towards maybe bigger data platform players. Curious if you're seeing I think customer of like your analytics use cases from Snowflake [indiscernible] versus Amplitude. .

I want to make sure -- I want to make sure I understand what you're saying, you're saying do we see competition from Snowflake and Databricks because data [indiscernible] to.

Well, it's not just data. They are thinking about doing the application as well. I mean if you think about white coating and then you think of application building, you can make it easier to build. I'm just curious if you see any being blurring between customer talking about just use cases between a customer you can use a data platform like Snowflake, they have Cortex versus what you will see with Amplitude?

So one of the big things that we see is that customers always want the most advanced and bleeding edge capabilities. Like I heard this great analogy the other day where software is very much like Sushi. So it's fine that the gas station at 7/11 offers it. But [indiscernible] in Japan is not probably not going out of business. In fact, it's going to create more demand for him. And so from our standpoint, what we think about is how can we offer the most advanced and robust system for analytics. So if you look at the benchmark that we -- with hundreds of evals that we released, where we got a 76% accuracy rate, if you look at the cortex or you look at DataBrick [indiscernible]. I mean they're going to be in the 10% or subbed. We're working on releasing full metrics on that. And that's because the Texas Sequel is only really one part of it.

The other -- there's 2 other big parts. The first is the contact layer. So what data sources are you bringing together in the right way, analytics data, session replay data, data from interactions and guides and survey data from other sources and interpreting those in the right way and then giving an LLM agent, the right set of tool calls so that they can iteratively quarry, okay, hey, what's my [indiscernible] funnel, where is the biggest drop on it. Why is the biggest drop on it? What's the biggest difference between users who went to the next versus the previous step, right? So that's like just in that example, that's 4 quarries that you're going to have to do in a row all correctly.

And to do that, you need to prop the [indiscernible] in a very particular way. You need to give the right tools, you need to give it the right context. And so we have thought really deeply about because we have the largest user behavioral data in the world, we have thought very deeply. We have seen what millions of analytics quarries, what good looks like for millions of analytics quarries and translated that into an agent that does the same. And so because because, again, you're going to need to give it all that context and then be able to [indiscernible] a data system, the differences and accuracy are really, really starts to roll your own or use a [indiscernible] or Cortex versus using an Amplitude. And when you're an analyst, that difference between 76% and 10% is a massive difference in terms of your ability to leverage agenetic analytics.

No, that's helpful color. Maybe one for Andrew. The profit definitely came in well ahead of expectations here. Maybe you guys are leveraging some agents internally that helps to drive better sales efficiency. But curious to see going into next year, like what can we expect, especially on the free cash flow that outperformed probably you saw an expansion by 4 points, like curious to see what your expectation in next year? And any other moving parts that we should be aware of or headwind from to be made from the strong performance this year?

I think what you're seeing is that the efforts we've been doing on sales and marketing, on our cost to serve and our G&A and operating more effectively as a company, is not just a 1 effort, 1 activity, there's multiple certainly, we're introducing agented capabilities into our own workflows within the company, and that's certainly contributing to it. But there are so many things structurally we've done to the business to create greater durability that that's ending in greater abilities for us to drive efficiencies. I'll give you one example. We've talked a lot about our ability to go drive increasing contract duration to our customers and that our RPO has been growing rapidly. if you don't have to renew your installed base every year or that installed base percentage goes down because you're executing more and more longer-term duration contracts with your customers, then the sales team has more time to dedicate towards selling new and expansion deals rather than working on renewals. And so it's just a great example of a strategy we put in place that's going to accrue benefits for a longer period of time.

And our last question is going to come from Arjun Bhatia, William Blair.

I'm Willow for Arjun Bhatia. Andrew, in terms of guidance, the full year revenue range seems a bit wider than normal at $8 million. Can you help us understand the reason for this? And what scenarios are contemplated at the low and high ends of the range?

I think when we approach, we approach our guidance. We approach it with what we think we can go execute in the period. And I wouldn't read too much into that other than we have a breadth of different opportunities that we're going to with our product set, with improvements in our targeting enterprise customers. So I wouldn't read too much into it.

And that will conclude our fourth quarter earnings call. Thank you for your time and interest, and we look forward to seeing you on the road this quarter as we attend conferences hosted by Baird, Citizens, KeyBank, Morgan Stanley and others. Take care.

Okay. Hi, everyone. My name is Taylor McGinnis and I head up the software application SaaS equity research here at UBS. And in this session, we have Amplitude's CFO, Andrew. So Andrew, thank you so much for joining us.

Thank you for having us. It's great to be here.

Perfect. So maybe we'll just jump right into the current momentum that Amplitude's business has been seeing. So you guys have had an impressive reacceleration to 18% growth. So maybe you could just give a little bit of insight in terms of the current demand environment you're seeing? What Amplitude has been doing over the last several months, plus the last couple of years in order to position the company to really capture some of these emerging new opportunities?

So I'd start by saying there's a couple of big things we've been working on for a while. Before I even joined, the founders were looking at the analytics ecosystem and all these other companies that have been created in experimentation and session replay and guides and surveys and activation and said, look, some of these companies, they really are not whole platform. They are by themselves creating one part of an ecosystem, which really should be consolidated.

And so there was an effort to start building those products around analytics and started with experimentation and then to activation, and then on to web analytics and guides and surveys and session replay and on and on. So we created this environment where we're setting ourselves up for a true strong platform play. Even though we weren't calling it at that point in time, we're saying like the value proposition for clients is one that all these things are together.

Now about the same time, Thomas Hansen, our President and COO, who runs all go-to-market joined and said, if we're positioning this as a broader sale and really going after Google and Adobe, especially in marketing and analytics, we should be focusing more on an enterprise sales process. And he started then that advancement of our go-to-market team, moving from very much a transactional model over to an enterprise selling model.

And they were going along. And I joined in August of 2024 and said, "Hey, guys, there's a lot of glue and gum on this process. Let's make sure this is actually going to result in the financial outcomes that we want". So we started building and instrumenting our processes better. We made a lot of changes that led up to the beginning of 2025. The products were coming along at that point in time, and it was a very strong value proposition for clients, looking to consolidate and reduce vendor spend and drive greater outcomes.

So I would tell you, the growth that we're seeing now, this is a discussion I have with the board just yesterday. The growth we're seeing now are the results of some of that core instrumentation we put in place late in '24 that's now coming to pass. And I think we're not done. We're doing a number of really good things this next year in order to really drive sales productivity higher, reach out to customers more effectively, drive better process optimization and drive greater profitability in our business model.

The -- you know this, but for others who don't, I've been talking about our planning is one that focuses on driving that growth on those 2 major areas growing more enterprise and driving our product portfolio, but also doing it in a way that we're doing -- showing leverage. We're showing increasing profitability associated with that operation.

Yes. I want to touch a little bit on the former and what you were talking about earlier, which is you've made a lot of changes in the business to get to the growth levels that you are today. You've seen significant improvement in NRR. I think last quarter, it was around 104%. So when you think of what's driven the improvement to date, how much of that was coming from the cycling through of a lot of these optimization efforts that we saw post COVID and the rightsizing?

And I think that's an important question to ask because then it will tell you, okay, now with all these expansion opportunities and everything you've been doing on the execution side, like what's actually left to come, right? So can you impact how much of what we've seen so far is just cycling through those efforts? And actually, what does that mean for the durability and improvement of NRR and top line as we look ahead?

Yes, because some of those things are actually in different ways instrumented as well as and durability. So I'll come back to that one. But the first 2 things I would say is you really have to understand it that Amplitude is not a traditional seat-based SaaS platform. We have 2 major mechanisms through which we monetize.

First is our classic, which is our product analytics level, we ingest event data into the platform. And event data is you can think of it simply as any type of user system that interacts with a product or a website or other technology that's creating some type of cursor click or an engagement or a movement from one place, that creates an event at a very detailed level.

And that's one way in which you monetize. The more event information that is ingested within the platform, the more we will charge higher and higher ARR. And since we've done our pricing packaging changes, which is another question we have, I would tell you that's at a decreasing rate for customers now.

The other mechanism is that we charge for additional modules. So you have both an upsell and a cross-sell mechanism in which we monetize, and the reality is, as you mentioned, there was a lot of overbuying during COVID. But I also say when you're going from a transactional sales model to one that's more value-oriented, there were a number of things that we created headwinds for ourselves, like some large contracts that had planned downsells upon renewal. Same capacity, but a reduction in the ARR. Like, why would you ever -- why would you ever put that in the ARR? Well, that's what happened. And we had to go remediate some of those things.

So if you look at over the last year, what's been driving our growth, we have had an increase in the level of data that's being ingested in the platform. But we also had these headwinds around overselling capacity that wasn't being used, and we had to rightsize. And in some cases, you lose contracts as well. That side of the upsell business has been relatively meager. So the growth has really been driven by the success of our new product offerings that we've been surrounding analytics.

Now here's the good news. As we progressed and gotten better in remediating some of these issues, we start to open up the ability to grow via upsell as well. So now we have both vectors that are starting to work for us. And that's what's the exciting part because the innovation pace for us has increased. And as we add our new Agentic capabilities that are coming in the January time frame, we're going to have a really great product launch around our new agent capabilities, our MCP server and a number of others that we haven't talked about much yet, but it's very exciting.

We got a demo of it at the Board the other day. We think that, that not only opens up more data ingestion in the platform with more use cases especially when you think about like MCP, which opens you up to whole different data sources that you can have behavioral heuristics exposed to, as well as it really emphasizes the power of our platform and one large set of applications working well together.

Yes. And I believe Amplitude has ambitions ultimately to get to 20%-plus revenue growth. We're not that far off from those levels today and getting NRR back to 110 plus. So in terms of what's still needed to get there? Do you feel like all of the changes that you've made over the last year plus position you well to go capture that, and it's really just a function of execution? Or are there still things that you guys need to do on the product side, go-to-market side, whether it be pricing and packaging that you mentioned earlier, like anything left to really go capture 20% plus more durable growth?

So when I joined Amplitude, I told Thomas, look, there are 15 things that I've identified we need to go change in go-to-market. And I would say we're about half of those things we've done at some level. And it takes time to drive change management. It's the hardest thing. It's not about the what. It's about how you do it and how fast you can do it without creating major disruptions. I mean, the first major thing we did was we moved the quota basis for our salespeople from gross ARR to net. And I always tell people show me a company that has a gross ARR quota basis, and I'll show you a company that has a churn problem.

So we changed that. We started teaching our sales reps how to sell the platform. We implemented a lot of changes in the comp plan design and emphasized more durability in our business and sales process, leaning in more towards multiyear contracts. And I can tell you when a customer looks at our platform and they see they want to replace multiple applications, they can't do it all at once. And many times, they have contracts that are with other providers that are staged over a period of time. And so they are looking for a multiyear engagement as well and cost predictability.

Well, you do all those things in the instrumentation side and you fit it well within the sales process and the value proposition to clients, and you see some of the progress we've made in our RPO. It's growing quite rapidly, especially in the long term. And why is that good? Because we're not having to overly discount for that. It gives us great revenue visibility. And here's the secret that a lot of people don't understand about the instrumentation part of this, which is when I first joined, we had to renew 89% of our installed base, okay?

In the back half of '24, we started doing more multiyear contracts and deal constructs that leaned into the multiyear agreements, but we only had 6 months to go after it. But that led us in '25, to only have to renew 74% of our installed base. And the progress you've seen in Q1 through Q3 would lead you to believe that that percent is going drop even further as we go into next year.

Well, what happens when you have an installed base that is much lower as you have to renew, you have one, less risk on churn dollars, just basic math, my gross retention is going up, that's great. It's applied to a lower balance. Guess what? Churn dollars are lower. But the other thing is you get a lot of capacity back in the sales team. They don't have to spend as much time working on the renewal. They can spend much more time going after new logos, and they can spend much more time in driving the value proposition of expansion with our clients.

Only 39% of our customers are actually on multiproducts, there's a huge opportunity just going into our installed base and just doing expansion. But we know too that some of these new products we're bringing out are opening up the aperture of where we can have use cases, especially in marketing and analytics.

So as we move into next year, focusing on driving many, many more enterprise engagements is where we now have the capacity, even though we're not having to bring on a lot of new reps. That helps with sales and marketing as a percentage of revenue dropping. So all these things I was explaining to the Board yesterday that we've done, it's helping us to instrument the financial expectations we have on our multiyear plan.

And so it's not a question of what anymore. It's how fast you can do it to achieve the results you expect. A great example of that is on the sales side is like when you do territory redesign, you can't rip apart all of your account and rep relationships at once. It would create chaos. So you've got to be very methodical about how you do it, and making sure that you're setting up new reps in new territories with existing customers to talk to.

Well, if you do that, you have to take existing customers away from the reps. So that level of churn in chaos is one that you want to -- you know you have to do, but you want to kind of keep it like 30%, 35%, not have a tremendous amount of churn. And if you do that well, then the system starts to expect it. And as the system expects it, then you're ramping your reps faster, they're reaching attainment faster, your attrition falls down, you're instrumenting the outcomes you're expecting.

Yes. And to your point, you have to be very methodical, especially on these older cohorts of customers. But maybe a way to bring better clarity to this is what are you seeing in terms of NRRs with the newer customer cohorts? So the post COVID, right, where you've had some of these changes potentially implemented faster, like what are the expansion rates look like there? And could that give us insight into what the older customer cohorts could potentially ultimately become?

Yes. So I think that what you'd find is a gross retention rate you think is nearing best-in-class in some of those, especially given the -- many of them are multiproduct when they're adopting and the best indication of a return customer is one that already adopt multiproduct. So we don't even have a long history of that.

So you can't tell, "Hey, well, it's a small cohort". But the reality is we're seeing expansions happen within the year to the tune of 5 to 20x your initial land. That happened with Empower, where we started very small and it went over to $1 million deal. I can tell you there are multiple customers I've talked with this quarter where we're starting off with a very small and the expansions are showing up or we're even landing with much larger deals.

So go back -- that goes back to the value proposition of selling the full platform and the customer journey and what they can expect as they adopt more and what outcomes we're instrumenting in the sales process. Those were things we put through the instrumentation I'm talking about and how we're changing our go-to-market. It's dovetailed very well with the products themselves. You can't have one without the other. But -- but yes, that's -- it's showing up very well. And I would tell you the dollar retention rates on customers in the enterprise, whereas I'd say that as we've shifted our focus on enterprise, they are 3 and 4 points higher than the average.

So the more our business becomes enterprise, and it's creeped up every quarter, the less and less you'll be asking me that question because the reality is enterprise customers are so much better than in the SMB mid-market with respect to gross retention and net dollar retention.

Yes. Let's focus in on the cross-sell opportunity. So -- when you -- so a couple of years ago, Amplitude was primarily product analytics. And over the last several years, it's really been building out to become a suite of a number of solutions. I don't know if you can give us exact numbers, but maybe just even providing color to help us quantify it, how much of the business is product analytics versus nonproduct analytics today? And when you have a product analytics customer and you upsell or cross-sell all of these other solutions, like what impact does that have to average deal sizes?

So I'll answer the last question. It's much larger. So average deal sizes have gone up quite dramatically with multiproduct. And part of that is the way we price. It's -- you pay for your analytics and the data associated with your use case is downloaded into our system, and that becomes the cost to serve, if you will. Session Replay has a little bit of that. But the other modules have very low marginal incremental cost to serve, and yet we're able to -- because of the value that they deliver, we're able to add on to that value.

So you can get a customer that is analytics, it's call it, 30% of value in the contract, but the rest of the modules make up 70% or more. So when we add on to an analytics customer, it's a 25% to 30% uplift typically by module. And what I'd tell you now is that already the way we disclosed it, and then I'll tell you what you're asking, which is the customers that have multiproduct now represents 71% of our ARR. Only 39% of our customers are multiproduct. So there's a big difference in ASP.

And then the other thing is, 2 years ago, we had virtually 0 revenue from other products. Now it's upwards to 30%. So you're seeing us leverage the strong product analytics foundation and broaden it into these broader use cases and customers are seeing and paying for the value of the activation layer that we've added.

Yes. And if I put AI aside because we'll touch on AI in a moment, but it seems like there's 3 big opportunities. So we've been talking about cross-sell, right? So I'd love to hear you've made an amazing improvement today, but where could that momentum go? And how are you thinking about that?

The second piece would be, I think, historically, Amplitude has been focused on product teams, and there have been efforts to push more into marketing and other departments. So two, can you talk about -- not to throw too many things, but two, talk about that opportunity. And then the last one, I think the big early adopters of Amplitude were truly the digital native companies, and there's a big opportunity to go after those traditional established conventional companies. So where are we on the tipping -- like the tipping point for those?

Yes. So the -- on the more traditional companies, I would tell you, every day we see health care, financial services, media and telecom -- that kind of First American titles when we shared it on an earnings call. That's a title company around transferring property and like their business process automation was digitizing that application environment, and they're using Amplitude as the underpinning for it.

A number of other -- Sutter health. I don't know if you're -- I'm a Sutter Health. My Sutter Health, Amplitude is the underpinnings of my Sutter health. And just a number of them that in very traditional industries that are figuring out ways in which they want to digitally engage with their own clients and are instrumenting and using Amplitude to do that. I don't know what your favorite fast casual restaurant is, but we have a lot of them under contract, Jersey Mike's, Chick-fil-A, Burger King, Dairy Queen, you name it. A lot of them have figured out that if they want to have a loyalty program with their clients, they need to meet them where they are and often that is on their mobile application.

And so they're delivering promotions and engagement through that environment, and they're using the Amplitude to do it. So I think there's been a great push, I think the progress that we will have in more, call it Global 2000 and less digital native clients will be the pace at which they're engaging digitally with their clients.

And for us, we think that as we move into adding more and more agentic capabilities to our platform and add more products, the reality is that's going to spur that interest and engagement. In fact, one of the things we showed on our earnings call, we know we do demos. We do demos of the product in our earning call was the whole MCP server capabilities. And for us, it was about opening up the aperture around who can engage with Amplitude without having to have a data science degree, understanding the data taxonomy or be a marketing analyst or a product analyst.

And you'll see in January that same prompt-related interface is both going to be within Amplitude and where you can easily generate charts or ask queries or you could easily engage it through your favorite LLM model.

Yes. Let's talk AI because Amplitude has been really innovative on that front. So you talked about the MCP servers, Amplitude's introduced AI agents, even more recently, you guys have entered the AEO/GEO conversation as well, too. So there's a lot going on at Amplitude. But just in terms of the areas that you as a CFO, right, and you think about the impact that could have on retention, right, or direct monetization, which one are -- maybe there's a couple are you most excited about? And what has early customer feedback then around adoption and readiness for these solutions?

So I guess I'm probably most interested in the core platform areas where we've had agentic capabilities. Like if you have an agent that can actually do experimentations continuously. I've talked to customers that talk about how they run an experiment every 2 weeks, and it's one. And they have to involve all these different parties within the organization, IT and data analysts and the engineering team, and they have to have handoffs and sign-offs.

And this discussion I have with I would call our marketing analyst, they have different names. We're talking about person, but she's the marketing house. And she leaned over to the IT person, let's call them Tom and said, "I don't want to talk to Tom. I just want to run the operation experiment knowing that he set up the framework directly in the data tech, they don't have to ask".

And it's very funny because Tom, I'm on. But the funny thing is, she's right. She's right, that is what they need to do. They need to increasingly think about how they can run the experiments more rapidly and the agents can do that for them.

You can set up the framework and run thousands of experiments continuously and make recommendations and propose a guidance survey engagement or a cohort targeting, different ways of looking at the analysis, deep dives into the reason and the causes of what you could do with it. So experimentation is a really, really interesting area. And I call it the death of the data scientists because you're going to need a lot less.

Another one is a evaluation of your Session Replay, that's -- if you think about Session Replay's video. It's watching people's cursors and how they're interacting with your application environment. Well, One of the things that agents are really good at is taking lots of data and summarizing it for you. Now I think in terms of we have an agent that's constantly doing it and providing you the right actionable insights from that qualitative and then quickly engaging workflow, how you can make the changes to your website or colorization. Now it's fast forward.

As you start to make that more autonomous, you enable personalization such that the application website that you would look at would be different from the one that I look at based upon the batch size of one. They're making the changes based upon the behavioral heist that you exhibit versus me. And that's something that's pretty powerful when you think about it in terms of targeting, marketing, advertising because you can then optimize your conversion rates associated with them.

So I'm really bullish about some of those things, and I see the platform becoming more accessible, less about the UI and understanding Amplitude, but more about Amplitude being catered to your business needs. The other areas on opening up the data, MCP is very, very interesting because you get access to not just what's ingested into Amplitude, they get to support tickets to edit feedback to how you're -- how your CRM data is structured. And it becomes less of a task to ETL information from one place to the other because your data taxonomy becomes something that is -- think of it in terms of -- it's not a relational database required anymore. It's unstructured data and the agent understands that.

So those things are exciting. And I also think some of our new products are pretty cool. I'm -- when we acquired Kraftful, they were in the voice of the customer space. And if you talk to Yana Welinder, who was the founder of that and is now part of our leader in our engineering team, to tell you that we're going to go disrupt Qualtrics and Medallia because they're doing basic survey stuff and that's one slice of customer sentiment, whereas that's important, you should have a broader sense with many more channels to establish that sentiment. And if you're not looking at how customers are interacting on blogs and websites and feedback areas, and you're missing some important feedback.

Yes. How are you thinking about monetization as it relates to all of these different solutions and initiatives if I recall, and correct me if I'm wrong, I think some of the past messaging has been you guys are really focused on just driving usage and adoption. But where -- even if that might be the goal today, when you look 2, 3 years from now, how do you see that playing out and impacting revenue potentially?

So there's a couple of different areas. I'd say the agents that we're very much in the platform people like to say it's free. I said they don't use the effort. The reality is that we have great monetization associated with usage and more of that increases usage, we'll monetize that and the more of that increases the adoption of the platform, we'll monetize that.

But there are also new products that we will charge for. There are new modules. AI feedback is one we're charging for, and we will continue to charge for. You saw AI visibility, right? And that is in the agentic search optimization area, that's going to lead to some products too because what we did, we gave it away for free, but it was one that just kind of tested and poked at the world that is like the profound of the world saying that that's not a new product, that's a feature.

And what you really need to do is not get the insight, you need to understand what you do with the insight, how you optimize from that insight. So we will definitely sell products in that space and be increasingly focused on developing our own capabilities to displace Google and Adobe.

Perfect. And on the topic of AI, I got to ask you about OpenAI's acquisition of Statsig, which has some overlap on the product analytics side, but they're more of an experimentation player from our understanding. But I think when OpenAI made that acquisition, there were some investor questions around, does that mean that OpenAI might further get into Amplitude's core market, right? So what are your guys' thoughts on OpenAI's acquisition, what that means for the direction they're potentially taking, do you view this as validating, right, the opportunity you guys are chasing? Do you see them as a potential rival in the future? How do you think about that?

So our initial response was very much, oh, it's validation, much like when Datadog acquired Eppo, experimentation is an important aspect for anybody who's trying to figure out how they bank their digital engagement better for clients.

What I can tell you is we saw Statsig once in a while in very scientific-related experimentation opportunities and data warehouse native opportunities, but it was never really that large. It was mainly in smaller implementations. But when it was acquired, that's what we thought. And then we got more and more feedback from people who are giving us insights, and they told us, well, half of that Statsig's ARR was from OpenAI. Oh, and 2 CEOs, they were at Meta at the same time. And they brought in and wanted him to be the CEO of Statsig, wanted him to be a Head of product. And suddenly it looks like an acquihire, not really a product continuation.

And I would tell you, we don't really see them at all anymore. Now that doesn't mean that they won't do something, but I suspect that this was purely an in-sourcing and talent acquisition for them. And I would tell you that increasingly, we're seeing more and more AI native-related companies use Amplitude, you know that Cursor does. You know that, Granola does, you know that Character.AI does and might be another one of those major LLM model providers out there that's using Amplitude.

Yes, those definitely are great, great proof points. Let's look on the P&L now and shift gears a little bit. So net new ARR growth has been a blowout metric. So 2 quarters ago, that grew 200%. Last quarter, it was still up an amazing 45% year-over-year, tends to be a little bit lighter of a bookings quarter seasonally for you guys. But I guess just as we look ahead for that metric, we're going to start to come up against some tougher compares. So is that something we need to be mindful of in our models? Do you guys still feel like there's tons of momentum and runway left that some of the levels that we've seen can continue to be more durable? I guess how should we just think about the trajectory potentially of that metric as we look ahead?

Yes. I think there's a couple of things that I would mention. So first and foremost, I talked about that we are kind of instituting a better and better sales productivity framework for our teams. And so they've always been good on the growth side, but we've had that the headwind on the churn issue. Well, if we're doing better on the churn, variably you get better on the net aspect.

So I think at the current capacity levels, we can continue to drive. And we've got -- even without the AI products, I think we still have a great one way on just driving consolidation in the marketplace. We're seeing great demand from both traditional and nontraditional customer bases.

So I think we're in good shape there. But I think longer term, I think you're right, it gets harder from a compare perspective on a dollar basis. And that's when we start pulling the lever on capacity. We start adding more. If we're executing well on an efficiency perspective, that's when you want to invest in your go-to-market team and start building the capabilities to drive that growth harder.

Perfect. And maybe in the last 30 seconds or so that we have. When we think about the balancing of growth and profitability going forward, I'd love to get your thoughts there. Obviously, you guys have been executing really well on the improving growth story. When we think about profitability and those 2 in combination, do you see a path getting back to a metric like Rule of 40? How are you thinking about that? What are going to be the key unlocks? Is it a function of you've been doing all the right things. We just need to see revenue growth accelerate and that will unlock a lot on the margin side? Is there still areas of cost efficiency that you're looking at closely. Maybe you could just give the group -- what your high-level thoughts are there?

I would say it's never one of those binary things, and you're always trying to balance it out. I hear there's a big opportunity for technologies out there for us to bring in at a very low cost. And so occasionally, we'll look at making those growth investments. But I will tell you that the rest of the company just saw our approved budget at the Board level for next year, and it's very much in line with growth with leverage focus.

And you're always making trade-offs, but I'll give you an example. The product team came up with a long list of things you wanted to go do. And the reality is I told them, here's the line that we can afford. So figure out with sales, what are the most important things we work on.

And I think that's what investors are expecting. They're expecting us to make those hard decisions and drive towards a more and more profitable business. I will tell you that we think about growth and free cash flow margin that we're mapping out a path for Rule of 40.

Perfect. Awesome. Well, we'll end it there. Thank you, everyone, for joining. And Andrew, I appreciate all the time. I learned a lot. So thanks for joining us.

Thank you.

[Technical Difficulty] everyone, and welcome to Amplitude's Third Quarter 2025 Earnings Conference Call. I'm John Streppa, Head of Investor Relations. And joining me today are Spenser Skates, CEO and Co-Founder of Amplitude; and Andrew Casey, Chief Financial Officer.

During today's call, management will make forward-looking statements, including statements regarding our financial outlook for the fourth quarter and full year 2025, the expected performance of our products, our expected quarterly and long-term growth, investments and our overall future prospects. These forward-looking statements are based on current information, assumptions and expectations and are subject to risks and uncertainties, some of which are beyond our control, that could cause actual results to differ materially from those described in these statements. Further information on the risks that could cause actual results to differ is included in our filings with the Securities and Exchange Commission. You are cautioned not to place undue reliance on these forward-looking statements, and we assume no obligation to update these statements after today's call, except as required by law.

Certain financial measures used on today's call are expressed on a non-GAAP basis. We use these non-GAAP financial measures internally to facilitate analysis of our financial and business trends and for internal planning and forecasting purposes. These non-GAAP financial measures have limitations and should not be used in isolation from or as a substitute for financial information prepared in accordance with GAAP. Additional information regarding these non-GAAP financial measures and a reconciliation between these GAAP and non-GAAP financial measures are included in our earnings press release and the supplemental financial information, which can be found on our Investor Relations website at investors.amplitude.com.

With that, I'll hand the call over to Spenser.

Good afternoon, everyone, and welcome to Amplitude's Third Quarter 2025 Earnings Call. Today, I'm going to cover 3 things: first, our strong Q3 results and progress in the enterprise; second, the AI opportunity within analytics; third, product innovation and our customers.

Let's go ahead and get started with our Q3 results. We delivered another strong quarter, continuing the acceleration we saw in Q2. We exceeded expectations on our core financial metrics and made solid progress against our enterprise strategy. Our third quarter revenue was $88.6 million, up 18% year-over-year and exceeding the high end of our guidance. Annual recurring revenue was $347 million, up 16% year-over-year and up $12 million from last quarter. Non-GAAP operating income was $0.6 million. Customers with more than $100,000 in ARR grew to 653, an increase of 15% year-over-year.

Our Q3 performance reflects our continued execution against our strategy. We are winning simple by bringing Amplitude to everyone with AI. We are winning the enterprise with broad-based success with both AI natives and traditional enterprises, securing larger multi-year contracts. And we're winning the category with multi-product adoption now representing 71% of our ARR. Finally, we're winning together by leading the shift to being AI native across the entire Amplitude team.

I wanted to take a little bit of time to talk about how AI is changing how software gets built. Every single product team runs the same loop: build, ship, use and learn. The left side of that loop, build and ship, has been transformed by AI. AI coding has made it faster than ever to turn ideas into products. We are now at a point where someone can create a product that's used by millions overnight. By contrast, the right side of this loop, use and learn, remains in the stone ages. Companies ask users what they want, but actions are more powerful than words. The best way to understand what people want is to watch what they do. This is what Amplitude solves. Our AI analytics platform helps companies understand how people engage in their product, what they like, where they get stuck and what keeps them coming back. These behavioral signals are the most powerful indicator for what to build. Automating and scaling that understanding is the next frontier.

In this context, analytics is the perfect problem for AI to solve. To use analytics, you have to do a lot of manual work in specifying and setting up your query in between rounds of thinking. AI can handle all of that manual work, freeing up humans to focus on the thinking that leads to great insights. Amplitude has a unique position to build the AI analytics platform of the future. We have the world's largest database of product behavior. We have spent a decade working with world-class analytics teams. Over the past year, we've rebuilt the Amplitude team to be AI native. We've reorganized product development twice, and we've acquired 4 AI companies. We've trained our engineering, product management and design teams deeply in AI. The company that disrupts the right side of this loop, the use and learn, the fastest will define the future of this space. We are all in here.

Let's get into the details on product innovation. In the last few weeks, we have launched several AI-native products here at Amplitude. I want to start with MCP. In October, we announced the public availability of our MCP server, the top requested feature from our customers. MCP is made for data analytics. It exposes all of Amplitude's functionality, so an AI agent can interact with it directly. It allows you to use Amplitude without knowing anything about the Amplitude UI or your data taxonomy. This is the best MCP use case that I have ever seen. Watching an AI agent think, reason, query Amplitude and then repeat that process iteratively is magical. It shows what is possible in the AI analytics future I talked about earlier. It brings the power of our data to anyone in any workflow. This opens Amplitude up to an entirely new cohort of nontechnical users, in turn driving our growth.

Let me show you with a quick demo. MCP lets AI tools interact directly with Amplitude data. You can ask a vague question about your product inside any AI model and have a query Amplitude iteratively. Our MCP already has native connections to Claude, Cursor and GitHub with more to come soon. For this demo, I'm going to use Claude. I'm going to start with a simple prompt, give me high-level web traffic metrics. Claude will then get contacts, search web traffic metrics and then it's identified that during the week of September 21, we've had a peak. I can then ask follow-up questions to drill down, investigate the September spike. What's driving this growth? Claude then accesses behavioral insights, checks traffic sources, marketing campaigns and content performance. It shows that our webinar campaigns and the release of our product benchmark report drove this traffic. To get deeper insights, I'm going to ask, what are the downstream growth metric impacts by these campaigns? Claude then queries the campaign data set, downstream conversion funnels and sales force metrics. It concludes that the September campaigns drove a higher number and quality of visitors to the site. Of course, I'm going to want to share these findings so I can prompt it to create an Amplitude notebook for the growth team. So in a few minutes, customers can get deep research, insights and a detailed shareable notebook that allows them to take action.

In addition to the launch of MCP, we expanded the open beta for our AI agents. These agents continually monitor product data, detect anomalies and surface insights automatically. In June, we launched our closed beta. And then 2 weeks ago, we opened the beta to all customers. Our focus is now on 2 agents. The first is the Dashboard Agent, which analyzes charts and proactively flag significant changes. And the second is the Session Replay Agent, which reviews thousands of user sessions, detects points of friction and then shows curated clips that highlight issues. Both are powered by the same behavioral data that MCP can access. They are already helping customers uncover opportunities and resolve issues faster.

In addition, last week, we also introduced AI Visibility. As consumers turn to AI tools like ChatGPT, Claude and Google's AI summary when they search, marketers are flying blind. They have no idea how their companies show up or rank within the results produced by these new tools. To solve that problem, we launched AI Visibility for free last week. Think of it as SEO for LLMs. It shows where a brand appears or doesn't across all major AI models, how they rank against competitors and how they can improve their position. We saw a lot of excitement around the launch with our customers and on social media. The conversation about the future of AI Visibility is still going today on Twitter. Let me show you a quick demo of this, too.

Understanding how your products appear in AI responses and improving it will be the key to increasing awareness. With AI Visibility, customers can now see the percentage of mentions of their product and AI responses. They can also see competitor mentions versus their own and then topics by visibility. They can dig into prompts to see the exact questions customers are asking and how AI is answering. For example, when people ask LLMs for product-led growth tools, they mention Amplitude 90% of the time. Customers can also learn how to improve their ranking. I can use the Analyze page to see how AI interprets the existing content or run a series of simulated changes to test updates before publishing. AI Visibility tracks your brand and helps you turn that visibility into growth.

Finally, next week, we will launch AI Feedback. This is our newest AI-native product based on the core offering from our Kraftful acquisition in July. We're going from acquisition to new Amplitude product launch in 4 months. AI Feedback takes user feedback and information from multiple sources and turns it into insights a company can use. By bringing feedback, behavior and action into a single platform, it helps teams hear customers and understand them. Let me show you with my last demo.

AI Feedback is the new way to listen to users at scale and act on their feedback. AI Feedback collects input from all of our customers' feedback sources. You can link Zendesk tickets, Gong call transcriptions, App Store reviews, comments on Reddit and other social media, first-party surveys and more with no engineering help. In this example, I have already set up AI feedback for a mobile app and connected it to reviews from the Apple App Store, as well as Google Play. Agentic AI processes the massive volume of unstructured data and sorts it into categories like you see here. Our proprietary AI analysis gives product teams the right level of detail. Users can see feature requests to help build a road map or can filter by topics like complaints to know which issues to address. Here, there are 26 mentions of reliable read, start and alert query. I can click in to see specific comments for more detailed information and subtopics. For example, there are 4 mentions and notifications not syncing across devices. With Amplitude, customers can then turn that feedback into action. We can create a cohort of these 96 users watching session replays of how they interact with notifications or surveying them for more on what they need.

Our innovation will accelerate from here. In addition to what I've just shown, over the next few quarters, we'll introduce new AI-first products like automated insights, global chat assistant and additional agents that extend our reach. These new capabilities expand who can use Amplitude and strengthen the value of the analytics platform overall. Every new product draws on the same behavioral data set and feeds back into it, creating a single system of improvement. That's what makes our use and learn opportunity so large. Innovation is the biggest driver of long-term growth at Amplitude and our strongest moat in this AI-first world.

Let's talk about customers. We had a great quarter for new and expansion deals with enterprise customers, including Bentley Systems, FanDuel, Thomson Reuters, Taco Bell, Global Radio, Empower, Granola, Algolia and Gusto, among others. I'm going to highlight how a few of them are putting this all to work. Three examples stand out this quarter, each showing the power of the Amplitude platform in a different way.

First is FanDuel. They continue to be a great example of platform consolidation at scale. FanDuel is constantly refining its experiencing -- its experiences and tailoring them for millions of fans. Connecting real-time feedback to customer experience is critical to their success, and Amplitude powers that loop. FanDuel uses Amplitude end-to-end across multiple product lines from analytics to guides and surveys to session replay. That unified view helps their teams test, learn and improve faster. Their expansion and renewal patterns remain among the strongest in our base.

Second, Granola. This fast-growing AI start-up adopted Amplitude before launch after hearing about us through the AI ecosystem. Today, more than half the company uses Amplitude every day to understand how people use their product and where to iterate next. Granola ships new features quickly and relies on real-time feedback to guide product decisions. Their story is a great example of how AI-native companies are choosing Amplitude to accelerate growth and scale with confidence.

Third is Bentley Systems, a global leader in design, construction and infrastructure software. Bentley selected Amplitude as a single analytics platform across all products. The company previously relied on siloed legacy tools but needed one system to understand usage, drive adoption and guide feature development. By activating historical data from Databricks and combining it with behavioral insights in Amplitude, Bentley can now test, learn and implement improvements far more quickly across its portfolio.

These stories all point to a common theme. From AI start-ups to global enterprises, customers are betting on Amplitude as the AI analytics platform that will help them thrive in this new era.

We are at the beginning of redefining analytics as an AI-native system that learns, reasons and acts. Over the next few quarters, we will bring a new wave of AI-native products to market that will reshape how companies use data to build better products. This is just the beginning of the AI era for analytics.

I'll now hand it over to Andrew to walk through the financials.

Thank you, Spenser, and good afternoon, everyone. We've delivered another solid quarter of acceleration in our ARR, improved operational efficiency and created greater durability in our future revenue base. Our customers are increasing their pace of innovation and in turn, need to understand how the changes are being received, how to adapt and how to implement those changes. As such, we believe Amplitude's importance to the enterprise is increasing.

On our business, we have continued to perform against our strategy that we communicated at our Investor Day earlier this year. We've increased the value that our platform can deliver by adding Guides and Surveys, AI agents, our MCP server and others. We've grown the base of our enterprises we are serving, the number of customers that are using multiple products and the full platform. We've done all this while improving operational efficiency of the business.

We continue to improve the durability of our business as measured by improving our contract duration and remaining performance obligations, or RPO. This quarter, our average contract duration grew to nearly 22 months, up from 19 months just 1 year ago. Our RPO growth has improved from Q2 with current RPO growth year-over-year accelerating to 22% from 20% last quarter and long-term RPO growth accelerating to 78% year-over-year, up from 64% last quarter. This results in total RPO growth of 37%, accelerating from 31% last quarter. The growth in our RPO is a direct result from building a more repeatable and scalable go-to-market strategy, focused on enterprise customers.

Turning to our third quarter results. As a reminder, all financial results that I'll be discussing with the exception of revenue are non-GAAP. Our GAAP financial results, along with a reconciliation between GAAP and non-GAAP results, can be found in our earnings press release and supplemental financials on the Investor Relations page on our website.

Third quarter revenue was $88.6 million, up 18% year-over-year and 6% quarter-over-quarter. This quarter's growth benefited from better linearity in deal signings earlier in the quarter, growth in our services business and the strong net new ARR growth we had in Q2. Total ARR increased to $347 million exiting the third quarter, an increase of 16% year-over-year and $12 million sequentially.

Here are more details on the key elements of the quarter. We had a strong quarter for both new and expansion deals in the enterprise. Platform sales were also particularly strong. 39% of our customers now have multiple products with 71% of our ARR coming from that cohort. The number of customers representing $100,000 or more of ARR in Q3 grew to 653, an increase of 15% year-over-year and up 19 customers since last quarter. In-period NRR progressed to 104%, led by cross-sell expansions. Gross margin was 76% for the third quarter, down 1 point from the third quarter of 2024 but up 1 point since last quarter. We continue to make progress on optimizing our hosting costs, driving multi-product contracts and monetizing services engagements. And we will continue to look for opportunities to incrementally improve our gross margin over time.

Sales and marketing expenses were 43% of revenue, a decrease of 1 point from the second quarter. We continue to focus on improving sales efficiencies, driving improvement through our changes in process, coverage and expansion of enterprise customers. At the same time, we are investing for future growth, while balancing this incremental investment with efficiency gains. G&A was 13% of revenue, down 3 points from the third quarter of 2024. We expect G&A to improve as a percentage of revenue over time. R&D was 19% of revenue, up 3 points from the third quarter of 2024. We expect to continue to invest in the talent and the capabilities of our team to drive greater innovation in the future. Total operating expenses were $67 million or 75% of revenue, down 1 point sequentially.

Operating income was $0.6 million or 0.6% of revenue. Net income per share was $0.02 based on 143.2 million diluted shares compared to net income per share of $0.03 with 131.3 million diluted shares a year ago.

Free cash flow in the quarter was $3.4 million or 4% of revenue compared to $4.5 million or 6% of revenue during the same period last year. In the third quarter, we managed our cash collections and made meaningful progress, shifting to contracts with annual payments in advance.

Now, turning to our outlook. As Spenser laid out, the world of development, test and ship is changing rapidly. Analytics and the use of data to understand outcomes and drive action will be more important than ever for enterprises. Our strategy remains consistent with our go-to-market. We will continue to focus on gaining new enterprise customers and driving cross-platform sales with our existing customer base. We also believe that with the release of our AI capabilities, the monetization of data ingested into our platform and cross-sell opportunities of new products gives us the right strategy to align the value of our customers receive with our growth opportunities and to grow our business in a profitable way.

For the fourth quarter of 2025, we expect revenue to be between $89 million and $91 million, representing an annual growth rate of 15% at the midpoint. We expect non-GAAP operating income to be between $3.5 million and $5.5 million. And we expect non-GAAP net income per share to be between $0.04 and $0.05, assuming diluted weighted average shares outstanding of approximately 142.6 million.

For the full year 2025, we are raising our revenue expectation for the full year due to the quarter's positive performance. We expect full year revenue to be between $340.8 million and $342.8 million, an annual growth rate of 14% at the midpoint. We are adjusting our range for the full year non-GAAP operating income to be between $0.5 million and $2.5 million, reflecting growth investments. We expect non-GAAP net income per share to be between $0.06 and $0.08, assuming weighted average shares outstanding of approximately 142 million as measured on a fully diluted basis.

In closing, we continue to execute our strategy of growing our enterprise customer base, expanding multi-product attach with our customers and growing with additional leverage in our business model. This has only occurred through the focused execution of our employees and our relentless drive towards creating value for our customers.

With that, we'll open it up for Q&A. Over to you, John.

Thank you, Andrew. We will now turn to Q&A. [Operator Instructions] Our first question will come from the line of Koji Ikeda from Bank of America, followed by Patrick Schulz.

I wanted to ask on RPO. I look at ARR, nice growth there, but RPO, real nice acceleration of 37%, most added sequentially in a long time. Andrew, totally hear you on the repeatable and scalable enterprise aspect of the execution here. But I was hoping you could break it down a little bit by enterprise, mid-market, vertical, maybe even annual contract size and contract duration. What's really driving that 37% growth there?

You may recall, Koji, that back in the beginning of the year, we made a lot of efforts at resegmenting our customers, refocusing our sales coverage on the enterprise group of clients. And we even have a stratification that's even larger than that, we call, [ Strat ] in the enterprise. And for those who don't remember, we define enterprise clients as anyone who has 1,000 employees or more or over $100 million in revenue. And when we did that, we really had a lot of marketing efforts, focused efforts to show how we are consolidating the space, how we could increasingly -- by customers, increasingly leverage the power of the Amplitude platform, how they get greater value. And I would tell you, when you have those conversations with clients, they often look to not just a single tool replacement, but rather a multi-year journey, how they're going to continue to drive value as their business changes, as their business grows. And that's led to more strategic conversations and thus the deal constructs, which have more duration to them. I mentioned earlier, our contract duration has increased overall to 22 months. I will tell you, in the enterprise space, it's even larger. And the reason is because so many of those multi-year -- million-dollar contracts result in multi-year commitments, and that's driving our RPO.

Now, you don't do that without getting the sales teams very focused in an execution way and aligning incentives [ proportionately ] to drive those outcomes. And I would tell you, they've done a great job in multi-year contracts and making sure that our RPO is growing quite rapidly.

And maybe a question here for Spenser. In the presentation, you talked about this, I'll call it, the build, ship, use and learn circle of life here. Where are enterprises today...

Product circle of life.

Product circle of life. Yes. Where are enterprises today? Where are they investing now? And where are they really going to be investing in the future?

You mean, in product or in analytics?

In that circle. Are they investing more in the build process, the ship process? Are they yet to really start to invest in the use and learn side, and that's the real opportunity for you guys?

I think it's much earlier on the use and learn side, right? If you look at the state-of-the-art for how companies do this, it's literally you go ask, talk to a bunch of customers, you do a focus group, you send out a survey, you're getting lots of qualitative feedback. The point I am trying to make is that if you can master that quantitatively, that is so much more powerful because you can -- actions are much more powerful than words. Users are great at describing the problem they have. But it's like that old great Henry Ford quote, if you just listen directly to what they do, they'd ask you for faster horses. And so, the companies like the Facebooks or the Netflixes of the world that have done a great job of that has this incredible edge over anyone else. And so, we see ourselves as bringing that infrastructure much broadly.

Now, to your question, it's much, much earlier there. I think we see a lot of excitement. Like I was just in Europe a few weeks ago, and every single customer wants to talk about, yes, how is my learning side going to get completely disrupted by AI. Like we're already on coding. They're having the developers use Cursor and Claude Code and a whole bunch of other coding -- AI coding tools. They want to know how the same is going to happen in analytics. And so, what I was presenting is like, hey, what's our vision for that future?

Our next question will come from Patrick Schulz from R.W. Baird, followed by Elizabeth Porter from Morgan Stanley.

Maybe first, just thinking about the growth for revenue and ARR, another very strong quarter. And looking at the Q4 guidance, there's some revenue deceleration implied there. Just wondering if you could provide a little bit more color on what went into that forecast. And maybe more broadly, there's always some trade-off between investing for growth and driving profits. But curious to hear how you guys are balancing these 2 as you head into next year, especially just given the velocity of innovation you guys are seeing.

So I'd tell you that our guidance is always based upon the lens of execution for us. And as we've gotten better and better at executing our enterprise play, that's given us a lot more visibility into future revenue, and it shows up in our RPO. And so, rather than go back on what has worked very well for us as we progressed this year, we're going to stick with that guidance pattern and stick with what we know we can go execute in the market, and we believe that will continue to deliver benefits for us.

Now, on balancing growth with leverage, if you were a fly on the wall within the Amplitude offices, you'd find that we're often making debates -- we have debates about when to make the right investment, how much leverage can we expect from it, and then how can we make sure that we are continuing to drive towards the expectations we set at Investor Day back in March. And so, it's always a little bit of a difficult conversation. But I think the teams have learned to manage within those constructs, and we're making the investments that really matter while still delivering progressions on our growth with leverage plan.

Yes. I mean, Patrick, I'll say you don't find -- the bottleneck to growth is actually not a lot of time spend. It's about how can I retrain this team to be more effective. So Andrew and I and with Thomas have been looking at sales productivity and how do we drive that for next year. The teams have really stepped up and done a good job of how do we be really effective. You can get a lot of leverage from AI tooling in this world without a crazy amount of spend.

Okay. That's very helpful. Maybe just one quick follow-up, too. Very impressive list of customer wins during the quarter. And one of your goals over the past year has been to improve the enterprise go-to-market motion. Just can you maybe talk about how some of these improvements have impacted the trends with initial deal wins? What are you seeing in terms of change to sales cycles, size of initial lands and maybe landing with multiple products?

Yes. I mean, we're -- I mean, as I kind of outlined in all 3 examples, I mean, I think we're doing everything across the board way better than we were a year or 2 ago. You're seeing many more multi-years, which is reflected in the RPO growth. You're seeing 71% on multiple products. That's a huge [ freaking ] deal. You're seeing customers willing to just invest in analytics, invest in the whole platform. They're very excited about what we're doing on the AI side as well. And so, I got to give a lot of credit to the go-to-market teams for how they've built much stronger relationships with those. Even though I didn't speak to it too much, they've done a phenomenal job, and that's why you see ARR reacceleration, as well as all those other numbers.

One other just data point to call out, we had a really broad-based set of 7-figure wins. It was about -- it was 5 7-figure wins in Q3, which was just fantastic to see, everything from some of the enterprises, the very traditional enterprises I called out, all the way to -- we actually had a foundational model company that we can't disclose who it is but sign for Amplitude. So it's just exciting to see kind of, hey, across the full spectrum, we're doing well.

Our next question will come from Elizabeth Porter from Morgan Stanley, followed by Willow Miller from William Blair.

Awesome. I think one of the really exciting parts about Agents is its ability to tie together multiple products across the Amplitude platform. And should we expect kind of a step function increase in multi-product adoption? I know that it was about 70% of ARR, but the customer base is still below 70%, which is -- 40%, which is a big opportunity. So how does the Agent strategy just overall accelerate that mix shift in the broader platform strategy?

Okay. So there's what we're doing today and then there's future. What we're doing today, we are focused on the analytics adoption piece because more people using analytics, the more people sending us data, the more people querying that data, the more value we create, and that will naturally expand and lead to a whole bunch of these other use cases. So the 2 agents I talked about, the dashboard monitoring agent as well as the session replay agent, those are all kind of insights-level agents. Now, what we do have coming and will be in next year is we're going to be doing a lot more on the action side, so suggested experiments, as we showed in June on the last earnings call, where we come up with a strategy for how you can improve conversion on your website, suggested guides for you to send out to users, hey, you don't have a new user onboarding guide. We've created one for you. Do you want to go ahead and deploy that? Suggested cohorts of users to target with messaging. So I'd say today, on the Agents front, we're specifically focusing on the analytics and insight layer because that's where we're hearing the most pull from customers, but we'll add a whole bunch of action layer stuff as a fast follow-up.

Great. And then, just as a follow-up more on the financial side. As Amplitude moves further upmarket, should we expect there to be any seasonality in in-quarter NRR, just given we can have some of these quarters that are heavy in the enterprise deals? And how would you frame kind of the cadence in NRR trend and how it could evolve through the year?

So I'd say that everything is better with -- in the enterprise. Our NRR is much higher than the overall average for the business. We still have cohorts in the SMB and mid-market that have higher churn rates. But as we get more -- a higher and higher percentage of our ARR in the enterprise, NRR will continue to progress. And we expect that we'll see a continued progression in our enterprise content. As we delineated at our Investor Day back in March, we'll see that progression continue to go up. That drives better gross retention as well. And so, although we didn't see a big progression this quarter, the reality is, we have a lot of faith in our long-term expectations of being that 115% that I talked about last quarter.

Our next question will come from Willow Miller from William Blair, followed by Ian Black from Needham.

I'm Willow on for Arjun Bhatia. I believe, in your prepared remarks, you mentioned that you believe the new AI functionality like MCP and Agents can open up Amplitude to nontechnical users. Can you comment more on this? And could this expand your market opportunity?

Absolutely. I mean, the hard part -- there's 2 hard parts in analytics today. One is, you have to learn an analytics product. And so, as easy as we made it with Amplitude, you still have to learn our interface, what drop-downs, what type of charts to use, all of that sort of stuff. The other thing you have to learn, and this is the biggest bottleneck in analytics, is the data taxonomy that you have. And the problem with particularly product data taxonomies is they are huge. Like the average product has thousands of different things you can do in it. So, no one is going to be able to keep that in their head. The huge leverage point you get, which I showed in the MCP demo, is that you can start with a very vague question, like show me my onboarding funnel, and then it will look across your data set to see, okay, which are likely the onboarding events, let me construct a funnel with them. And then, if you're not happy with the result, you can iterate with it. But the point is, you're not having to learn our interface or any analytics interface. You're also not really having to learn your data taxonomy. It's leveraging -- it's doing the thinking and reasoning for you on specifying the query and constructing it and putting that back in a chart. So you're just getting the result and then kind of going back and forth with it. So I -- so the answer is, highly yes. We've set really aggressive internal targets for next year for what we expect the -- how we expect the adoption of Amplitude to grow because of what we're building.

Our next question will be by Ian Black from Needham, followed by Claire Gerdes from UBS.

Congrats on the great quarter. As you guys start to lap the switch to multi-year contracts that you started last year, is there an impact on sales productivity?

No, I don't think that -- we certainly look at sales productivity both on a gross and a net basis. But I think that the opportunity for us to continue to drive expansions within our existing customer base is well over $150 million. We said something similar at our Investor Day when we had even fewer customers. We've done a pretty good job of adding new enterprise clients. And even at the rate that we're adding them, we still think that there's a large footprint within each one of them where we can show great expansion opportunities.

Awesome. And then, you mentioned customers increasingly viewing you as key strategic partners. Is there an opportunity to expand your service partner network as you go deeper within your customer base?

Yes. That is one of the other growth engines that is somewhat nascent for us today. I think that as Amplitude becomes increasingly important to enterprise clients, you'll start to see more and more partners look at ways in which they can develop on our platform and create their own value add. And that's something that, over time, we believe will happen.

Our next question comes from Claire Gerdes from UBS, followed by Jackson Ader from KeyBanc.

I'm on for Taylor McGinnis. Great to see the results today. I wanted to ask about some of the newer AI products that you're offering. I know some are still in beta. But as we think about where you are with those and balancing adoption and monetization in the future, I know, again, still very early, but is it more that you're wanting customers to adopt and get familiar and we can expect more of those free at this time? Or how are you thinking about that right now?

Yes. Adoption is our key focus. Like they drive a bunch of value. Customers will track more data. They'll use that data a lot more broadly across more users, and that will very naturally lead to more value and upsells and everything else. I mean, if you look at our ASP, we're hovering around $400,000 a year. In a lot of companies, we are the largest spend after their cloud hosting. So we have no problems already commanding a premium price point. The key is to be able to consistently deliver value and make that barrier lower and lower, which is what we're doing with MCP and Agents and all of the stuff that's targeted at nontechnical users.

And then, just maybe a quick follow-up, a bit broader, but can you just comment on the selling environment right now? Obviously, still a lot going on, but you're putting up a great acceleration. So just if you could comment on that, that would be great.

Yes. I mean, I think there's been -- I don't think there's any change quarter-to-quarter. There's a continued cost consciousness among all customers, the same way it's been for the last few years. I think what they're really excited to see is like, hey, does this AI stuff work? Is it legit? Every B2B company under the sun is like pitching them on AI this, that or the other thing. And so, what I always say is show me the demo. So that's why I kind of did that on today's call. And they're really, really -- yes, a lot of customers are very, very excited about that. They want to know how can I get 5, 10, even more times leverage with leveraging AI Agents, MCP, get a lot more insights than I would doing all of that work manually.

Our next question comes from Jackson Ader at KeyBanc, followed by [indiscernible] from Citi.

The first one is kind of a product question. One of Francois' major initiatives, I guess, if you want to call it that, is a tighter integration between customer feedback and the product road map and kind of getting that flywheel going. So I'm curious, what are customers -- agents are rolled out. You've got AI features and functionalities. What are customers kind of looking for as we head into 2026 that might be able to like continue to compound the growth that we've seen?

Yes. I mean, look, just to be really clear, we're in the really early days of Amplitude in this category. I think -- let's see what would I call out. So first, take all the AI stuff and set it to the side. Just with the enterprise execution, plus the expanding breadth of the platform, that already, we expect to continue to accelerate us to the ranges that we talked about on the Investor Day earlier this year, so beyond the 20%, which, in my mind, is kind of the bare minimum for this category. And then, you add how this category gets dramatically reshaped by AI, as I've been outlining, like the same thing that happened to software engineering with Cursor and with a whole bunch of other AI coding products, that is going to happen in analytics, and we see what that opportunity is. We're going to go really aggressive after it, and there is a lot of appetite to adopt that. So I'm very excited and bullish on it. And I think as we see customers adopt and use and get value on that, that will all translate downstream to Amplitude revenue growth.

Okay. All right. Great. And then, the follow-up question is just on the split between -- or I guess, not split, but just product analytics versus marketing analytics. I know we're getting into kind of new budget territories with the marketing piece. At what point does that split become material enough to where [ I'm sure ] that the high growth rates coming from marketing analytics is able to contribute to like a material amount of your overall revenue? Or are we there already?

No, no. It's early on it. I mean, there's customers that have switched off of legacy marketing analytics products and moved wholesale to Amplitude. Like DECATHLON is one. [ Realtor ] is another. There's a few that have that -- that do that, that are like saying, "Hey, Google Analytics or Adobe is not working well enough. So let me move on to something like an Amplitude." Now, there's still more for us to build there, and we're going to be -- we haven't really talked about what we're going to be doing there, but there's -- yes, we're going to be doing a lot in 2026 that will allow us to capture a significant amount of the marketing budget and have that contribute to growth.

Our next question will come from [indiscernible] from Citi, followed by Clark Wright from D.A. Davidson.

Spenser, the AI Agents launch over the summer was like pretty big [ fun for OSF ]. So today, with better availability, could you kind of give us an update how has customer adoption been? What are some of the most effective use cases you are seeing and the ROI from customers? And then, also if there's any update on pricing monetization effort for AI Agents as it goes better?

Yes. So let me hit the pricing one first. So, as I said to Claire earlier, focus is just let's make Amplitude more valuable. Like we command already an extraordinarily high price point with a lot of our customers. And so, we're not looking to squeeze them for more. We just want to give them a lot of value. So I think it's a common misconception -- sorry, misunderstanding about our business relative to other SaaS is, we already command like very, very high price points, which is great, but you obviously want to make sure you're consistent about delivering the value.

In terms of what use cases work very well, so the 2 I highlighted, dashboard monitoring and session replay are the big ones that customers use a lot of. They -- I think customers are not ready to trust AI agents to come up with a whole strategy of wholesale for them and then recommend a series of actions. But they are very excited about using things that can monitor the data and just proactively tell them -- like this is, in a lot of ways, the holy grail of analytics, like can you tell me when something has changed in my data and deliver that insight as to why it changed? And so, in the MCP demo, we have a -- I didn't do the demo today, but we have a dashboard monitoring agent that can do the exact same thing at that, deliver kind of a root cause analysis saying, "Hey, you had a spike in September and here are the particular causes the product benchmark report and this other launch that you did. And that customers are using and getting a bunch of value, and that's what our focus is in the short term. And then, same with session replay. With session replay, you'll have often millions of sessions, and it's like you can't watch them all. So how can you have AI summarize the highlights for you? And then, you can maybe watch 3 or 4 that are relevant you're like, "Oh, okay, I can see how the product experience can improve if we make these changes to the app."

Got it. That's clear. And Andrew, maybe one for you. I think the guide implies like a stronger kind of momentum going into Q4. Can you kind of help us understand the dynamic that you're seeing that have made you raise the guide bigger than [ a bit ] in the quarter and kind of how much conservatism you see in the guide and how is the pipeline going to year-end?

Well, as we moved our business more towards pursuing enterprise opportunities, you still -- you have a pattern like many enterprise companies, and we're a calendar year company. So, all of our reps have end-of-year incentives to achieve and exceed their targets. And so, we typically have our strongest quarter in Q4. So that's the first thing. Pipelines and the maturity of those pipelines are something we look at very closely when we're putting together our guidance. What we've seen so far, that gives us confidence. And it would be remiss of me if I didn't say the RPO gives us great visibility into future revenue periods. And the more that the sales team executes on driving greater relationships with clients, more strategic agreements, we're getting those multi-year agreements booked. And so, as I said earlier, the guidance is based on the lens of execution for us, and we've pursued that same philosophy throughout this year, and it's led to strong performances. So we're going to continue that.

Congrats on the great result.

And our last question comes from Clark Wright from D.A. Davidson.

Spenser, your product teams have been hard at work this quarter.

Every quarter. But yes, we're accelerating.

Yes. And the pace of innovation has been great to see. I was wondering if you could talk about how you think right now about building versus buying incremental capabilities at this point in time in order to keep up this pace of innovation.

Yes. So you always want to be able to do both. Like I think experiment -- we actually looked for a while for the right team to come in but ended up deciding to be able to build was the faster way to get that to market because there wasn't a company at the time that was the right fit. Conversely, like we bought when we think it makes sense, so Activation back in 2022, Guides and Surveys, now AI Feedback, that made a ton of sense. So you're kind of always doing both. I think the important part is we're not religious. And so, whatever gives us the shortest path to market, like with AI Feedback, that would have taken us a year -- I mean, I'd like to say only a year, but the reality is probably more like 2 to get to the same point a state-of-the-art product like Kraftful was. And the same with Command on Guides and Surveys, that would have taken us 2 to 3 years, no question, at least to build internally. So it's just a much faster path to market. So we're always looking at both.

The other thing I want to point out is that I think -- I mean, I've been really excited to work with the founders that we have. Yana on -- from Kraftful has done a phenomenal job of kind of educating me about AI and educating a lot of the rest of the company. Same with Enzo and Ferruccio from June, same from Frank and Eric from Inari, like they've all done a phenomenal job of like they've been building in this space in an AI-native way for a bunch of years. And so, as we're looking to change the makeup of the team to be able to build -- to look much more like an AI start-up, then like they're kind of the tip of the spear on that. And so the explosion of products you've seen has come from ideas seeded by different folks from that group. So just as an example, like AI Agents was led by James Evans from Command, and then now Frank from Inari is taking it. And then, like Yana is working on LLM analytics, as well as a bunch of automated insights that we haven't even got to talk about yet. And then, you combine that with folks who have been at Amplitude for a long time and you get just incredible results. So like, yes, you kind of -- you pair a lot of long-time product veterans at Amplitude with some folks who have been in the AI ecosystem and you're just able to be really aggressive about the innovation.

Awesome. And then, just in terms of monetization, you talked about the fact that you're going to be giving away some of these features. How do you think -- are events going to be the denominator going forward in order to look at really kind of pricing and packaging? Or is there another way to think about that, that we should be considering going forward?

Okay. So this is a topic I'm very passionate about. A lot of people in the AI space talk about outcomes-based pricing. I think it is an awful idea because you spend out a lot of time arguing with your customers about, was that a good outcome? It's like -- what works much better is the same meter-based pricing that you had in SaaS. So whether that's, in our case, events, some seats, some cases like tickets handles or size of customer base or what have you. But the customers have been most receptive. Like they don't want to feel like they're getting screwed. And like whenever you try to introduce a new metric or a new way of doing pricing, it raises a huge number of red flags. Whereas if you go with something they've already been used to using like events and data volume, like they're extremely comfortable with that. And so, we're not -- yes, we're not planning -- while I think there are things we're going to continue to improve about our pricing to better make sure we have a fair exchange of value between us and our customers, we don't want to -- we actually are looking to simplify our meters, and that's something we're going to be coming out with next year where we're just going to focus on the events piece.

Great. Thank you, Clark. That will conclude our third quarter earnings call. Thank you for your time and interest, and we look forward to seeing you on the road this quarter as we attend conferences hosted by UBS and Needham. Thank you.

Thank you, guys.