Fiverr International Ltd. Aktie

All right. We're going to go ahead and get going. I'm Doug Anmuth, JPMorgan Internet analyst. We're pleased to have with us Fiverr's CFO, Esti Levy Dadon; and Chief Business Officer, Jinjin Qian. Fiverr is a global marketplace that connects freelancers and businesses for digital services in more than 750 skilled categories. In the past 12 months, almost 3 million customers bought a wide range of services from freelancers working in over 150 countries. Esti recently became CFO in March and has been in several finance roles at Fiverr since 2016, most recently as EVP of Finance. And Jinjin recently became Chief Business Officer in March and has been in several roles at Fiverr since 2018, most recently as EVP of Strategic Finance. So welcome.

Thanks for having us.

All right. Maybe we'll start kind of big picture. You talk about the U.S. freelance market is nearly $250 billion of TAM and then including international, it should be much bigger. AI is also reshaping the nature of work as it expands kind of project ambition and compresses task duration. But how do you think about the TAM evolving in an AI-enabled economy?

It's an interesting question because, as you said, AI is really changing the nature of work and the skill landscape across the talent. We see the simple jobs on the marketplaces getting less demand, whereas there's explosive demand in AI-related services and complex projects as human is needed for more sophisticated strategic outcome-based work. I would say from a long-term perspective, there's definitely going to be continued growth in the TAM because the work is going to become more fragmented as human gets plugged into the agentic workflow. And freelancers offer this tremendous amount of flexibility, elasticity and also helping SMBs stay at the forefront of the technology. So we definitely believe there's a significant growth runway for the role of freelancers and the role of Fiverr.

Okay. Great. What are -- when you think about kind of TAM and how that expands and evolves, what are some of the key friction points that remain to unlock this opportunity, particularly as you transition from kind of a transactional marketplace to a more kind of trusted work platform?

I think there are a few things. One is the platform was built like over a decade ago, where it was really optimized for simple like voiceover, blog post writing type of jobs. And as jobs get more complex, we are going through this transformation to really upgrade the underlying data infrastructure and the matching infrastructure to enable a more sophisticated matching between the skills and the talent. That's one.

Second is as any technology shift, you are going to see demand of the new skills leads the supply of the skills, especially we're talking about this drastic shift. So we definitely see a lot of AI-related needs there where the freelancers are going through this upskilling, adapting to the new and -- adopting the new workflow, all of that. And as a platform, we're helping them to go through this transition as well. So you are going to see this mismatching timing. But overall, I think it's a huge opportunity because every time you have supply skill gaps and shortage, this is where marketplace really plays a role in empowering both sides.

Okay. Great. So the company has made some key leadership changes as well as undergone a fairly significant restructuring over the past several months. Maybe you can just talk about how the changes sharpen accountability, accelerate decision-making as you execute this transformation.

Yes. I think we've been kind of going through upmarket for a few years now, and we've seen lots of progress in terms of upselling our existing customer base. That said, we realize that the world is changing much faster and to really become more relevant, we have to take a more drastic step and really lean into the opportunity in a more aggressive way. And that's kind of why we did the changes both from the execution standpoint, also from a strategic pivot perspective to really be more focused into empowering the complex projects and the high skill talent.

And how do you think about kind of the right size for the organization?

I think we're in a pretty good spot right now. Obviously, every organization is going through this AI adoption and becoming AI native. I think there's a lot of opportunities as we see it. And I think we're -- we kind of did the right transition, and we are always kind of the first to do these transitions in the market and really very focused on transitioning the workflows internally and also getting the talent that we needed to be there. Yes, very excited.

Yes, we removed some of the management layers so we can move fast. So what we're seeing now that when the team is focused and smaller, we can actually move faster and execute more.

Got it. Okay. So let's just dig into the kind of transformation more, which you've talked about as kind of multiyear and built around like 4 key pillars. And I think those are matching and product and go-to-market and then operational excellence. So maybe you can talk about how we should think about progress here over the next several quarters. And if you could kind of give us an update on where you are on each of those.

Sure. So we are only 2 months into this. We're just getting started. But already, I think we're already seeing a lot of focus and momentum into this transformation. On the matching side, like I said, it's really focused on the underlying brain of the marketplace where we are building a comprehensive knowledge graph to really capture all the data that as a transactional marketplace, the proprietary data we have. Historically, we know we're only scratching the surface of unlocking the potential for data. But now we are extracting the information from those unstructured data to enable the matching algorithms to really increase the conversion and the satisfaction rate from matching both sides. So that's on the matching.

On the products, we see a lot of success with the dynamic matching and managed services product. We are doing a lot of the upgrade on the product experience from enabling higher complex jobs to collaboration services to also enable the talent to be able to onboard and manage their business in a more sophisticated way because now when we were talking about high-scale talent, their need is very different from the less sophisticated freelancers. So that's on the product. And on the go-to-market, we are doing many different things, opening new channels that historically we don't do much, things like account-based marketing. We're doing more social listening. We're being more creative, leaning into more targeted efforts to drive high-value projects to Fiverr compared to historically a more broad focused volume-driven business. And then lastly, operational excellence is all about applying AI across org from customer support to production to data. So there's a lot of projects there to unlock productivity.

Got it. Okay. Let's dig into a few of those a little more. Just on matching, you kind of talked about transitioning from traditional search to agentic matching. What does that mean? How far along are you in that vision?

It's all about ultra personalization, right? Think about the traditional search is a pretty average experience and you optimize based on keywords and impressions. Now in the agentic workflow, you almost have this one-on-one kind of chat with ChatGPT where every conversation is contextual, every conversation is personalized. And we are trying to leverage that LLM capability into the matching because if you think about it, the amount of unlock you can do with the service marketplace is significantly larger with the physical goods because when you're trying to hire someone, it's very subjective. It's very contextual versus if you're buying iPhone, it's a standardized product. So we do see that LLM is a huge enabler for a service marketplace for us to really take it to the next level.

And because we have the -- so the transactions are happening in our platform. So we have a lot of data now. And now with agentic, you can actually better understand and do the right matching because you really know the buyer from past experience if you had and the TAM definitely from past experience that you have. So together with agentic, the matching is much better.

Okay. And then on go-to-market, you've talked about expanding into more AI-native distribution channels and enterprise partnerships. Maybe you can just talk about how you're thinking about integrations with LLM platforms? And is there anything meaningful coming out of that LLM source traffic kind of contributing at the top of the funnel today?

Yes. We are investing a lot into the Geo technologies. And we've seen kind of really encouraging results in the last few months where like there's a lot of know-how into how to impact LLM organic search to really increase the relevancy and increase the brand awareness in the LLM channel. So this is one area of investment. Second is we are one of the first partners to kind of participate in the ChatGPT like paid programs.

So we're also testing. And obviously, I think it's very early for OpenAI. A lot of the infrastructure is in the process of getting built out. But like I said, we want to be kind of ahead of the curve and really be there. And from an integration perspective, there's a lot of ongoing dialogue. I think this is a space where it's very dynamic. I would say most players, we haven't seen like significant traffic coming from these channels yet just because everything is so new and early. But down the road, definitely, it will be a very important kind of channel.

Okay. All right. Let's shift gears a little bit. Marketplace revenue has remained under some pressure, right, broader softness just in the SMB -- with SMB sentiment. And competition has recently called out some slowdown just around tariffs, energy prices, high interest rates. What are you seeing just in terms of macro kind of year-to-date?

Yes. I think definitely both SMB and macro has been under pressure. We kind of called this out last quarter. I think we kind of always see this a little ahead of other players in the market. So nothing kind of new to call out since the beginning of the year. I think we've -- it stayed relatively stable, and we're kind of executing according to the plan laid out from the beginning of the year. But definitely, we saw similar trends kind of earlier.

Okay. And then when you think about kind of how you're projecting revenue growth for '26 or revenue, which is down some. How do we think about how much is kind of more self-imposed from the transformation that you're undergoing and like deprioritization of low-end work versus some of those macro characteristics?

I think we're trying to be very disciplined on the transformation. I think part of the deprioritization is investing less into kind of the local optimization of the low end, but doesn't mean we're doing things to hurt the existing business. Quite the contrary, we're trying to be very disciplined and let the -- what works continue to work and drive majority of the company efforts into developing and investing into the future opportunity. We -- when we gave the guidance at the beginning of the year, we kind of factor in kind of a wider range of the guide to factor in the uncertainty in the macro. But overall, I think the guidance methodology and kind of the visibility we have into the year remains pretty consistent to what we did in prior years.

Okay. Let's talk more about AI. Just obviously, the impact remains kind of just key debate for investors. How are you seeing clients engage in terms of AI-related work? And any color just on how AI-related GMV is trending?

It's trending really healthily. I think after the initial AI, I would say, craze, I think we're seeing more and more customers realize that the human talent is needed for them to implement AI and really unlock the business impact of AI. And we see customers take what they did on AI platforms and come to Fiverr to find experts to finish -- get them to the finish line. So across the board, from AI development to automation workflows, all of that is like very popular, triple-digit growth, high double-digit growth.

Okay. And are there particular categories or services where you're seeing overall traction stand out more?

Yes. I think AI development is kind of one of the bigger categories within AI. Automation, all the Zapier, Make very popular. A lot of the content creation, also popular. marketing, another area, SEO, GEO, marketing copies and marketing automation, all very popular.

Okay. And then just in terms of agents, how do you think about implementing agents on both the buyer side and then also for freelancers to improve the overall experience?

I think agents are -- today, it's everywhere. We haven't really done a lot of integrations, but we know both sides are deeply plugging into all these agent tools. For us, I think agent is what I call in the collaboration product pillars, right, where it is an important layer to enable workflows and enable our buyers and sellers to work with each other more efficiently. But it's not the core moat of the marketplace, which lies within the human connections and then agent could be useful top-of-the-funnel tools. It could be useful retention and monetization tools. But we have to -- it has to anchor around the core human connection, which is Fiverr's core proposition in the AI world.

Okay. All right. Let's talk about just top line drivers a little bit. We talked a little bit just about GMV. We are seeing active buyers decline, but offset by strong growth in spend per buyer. And I know the transactions above $1,000 grew double digits in 1Q. Maybe you can just talk about kind of what that path looks like for high-value work to become a larger portion of GMV.

Yes. We are already seeing GMV from over $1,000 growing double digit year-over-year. And the goal is through this transformation, they are going to become a majority part of the business. From a buyer perspective, I think the metrics will be under pressure in the near term just because from a buyer count perspective, we've got substantially -- we're carrying kind of a large population who's kind of focused on the smaller side. That said, I think as we go through this journey, we will provide more clarity on what we're seeing in terms of the larger buyers and larger projects that will hopefully give more color to investors on kind of the trends and the trajectory we're seeing there.

Okay. And what are some of the key investments that are required? Just when you think about kind of increasing that mix of $1,000-plus jobs, what are the key investments you need to make on your side?

I think it's around the core pillars we talk about, the underlying brain, which is the matching, the experience layer, which is the product and then the go-to-market, which is how do we scale with strong unit economics. I think these 3 layers will together get us there.

Okay. Services revenue grew north of 50% last year, but we know it's going to exit at a low single-digit growth rate in '26. Maybe you can just talk about that path and kind of trajectory through the course of the year.

Yes. We do expect the services revenue growth rate to come down Q2 and then continue into second half of the year, largely because of the lapping of the acquisitions and also just the lapping of some of the investment we did last year around the modernization programs, which we did a lot of expansion last year. And this year, this is not going to be the core of the transformation. Yes, I think overall, really focused on driving GMV reacceleration and high-value projects acceleration with the plan.

Okay. And then can you just talk about current Seller Plus subscription penetration, how you think about kind of where you are today? How much room there could be as you drive more value for sellers?

It's a really healthy program. I think it's kind of -- we had a really nice ramp in the last 2 years. I would say it's at a good spot where we've got a very core strategic seller pools who are very active in the program. Now I think it's kind of into a more relatively mature and stable growth at this stage compared to the previous kind of early expansion period. So expect the program to continue to be healthy and -- but it wouldn't kind of ramp at the same rate compared to previous years.

Okay. All right. And then I guess just as we think about '26, we know this kind of transformational year, positioning for growth in '27. I know it's early, but any kind of view on what normalized growth looks like on the other side?

I wish I can give you a golden number. But obviously, you've known us for a long time and you've known me for a long time. I think growth is always top priority and us, really driving growth is top of mind. And I think labor market is an area where the TAM is so large, and we're going through the shift. So we really believe we're -- we have a unique right to win here and kind of really investing into that. So definitely not going to grow at the current rate.

Okay. Let's talk about profitability a little bit. So you've guided to 18% EBITDA margins at the midpoint for '26, the core business at around 20% and then a couple of points of drag just around transformation investments. Maybe you can just talk more about some of those investment areas and kind of how you think about overall margin trajectory.

Yes. So I think in 2026, as you said, this is the transformational year. So investing on all of the pillars that Jinjin said, it's -- so it's mainly about hiring that will pick up during the year. Generally speaking, the marketplace is healthy, is generating healthy cash flow. Now as for the future, we plan to invest in the transformation, but also to be very cautious on the bottom line. So not to sacrifice the bottom line in the future, but -- and invest in the top line, as I said, very disciplined.

Okay. You've previously talked about -- you kind of previously targeted 25% margins in '27, but I know that was a little while ago. I guess just any -- again, kind of early, but any view on just how you think about what the right long-term margin framework is?

So 25% is still the long-term EBITDA target. But given the transformation that we're going through, it will take us more time to get there, but it's still definitely the target.

Okay. All right. And then just thinking about capital allocation, I think you have about $60 million remaining on the buyback authorization. You have been active in different quarters recently, but maybe you can talk about kind of thought process there.

Yes. So from a capital allocation perspective, so #1 priority is growth of the business and investment in the business and being able to execute the transformation. We believe this would really bring the highest value in the long term. We've been doing buyback in the past. We have an authorization, as you said, $60 million. So we'll continue to do that, but in a thoughtful manner. And then on the M&A side, from a capital allocation perspective, we're always opportunistic, but nothing currently to call out. Only if something makes sense to us and support the strategy, we'll go for it.

Okay. And then just on -- going back to AI for a minute. Just when you think about how AI is changing how companies work, what stands out to you just in terms of some of the limitations around traditional staffing and hiring models that you're kind of ultimately much better equipped to handle?

Yes. I think we kind of pioneered this outcome-based, project-based hiring many years ago. And in some argument, like we're kind of ahead of the market. Now with agentic, I think everyone has talked about this outcome-based and which is really fitting into kind of Fiverr's model. And if you think about the future agentic workflow, you're going to have agents doing 80% of the work and human gets plugged into the 20%. You have agent, agent and a human comes in and then agent continues. So the outcome-based like model that Fiverr has is much more seamlessly can get plugged into this agentic workflow, where we see the scope of work, we see the price, we see the efficiency and the quality. And that gets us really aligned with agentic workflow, which is very hard if you think about job board or the traditional hiring process. Yes, that's definitely -- and then the data we have is -- we kind of build this wealth of transaction data is almost the training data with evaluation with it. So like this is part of what we are investing with this transformation is to unlock that potential.

And to create the upmarket flywheel that worked for us for the low transactions, so to recreate that in more complex work.

Okay. Got it. Okay. And then maybe just last one. Anything that you think is maybe misunderstood or kind of underappreciated either about the company or the opportunity in the industry?

I think there's this general fear about AI is going to take over all of our jobs and you don't need human anymore. I think this is probably overdone in my opinion, because at the end of the day, for any business to succeed, if everyone have the same access to the same AI models and human is still the differentiation. So we really hold that view, and we are starting to really see it from the activities on the marketplace. So I think this is a part that market doesn't fully give us credit for.

Thank you to JPM for the forum. I've looked at the sector quite a bit. I have a number of questions, one of which is we're living now in a siloed world where countries are putting up borders. It seems to me it might make -- increase the value of accessing labor markets synthetically through marketplaces like yours where we can find specialist pools of labor around the world. So I was wondering whether that's -- it seems to me that would be probably a net positive rather than a net negative in a siloed world. And then I was also interested in -- so that's the first question.

The second question was interested in how you add value to the labor pools that you address. So whether you offer them services, benefits, other things that they need, particularly in the independent employment market as well as getting people paid, whether you link in with good companies like Wise, Revolut, so in order to efficiently pay people in different currencies as you're, I would assume, inherently sort of a multicurrency, multi-country business model.

Yes, definitely. I think Doug said in the introduction, so we are global marketplace. We cover almost every corner of the world. And over half of our transactions are cross-border. And it's not always from developing -- developed countries higher in developing. There's the other way around as well, depending on the use cases. So we do see a very dynamic cross-border kind of a marketplace. On the payment side, I think we cover a basket of currencies, and we really enable both the customers to pay in their local dollars -- local currency and also the freelancers to get paid in their local currencies. And we provide quite a few range of like financial products to help them manage their business better.

We've also done a lot of learning and education for the community. I think community is a big part of how Fiverr is different. We have local communities in, I think, dozens of cities who runs like hundreds of events annually to -- especially at a confusing time like now, like we -- freelancers are very lonely professionals, right? They don't really have a company. They don't have teams. So Fiverr is kind of their "home" like we help them navigate the current environment. We have on the benefit side, I think we -- there's a lot of future opportunities. We do some partnerships and providing like insurance benefits. But I think this is an opportunity where the whole industry really is going through. There's a lot of opportunity to do more for the freelancers. I think especially in the U.S., I think not much. It's hard to be a freelancer. So...

Would you consider dealing with private company like Deel where you bolt on with them or [indiscernible] partner?

We -- Deel is not super relevant for us because they do more EOR, like long-term contractors, EOR, like helping right enterprises to set up like external offices. Our freelancer engagement are more on demand. So they actually directly pay through Fiverr, like we don't necessarily need Deel because we don't issue kind of payrolls. Deel is more like a payroll provider.

PEO.

Right. Like we don't -- like for our use cases, we don't necessarily need PEO.

Why not the micro...

It's a long question. So right now, we don't help -- we don't do long-term contractors currently. The most use case on Fiverr is very on-demand project-based, few weeks kind of -- think about Uber. You can plug in, plug out any week of the -- day of the week. So this is the primary use case.

All right. I think we're going to wrap up there. Thank you, Jinjin. Thank you, Esti.

Thank you for having us.

Good morning, and welcome to the Fiverr First Quarter 2026 Earnings Conference Call. [Operator Instructions] Please note that this conference is being recorded. I would now like to turn the conference over to Ms. Emily Greenstein. Thank you, and over to you.

Thank you, operator, and good morning, everyone. Thank you for joining us on Fiverr's earnings conference call for the first quarter that ended March 31, 2026. Joining me on the call today are Micha Kaufman, Founder and CEO; and Esti Levy Dadon, CFO.

Before we start, I would like to remind you that during this call, we may make forward-looking statements and that these statements are based on our current expectations and assumptions as of today, and Fiverr assumes no obligation to update or revise them. A discussion of some of the important risk factors that could cause actual results to differ materially from any forward-looking statements can be found under the Risk Factors section in Fiverr's most recent Form 20-F and other filings with the SEC.

During this call, we'll be referring to some key performance metrics and non-GAAP financial measures, including adjusted EBITDA, adjusted EBITDA margin and free cash flow. Further explanation and a reconciliation of each of the non-GAAP financial measures to the most directly comparable GAAP measures is provided in the earnings release we issued today and our shareholder letter, each of which is available on our website at investors.fiverr.com. And now I will turn the call over to Micha.

Thank you, Emily. Good morning, everyone, and thank you for joining us. Let me start with the headline. Q1 was a solid quarter of execution with both revenue and adjusted EBITDA coming in at the high end of our guidance range. Esti will walk through the details shortly, but the underlying message is this, we are focused on executing the strategic transformation while being methodical in managing the existing business across both top and bottom lines.

Maintaining financial discipline and transparency throughout this transformation is critical, and we are committed to doing that consistently and credibly. Let me now turn to the transformation or as we mentioned last quarter, we are in the early stages of a multiyear journey to reposition Fiverr from a transaction-oriented marketplace into a trusted work platform for complex high-value outcomes.

This is not a cosmetic shift. It is a fundamental evolution of how work is matched, delivered and orchestrated on our platform. Our North Star is clear: to become the most trusted platform for completing high-value, high-trust work. This means enabling businesses and talent and increasingly AI-driven workflows to collaborate effectively on complex outcomes.

Two months into the transformation, the early signals across all pillars of this transformation are consistent with our plan. First, we are strengthening the high-end talent flywheel and expanding into more complex, higher-value projects.

Projects over $1,000 continue to grow at a strong double-digit rate with clients completing $1,000-plus projects, up 18% year-over-year. We are also seeing increasing participation from talent delivering these engagements. What's important here is not just the growth. It's the nature of the work. We are seeing businesses come to Fiverr not for isolated tasks, but for multiphase mission-critical projects.

For example, one, a global health care company is working with talent on Fiverr to produce over 30 multilingual animated assets for a product launch with ongoing spend across multiple engagements. Two, a C2C sports platform in New Zealand built a full mobile application through multiple development phases on Fiverr. Three, a European entrepreneur is building an AI-enabled invoicing SaaS platform to comply with regional regulatory standards.

These are not one-off gigs. They are sustained high-value engagements that require coordination, iteration and trust. This is exactly the segment we are targeting and exactly where the market is moving towards more strategic outcome-based engagements. Second, we are investing heavily in matching infrastructure and experience. This is our main differentiator and the key to driving trust and quality, which are the core primitives of the market.

Our research and internal data confirm this. The primary differentiator in hiring platforms is not price. It is talent, quality and trust. Historically, Fiverr has won on ease of use and speed. Winning upmarket means extending that advantage into quality and trust, and that is exactly what our infrastructure investments are designed to deliver. That is why we are rebuilding our matching infrastructure from the ground up.

We are moving from keyword-based matching to context-aware, outcome-driven matching powered by a knowledge graph that captures not just who the talent is, but what they have delivered in what context and with what results. At the same time, we are shifting ranking from optimizing for conversion to optimizing for expected project success and buyer satisfaction.

The data is already moving. Recent tests in Fiverr Pro show mismatch rates down nearly 10%, and we are consistently seeing higher value engagements, leading to stronger repeat behavior. These are the early proof points of a durable trust mode. Third, we are evolving Fiverr into a comprehensive work platform. Today, most high-value projects on Fiverr run on infrastructure built for a different era of the platform. We are addressing this by building an end-to-end fulfillment layer that includes visibility into project progress, early detection of risk, structured feedback loops and active orchestration by Fiverr.

This is a fundamental shift in responsibility and perception of responsibility. We are becoming an active partner for our clients and talent, not just a passive connector. Over time, this infrastructure will also allow Fiverr to integrate seamlessly into agentic workflows, where AI handles coordination and humans provide judgment and accountability.

Fourth, we are expanding our go-to-market capabilities to scale more aggressively into high-value work. We are now building 3 new growth engines. First, talent-led growth engine, driving high-quality demand directly to high-performing freelancers. Second, industry-led growth engine, building tailored experiences for specific industries such as e-commerce and early-stage start-up companies. And third, partner-led distribution, embedding Fiverr directly into workflows and platforms where high-value demand already exists. These initiatives expand beyond traditional performance marketing and are designed to create scalable, durable growth engines aligned with our upmarket strategy.

Finally, we are improving execution across the organization. We are optimizing production workflows through better telemetry, identifying bottlenecks and increasing discipline in delivery. At the same time, we are rebuilding how work is executed with AI agents at the center and human judgment where it matters most.

This approach enables faster decision-making, reduces handoffs, improves product quality and drives efficiency across the organization. Mastering this as a company will also allow us to generate a reusable blueprint for our customers and talent to replicate and enjoy.

Stepping back, the fundamental dynamics of this market are moving in our direction. AI is increasing, not reducing, the complexity of matching the right talent to the right work. The demand for trusted outcome-based platforms is not a future possibility. It is already showing up in our data, in our customer examples and in the infrastructure we are building.

Fiverr has a differentiated model, a compounding data advantage built on real transaction outcomes and an end-to-end platform that no point solution can easily replicate. We are executing with urgency and discipline, and we are confident in where this leads. With that, I'll turn it over to Esti for the financial details.

Thank you, Micha, and good morning, everyone. We delivered a strong first quarter with both top and bottom lines exceeding the midpoint of our guidance. Revenue was $105.5 million, down 1.6% year-over-year, reflecting continued growth in high-value work, offset by headwinds in low-value transactional activity on the marketplace alongside a continued growth of service revenue. Adjusted EBITDA was $22.6 million, up 16.3% year-over-year and representing an adjusted EBITDA margin of 21%. This is an improvement of 330 basis points from a year earlier as we continue to execute with strong financial discipline.

Turning to our revenue segments. Q1 marketplace revenue was $67.1 million, driven by 2.9 million active buyers, $356 in spend per buyer and a 27.7% marketplace take rate. The continued momentum in our upmarket strategy and shift towards more complex engagement is clearly showing in our cohort behavior with spend per buyer growth of 15% year-over-year.

Projects over $1,000 grew at a strong double-digit rate, driven by 18% growth in clients completing these engagements. This growth is coming from both new adoption and repeat behavior as buyer expand into larger use cases, along with increased usage of dynamic matching and managed services.

Looking ahead, macro conditions remain largely unchanged. Based on current trends, we expect marketplace growth for the remainder of the year and on a full year basis to track broadly in line with Q1 performance. Service revenue in Q1 was $38.4 million, up 30% year-over-year and accounted for 36% of total revenue. Services revenue came in slightly higher than expected as AutoDS ran a successful campaign at the start of the year, pulling certain user sign-ups and revenue forward from Q2 to Q1.

Overall, our expectation for services revenue for this year remain largely unchanged with growth moderation in Q2 and continuing into the second half of the year. As Micha mentioned, 2026 is a transformational year for us as we make critical foundational investments to strengthen our high-end talent flywheel. Our decisions are centered on improving marketplace quality and trust, prioritizing high-value work and driving more focused execution with strong financial discipline.

On capital allocation, we continue to take a disciplined and balanced approach. Our strong balance sheet allows us to invest in growth, returning capital to shareholders and remain opportunistic on M&A. We generated $21 million in free cash flow in Q1, and we expect to continue executing our buyback program in a thoughtful manner. As of March 31, 2026, we had $59.5 million remaining under the current authorization.

Now on to guidance. For the full year 2026, we expect revenue to be in the range of $380 million to $420 million, representing a year-over-year growth of negative 12% to negative 3%. We are raising our full year adjusted EBITDA guidance and now expect it to be in the range of $64 million to $80 million, representing an adjusted EBITDA margin of 18% at the midpoint.

For the second quarter of 2026, revenue is expected to be between $95 million to $103 million, representing year-over-year growth of negative 13% to negative 5%. Adjusted EBITDA is expected to be between $16 million to $20 million, representing an adjusted EBITDA margin of 18% at the midpoint.

Our revenue outlook reflects solid execution in Q1 and the continued uncertainty in the market conditions. Our adjusted EBITDA guidance reflects the strength of our core marketplace profitability and our continued commitment in maintaining disciplined margin profile while investing in the transformation.

As we look at the rest of the year, we are staying focused on our core priorities, driving progress in higher-value work, improving trust and quality and building scalable growth engine. We believe these are the right indicators to evaluate the business as we transition to the next phase. With that, we will now turn the call over to the operator for questions.

[Operator Instructions] We have the first question from the line of Eric Sheridan from Goldman Sachs.

Maybe 2, if I could. One, just coming back to the transformation strategy. I want to know a little bit more about the duration of sort of completion of what you call sort of the infrastructure layer and putting the pieces in place, and how we should be thinking about when you exit that phase of the transformation and some of the execution shifts more predominantly to go-to-market or what the mix is of building blocks relative to execution on the transformation strategy, that would be one.

And then the second one would just be, you talked a little bit about partners and evolving the go-to-market strategy. I want to know if you could go a little bit deeper in terms of what those types of partners might look like and what market opportunity they might open up that maybe you're under-indexed to today?

Essentially, the transformation is an ongoing process, and since we just started it mid last quarter, we are anticipating to see results over the remainder of the year with more emphasis, because it takes time between the things that we develop and release until they show up in the numbers, to see this more in the second half of the year and definitely towards the end of the year. And as we said, we will continue to be transparent on what we're seeing and the progress there.

As the transformation, my belief is that the entire market is in a transformational moment where every business needs to adapt to a new reality where AI plays a critical game, not in just making products better and more efficient, but also being able to connect with agentic realities where agents are actually using the platform. This is not limited to this year, I think that this is going to be a transformation that every business out there will have to implement in the coming years.

It's very similar in my mind to the digital transformation when businesses went from the offline to the online and now are seeing a new reality. Now we are already seeing some initial signals that we called out in the opening remarks of areas where that transformation has started, and we started rolling out experiments and new products and how they influence a higher quality matching and focuses on better conversion and better retention around high-end talent and larger scopes.

So over the next few quarters, we will continue to report on what we're seeing. The progress, and obviously, the more history we have in doing this, the results should accumulate. And as we said, this is going to be a turnaround year where the next years are going to be years of growth.

In terms of the other question regarding partners and go-to-market strategy. Again, we very much focus on this idea of human-in-the-loop partners where the requirement for a skilled talent network to make judgment calls on AI's work and on calibrating models and checking integrity and ensuring accuracy is paramount, and I think that this is an area where Fiverr can play a major role.

That together with agents that we're developing to automate some of this work to make sure that the experts are actually focusing only on things that humans need to focus is a very important and critical role in what we're doing. It is still early, there's a lot of AI automation use cases. We're running successful pilots with some initial customers, and we see that there's a lot of demand for Fiverr to become a fulfillment partner for SMBs to adopt automation. So again, early in the process, but we will have more things to call out in future quarters.

We have the next question from the line of Jason Helfstein from Oppenheimer.

Kind of like a 2-part question, but on the same theme, so obviously, you've had a front row seat to this whole evolution of how agents are evolving the business. As you're seeing kind of even these more cutting-edge frontier models coming out, how is that further evolving your view on kind of how both you will leverage this technology, how your companies -- how your customers will leverage it?

And then there's also been discussion among investors that AI agents are like lowering the barriers to new business creation. There's like more -- new domains coming online, I think a record number of apps being submitted to the app stores. I guess like how do you think about that, like is that a positive for Fiverr or a negative for Fiverr? Can you leverage that? Just kind of broadly all bring those topics together.

So essentially, the way we're thinking about how agents are becoming a part of what we're doing, essentially, agents are very much learning from human -- skilled people on how to run workflows much, much faster, much more efficient, 24/7. But at the same time, a lot of what agents are doing require ongoing judgment.

And it's much like everything else with AI. Everybody has the access to the same AI which means that also everybody has access to the same agents that are available out there. Just having access to this technology doesn't give you a competitive edge, it just flattens everything and it -- maybe it elevates the floor. But on top of what agents are doing and how you create skills for agents, how you create workflows that combine multiple skills, multiple agents, that is an art.

And that is what a lot of companies are actually focusing in and providing their employees, their expert skills onto agents. Now in the case of businesses, not all businesses have the talent to actually train an agent and oversee what the agent is doing and providing judgment and calibration and fine-tuning.

We see this on Fiverr. The implementation of agents across our system internally require tremendous amount of calibration to overcome hallucinations, inaccuracies or just moderate execution. So definitely, the role of an expert, of an employee, of a freelancer is changing, but it is highlighting the uniqueness of what they can build, bring to the table to provide an advantage.

Now, when we think about lowering the barriers, or agents lowering the barriers for business creation, this is amazing news for us. I've seen lately staggering numbers on the launches of new products in recent months. I believe April was the highest month with over 19,000 new announcements on product and company releases.

On the one hand, the signal to noise is extremely complex because it makes everybody a builder, but building something gives you nothing. It's all about the deployment, it's all about taking you to the market. It's getting noticed, it's validating and then it's scaling. These tasks are largely unresolved yet by AI.

Can AI help in this? Yes. But generically speaking, because it provides the same help for everybody else, again, flattening everything. What gives you that competitive edge when you create something or you almost create something and you want to improve it and then you want to deploy it and then you want to scale it, this is where experts come in.

Now the reason why I believe that this is not still showing up in the numbers is this is a transformative period. I remember the digital transformation from 2000. It took time for businesses to understand that if you don't have a website, you're going to be out of business over time. The same goes with AI, and the same goes with experts that need to come with AI to make your AI or your execution better than your competitors.

And I think that we're in the early innings. It's going to take some time. But all in all, I actually think that this is really a great upside for us. And when I look at the marketplace, we see AI consulting, business formation, all grow really strong double digit. So this, to me, I think it's an early sign of what would come.

Also AI-related categories continue to be super strong. AI development up 118% year-over-year. Marketing automation, also growing really strong double digits, and I can go into -- my answer is really long, so I'll stop here, but I can give more color around this.

I guess it hasn't automated us doing this earnings process yet, but maybe someday.

We have the next question from the line of Ron Josey from Citi.

Automation is the future, right? Can't wait. I wanted to ask a little bit more -- 2 questions. First is on just attracting the talent to the marketplace as we go to more upmarket projects -- upmarket and towards these multiphase projects. We're clearly seeing continued strength on spend per buyer. We're seeing that growth reaccelerate. So talk to us just about the talent on the marketplace as we go more upmarket and these multiphase projects.

And then one of the things that struck me, matching is a key part of the marketplace. And I think I heard the team talk about mismatch rates being down 10%. So during this transformation era, talk to us just about the ability to continue to execute on some of the key tenets of the marketplace like dynamic matching and the results that you're seeing.

On the first question, talent is super important. As we know from research, quality is core and the ability to match quality, drive quality perception is super critical, and this is very much in the center of this transformation for us. Now getting access or getting talent to the platform has never been an issue. Actually, we always had, I would say, an abundance of talent.

What we're more adamant right now is really understanding on the meta skill level, what does it mean to be a talent and for what type of customer and what type of an outcome, and creating this skill graph is super critical. In other words, what this means is, a, we're more picky about talent. But two, by improving the algorithm, improving the matching, we can create -- we can anticipate better outcomes, better happiness, and as a result, we also anticipate better retention in our customers.

Those are the key things. So when we call out the reduction in mismatch, this is key because this actually means -- and it's like hiring for any job, right? Some people that you hire turn out to be amazing, some you later on figure out that there is something that you missed or they missed, which makes the match not optimal. We don't want to tolerate this. And we actually think that if there's one huge advantage based on data that we've accumulated over 16 years, billions of interactions, tens of millions of transactions, is being able to take that data and actually make matching like anything that was done before by anyone.

And this is a reason to win, this is a reason to exist. So we're putting a lot of pressure there and seeing numbers, seeing the amount of actual matches that were mismatched in hindsight, getting down is a very positive signal, and we're far from that. We're just starting right now. And obviously, over time, as we accumulate more signals, deeper signals, we will continue sharing it with you guys.

We have the next question from the line of Bernie McTernan from Needham.

This is Stefanos Crist calling in for Bernie. I wanted to follow up on Ron's question on the matching, could you maybe give us any more details on what a baseline mismatch rate is or maybe what the revenue impact is of that 10% reduction? And then I also wanted to ask on the AutoDS pull forward, could you talk about what went right with that campaign? And is the pull forward just a dynamic of annual subscriptions? Or is there anything else?

In terms of matching, we haven't publicly shared any specific numbers, but with this transformation, we're really focusing on trust and quality as core primitives, so to us, mismatch is really about making sure that we have this deep understanding of what are the things that will drive a perfect match between a customer with their specific circumstances and needs and the very specific skill and validated experience of a talent to do this task, okay?

And again, as we move forward to do this restructure and refocus, we are going to be able to provide more color, more specific color as we really focus on those KPIs. And this nuanced understanding is super important. It's not just driving revenue today, but it is driving the flywheel and driving the repeat rate.

Now on the AutoDS, essentially, we had a very strong influencer campaign that we found great timing to do. In Q1 -- essentially, we were kind of focusing this on Q2, but we were able to actually execute this slightly earlier, not something that we plan to replicate, but -- which is baked into the numbers, which that has drove strong sign-ups at the beginning of the year, okay? So we called it out because this was a great opportunity for us to move something from Q2 to Q1 and do it earlier.

We have the next question from the line of Doug Anmuth from JPMorgan.

I have 2. Micha, can you just talk about where you are in terms of hiring AI native personnel within your own company and how you're thinking about that? And then Esti, can you just help us bridge the EBITDA margins from the 21% in 1Q to the 18% or so for the full year?

In terms of hiring AI native, we're on track. We continue to do this, and it's -- obviously, the -- I think the competition over talent is pretty brutal, but we've added incredible people into the team. And what's interesting is that if you really can find, identify and attract the right people, it's really different than it used to be before in terms of the amount of people that you need to do this because essentially, those are really AI natives are very much what I found in common, and I actually wrote about this. It is really this idea that they have this founder mentality, this entrepreneur mentality. And what's really common around them is that they're 10xers.

Essentially, there are people that can do 10x, and a lot of what they do is really put up these systems, these agents, these workflows and be able to connect them for the rest of the company and continue evolving this, tweaking, calibrating, validating. It's incredible. And this is really, I think, also points to future corporates being leaner, smaller, but having people that can actually multiply the work. And talent strategy is important, not just for us, but for all companies, top of mind, in my opinion. I'll let Esti address the EBITDA question.

So as for the 18% full year margin guide, so that actually reflects the hiring and the investment that we're doing in the transformation, and that picks up over time during the year, so it's consistent with our expectation at the beginning of the year. As you know, overall, we're very committed to execute the transformation and -- but that is with a strong financial discipline, so we are planning to execute that together with higher profitability and to continue to generate healthy cash flow.

We have the next question from the line of Brad Erickson from RBC Capital Markets.

I guess all this transformation talk, larger buyers, et cetera, I wonder, do you think about adjusting the economics or take rates or pricing or how you merchandise your services at all to kind of serve that type of customer? And then along those same lines, what would you say you want to kind of be signaling here this morning on overall marketing intensity as you pursue, again, this kind of maybe different customer profile than you have historically?

As for the first question, there's nothing to call out at the moment. The -- what we see from the dynamics is as expected. I don't want to speculate on future models. Obviously, it's a very dynamic company. We look at it all the time, but I don't have anything to call out at the moment. In terms of similar to the market with the customer profile and marketing, so we gave some example of use cases in the prepared remarks, and these types of examples are rising. That portion of the business is growing. It is taking a larger size of our overall activity, and as it continues to grow, it will drive the business for growth.

Now as we create more efficient, higher trust, higher quality solutions with everything I've outlined, again, I'm happy to go through it, but I was pretty long in my opening remarks about what we're doing with the transformation. What we're doing with acquiring customers, with creating the flywheel will become more efficient, allowing us to also invest more aggressively in marketing to feed this flywheel as it grows.

That's the plan. This is why we said at the beginning of the year, and I'm reiterating this, we're building ourselves for growth in the next couple of years. And this is really important and the foundational work that we're doing is not -- are not buzzwords, it is really the essence of the business.

We have the next question from the line of Matt Condon from Citizens Bank.

My first question is just on as we look at the green shoots that you're seeing in success moving into more complex projects, can you just talk about what you're seeing today as far as the product launches or go-to-market that's really driving that success in the transactions of $1,000-plus growing, clients purchasing projects $1,000-plus growing?

And then my second question is just you talked in the letter a lot about this comprehensive work platform, can you just talk about the specific products that you are really focused on today is really enabling that end-to-end platform, as you called it?

The truth is we're only 2 months into the transformation, so what we've progressed so far is not big product launch yet. What we're doing is really dealing with the fundamentals of the business, the infrastructure of the business, the data infrastructure, the matching algorithm, the quality improvements, all of these. It's not really about new product launches, but it is very much about how we enhance everything we do. In some cases, we completely rewrite those solutions.

Now we -- I called out job post as an example, dynamic matching, managed services, and they're driving large engagements on Fiverr. And we highlighted a few examples in the prepared remarks, but more broadly, we're seeing a clear shift in how customers use the platform. These products are fundamentally changing Fiverr from a place to complete isolated task into a platform to execute multiphase high-value projects.

So again, 2 months into it, it's not about shiny, new launches and creating promises. It's about going back to the basics and making sure that we provide a level of service, a quality of service that is unmatched. That is the focus, and this is where we're starting to see the numbers provide signals.

Does that answer your question, Matt?

Yes, that's great. I just had another question just on the comprehensive work platform, just the end-to-end platform that you're launching, like what are the capabilities that you really need to launch to enable this end-to-end service?

Sorry for skipping this. So on the product front, we talked about investing in an end-to-end fulfilling layer, which we think is key to identifying and increasing the value of Fiverr as an active partner in ensuring that the customers and the talent is engaging efficiently, and if there's anything in the process that we need to identify to course correct, we're there.

This is really important because as you go to more bespoke, more complex types of projects, being there and being a part of that transaction and making sure that you understand the scope, you understand the progress, you can create transparency. You identify early on if things are on track or are deviating from track is super important. And this is a whole new layer that we are creating, and it's important as we think about changing the perception of Fiverr as a high-end solution for high-end scope and talent.

And as we -- I can talk about those core pillars. I just don't want to reiterate the -- my opening comments, but it's all about the matching and the brain behind things. It's about the product itself. It's about the go-to-market and how we entertain, how we engage with customers, and it's all about this idea of operational excellence where we're creating this extremely high execution capability, which we want to also provide for our customers.

The learnings that we have as a team on how to be more efficient, where do you need to have a human-in-the-loop and where can Fiverr help with both providing you with the right tech, but also with the human-in-the-loop is critical. And all of these learnings are transforming from our internal execution into the tools that we are and we will provide for our customers and for talent.

We have a last question from the line of Josh Chan from UBS.

I guess with your move upmarket, what's the profile of the customer that you're ultimately targeting? You mentioned projects above $1,000, so is that the benchmark of what you're targeting? And then secondly, on free cash flow, could you talk about whether the Q1 level of free cash flow is roughly sustainable for the rest of the year and then your willingness to more aggressively buy back the stock at these levels?

So in terms of focus, it's still largely focused on SMBs, probably larger than the micro businesses, but still SMBs. There's a lot of untapped demand, both into larger customers and larger use cases. If every business and definitely midsized businesses and up are building their own new tech stack that includes agents, APIs, MCPs, all of these use cases require a tremendous amount of validation to check accuracy, integrity, compliance, security, all these things require the mix of both technology and human-in-the-loop to continue calibrating, validating, in some cases, building, fixing.

Those are areas that Fiverr is a perfect fit. And so we'll see those -- more of these use cases as we continue exploring this. But largely speaking, it's SMBs. And the projects of $1,000 and more is a proxy, it's a way to identify the spend capacity and willingness on digital services, and so that has provided us with a good point. Now it's $1,000 and above, and within our marketplace, we have everything in ranges of hundreds to tens and sometimes hundreds of thousands of dollars on transactions. But that's kind of a reference point to help us identify the seriousness and willingness to invest in your business.

As for cash flow, we generated $21 million free cash flow in Q1, and we plan to continue to generate strong cash flow and to be consistent and disciplined on capital allocation. Obviously, our capital allocation priorities remain the same. First and foremost, we are investing in the business, and we will continue to invest in the transformation and generating cash flow. Now as for buyback, we have authorization of $59.5 million, and we will use it and act on that thoughtfully over time.

This concludes our question-and-answer session. I would like to turn the conference back to the management for any closing remarks.

Thank you, Marin, for moderating the call today. And for everyone joining, wishing you a great day and looking forward to speaking soon.

Thank you. The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good morning, and welcome to the Fiverr Fourth Quarter 2025 Earnings Conference Call. [Operator Instructions] Please note this event is being recorded. I would now like to turn the conference over to Emily Greenstein, Investor Relations. Please go ahead.

Thank you, operator, and good morning, everyone. Thank you for joining us on Fiverr's earnings conference call for the fourth quarter that ended December 31, 2025. Joining me on the call today are Micha Kaufman, Founder and CEO; and Ofer Katz, President and CFO. Before we start, I would like to remind you that during this call, we may make forward-looking statements and that these statements are based on our current expectations and assumptions as of today, and Fiverr assumes no obligation to update or revise them.

A discussion of some of the important risk factors that could cause actual results to differ materially from any forward-looking statements can be found under the Risk Factors section in Fiverr's most recent Form 20-F and other filings with the SEC. During this call, we'll be referring to some key performance metrics and non-GAAP financial measures, including adjusted EBITDA, adjusted EBITDA margin and free cash flow. Further explanation and a reconciliation of each of the non-GAAP financial measures to the most directly comparable GAAP measures is provided in the earnings release we issued today and our shareholder letter, each of which is available on our website at investors.fiverr.com. And now I will turn the call over to Micha.

Thank you, Emily. Good morning, everyone, and thank you for joining us. Let me start simply, 2025 was an execution year, and we delivered. Revenue grew 10%, accelerating from 8% in 2024. Adjusted EBITDA reached $92 million, up 23% year-over-year with a 21% margin. We met the revenue and profitability targets we set at the beginning of the year while continuing to generate strong cash flow. Importantly, we achieved this while repositioning the business for where the market is headed. Products like Dynamic Matching and managed services are enabling us to expand into larger, more complex projects and drive sustainable wallet share growth. Spend per buyer increased 13% year-over-year, accelerating from 9% in 2024. Buyers spending over $10,000 annually grew 7% and GMV from projects over $1,000 increased 23%. These are not just product milestones. They reflect a broader shift in how businesses engage with talent.

As many of you saw in the shareholder letter we published this morning, following the restructuring we undertook a few months ago, we have since developed and begun executing a comprehensive multiyear plan to transform Fiverr from a transaction-oriented marketplace into a trusted work platform, one that enables businesses, AI models and agents to collaborate with talent on complex, high-value outcomes through intelligent matching, integrated workflows, end-to-end orchestration and fulfillment and durable trust. Before I go deeper into that transformation, it's important to step back and understand the broader environment that led us here and why we believe this is the moment to act decisively. There's a prevailing narrative that AI eliminates labor. That framing is incomplete.

What AI actually does to work is, one, it compresses task duration. What took weeks now takes days. Two, it expands project ambition. When execution becomes cheaper, scope grows and the number of projects grows exponentially. Three, it democratizes capability. Individuals can operate with domain expertise beyond their original knowledge base. The result is not less work. It is more ambitious work. Human talent remains essential. What changes is where value resides in context, judgment, orchestration, trust and ownership of outcomes. There will be displacement in lower-value transactional work. We are already seeing that dynamic. At the same time, demand for higher-value specialized work is accelerating at a healthy double-digit rate. As work becomes more nuanced and complex, matching talent becomes harder, not easier. That makes Fiverr's core mission of connecting businesses with the right human talent more relevant than ever.

Looking ahead, much of the workflow will become human in the loop. Hiring decisions will increasingly be influenced and in some cases, initiated by AI agents. In that environment, traditional resume-driven hiring models become inefficient and unreliable. Precision matching contextual data and outcome history becomes critical. So what does this mean for Fiverr? First, we see a significant opportunity. Today, projects over $1,000 represent less than 15% of marketplace GMV, yet they are growing 23% year-over-year. With focused execution, we believe this segment will become a materially larger contributor to our business. We are prioritizing 2 categories of high-value work. The first is complex orchestrated engagement requiring collaboration between businesses, talent and Fiverr. For example, through managed services, we support a Georgia-based automotive technology company with ongoing multilingual UGC production for the Canadian market, coordinating multiple creators each month.

This reflects growing demand for scalable, always-on creative production powered by global talent. The second is AI-native work building the AI-enabled economy. For example, we are partnering with AI model safety companies to provide domain experts who help identify vulnerabilities in foundational models. In another partnership, we are enabling enterprises to build AI workflows automation through white label solution that allows them to deploy AI agents quickly and cost effectively. In one case, we streamlined a historical case discovery workflow across Salesforce and Jira, reducing knowledge gaps and improving support efficiency. What would have required 2 weeks of internal implementation was delivered in 1.5 days at the cost of $6,000, reducing deployment time by roughly 90%. These examples illustrate where the market is moving and where Fiverr is leaning in.

Fiverr has a strong right to win in this AI-enabled talent economy. First, the future of work is human in the loop. Scarcity lies in high-quality human expertise, not AI agents. Fiverr operates one of the largest global talent networks and has deep expertise managing liquidity, quality and engagement at scale. Second, our end-to-end transaction model is built around outcomes. That structure integrates naturally into AI-enabled workflows and eliminates much of the friction inherent in traditional hiring system. Third, our data is a durable advantage. Over 16 years, we have captured not only millions of transactions, but the contextual relationship between buyers and sellers, what was delivered in what context and with what results. That depth of data enables precision matching in increasingly complex environments.

Capturing this opportunity is why we are making foundational investments across data infrastructure, back-end system and product experience, accelerating Fiverr's evolution into a fully AI-native talent platform. While we have made steady progress over the years, the velocity of AI innovation requires us to move faster and more decisively. A few months ago, we initiated a company-wide restructuring to accelerate this shift. We have since developed a multiyear execution plan built around 4 pillars: the first is matching, building advanced semantic and reasoning layer powered by proprietary data to enable AI-native talent matching. The second is product, transforming the experience across matching, fulfillment, collaboration and talent management. The third is go-to-market, expanding into enterprise and AI-native distribution channels with scalable growth engines. The fourth is operational excellence, becoming an AI-native organization across engineering, product and operations.

We expect tangible impact within 4 to 6 quarters, including a stronger high-value work flywheel and proven AI native growth loops. These milestones will position us for meaningful revenue expansion in the years ahead. Let me be clear, this is the moment to lean in. AI is not shrinking the market for human talent. It is reshaping access and expanding ambition. Platforms that own the intelligent matching layer between business demand and human capability will capture significant value. Fiverr has the assets, infrastructure and strategic clarity to lead in that environment. 2026 will be a transformational year, positioning us for accelerated growth in 2027 and beyond. Before I turn it over to Ofer, I want to acknowledge that today marks his final earnings call as CFO.

Ofer will continue as President, focusing on strategic investments and M&A as we execute this next chapter. Esti, after 10 years at Fiverr and 4 years as EVP Finance, will assume the CFO role. Her deep institutional knowledge and disciplined financial leadership provide important continuity as we execute through this transformation. Jinjin, after 7 years leading IR and strategy, will step into a newly created Chief Business Officer role, overseeing revenue, talent, fulfillment and business operations. As part of this transition, she will be relocating with her husband and 2 young children from San Francisco to Tel Aviv to take on this expanded responsibility. I'm truly excited about our expanded leadership team that will strengthen our ability to execute with focus and velocity as we move forward. With that, I'll turn it over to Ofer for the financial details.

Thank you, Micha, and good morning, everyone. I'm excited about the transformation we are undertaking and very happy to welcome Esti and Jinjin into their expanded leadership role. The work ahead is ambitious, but across the management team and the broader organization, there is a strong alignment, clarity and conviction on the direction we are taking. Most importantly, there is a shared sense of purpose that ties us back to Fiverr's founding 16 years ago. That shared sense of purpose brings tremendous focus, energy and confidence as we enter 2026. With that, let's turn to financial highlights. As we wrap up 2025, we delivered fourth quarter revenue of $107.2 million, up 3% year-over-year, while achieving record adjusted EBITDA and adjusted EBITDA margin.

Adjusted EBITDA for Q4 was $26.5 million, representing an adjusted EBITDA margin of 25%, an improvement of 470 basis points from a year earlier. We continue to generate healthy cash flow with $21.8 million of free cash flow in Q4 '25. We had a convertible note with a principal amount of $460 million, which was fully repaid during Q4 '25. We continue to execute a disciplined, thoughtful capital allocation strategy, and our strong balance sheet allow us to invest in growth, return capital to shareholders and remain opportunistic on the M&A front. Diving into our Q4 results. In Q4, marketplace revenue was $71.5 million, driven by 3.1 million active buyers, $342 in spend per buyer and a 27.7% marketplace take rate. Growth in this segment continues to be influenced by broader softness in the SMB sentiment and muted freelancer hiring demand.

More importantly, we continue to see diverse trends on the marketplace between low-end transaction and high-value work. GMV from transaction over $1,000 grew 22.8% year-over-year in Q4 and continued to accelerate. Looking ahead, we expect elevated volatility in marketplace revenue this year compared to last year as the transformational work we are doing intentionally deprioritize effort to optimize low-end transactions, which today represent the majority of the marketplace. As we make progress towards strengthening our flywheel for high-value and AI native work, we expect this focus area to become a larger portion of our overall business, which will lead to reacceleration of this segment. Services revenue in Q4 was $35.6 million, representing year-over-year growth of 18% and accounting for 33% of our total revenue in Q4.

The upside was driven by the continued strength in Fiverr ads, subscriptions and e-commerce solutions. For 2026, we expect more moderate growth in service revenue as the impact from the AutoDS acquisition normalize and the pace of expansion for Fiverr Ads and Seller Plus moderates compared to 2025. As Micha mentioned, 2026 will be a transformational year with critical conventional investment across data infrastructure, matching technology and product experience to strengthen our high-end talent flywheel. It is important to note that we are committed to executing the plan with strong financial discipline. The structural profitability of our core marketplace remains strong and is expected to stay north of 20% as we retain significant control to maintain the health and the profitability of the business.

At the same time, we will use a portion of the cash generated to fund the transformational work ahead. We expect that to impact the adjusted EBITDA by approximately 200 basis points in 2026. On capital allocation, we maintain a disciplined approach and expect to continue executing our buyback program in a balanced manner. As of December 31, 2025, we have $67.5 million left on the current authorization. Now on to guidance. For the full year 2026, we expect revenue to be in the range of $380 million to $420 million, representing year-over-year growth of negative 12% to negative 3%. Adjusted EBITDA is expected to be in the range of $60 million to $80 million, representing an adjusted EBITDA margin of 18% at the midpoint.

For the first quarter of 2026, revenue is expected to be between $100 million and $108 million, representing year-over-year growth of negative 7% to 1%. Adjusted EBITDA is expected to be $19 million to $23 million, representing an adjusted EBITDA margin of 20% at the midpoint. The wider-than-normal revenue guidance for the full year and the first quarter reflects the elevated uncertainty as we execute our transformational plan focused on high-value work alongside evolving market conditions. On the adjusted EBITDA side, the updated guidance for this year reflects the revenue trends we see as well as the impact from the investment into foundational works. That said, we do not expect structural change to the core business unit economics, and we expect our ability to drive increasing leverage for the marketplace business model remains intact. With that, we'll now turn the call over to the operator for questions.

[Operator Instructions] First question comes from Ron Josey with Citi.

I had 2, please. Micha, you talked about an execution plan around matching product, go-to-market and operations. And given the progress, I think, that we've seen around managed services and Dynamic Matching, just talk to us about how you see these investments in those 4 core areas sort of unfold? Meaning do you have more work to do on the product before we start investing in enterprise go-to-market? Any insights there would be helpful as we think about this multi-quarter transition. And then Micha, over the balance sheet, we ended the year with approximately $300 million in cash. I know we've mentioned M&A a few times on the call and Ofer's new role. Talk to us about what you're looking for maybe on M&A or just overall capital allocation.

Thank you, Ron, for the questions. So essentially, the way we're thinking about the investment is really to deprioritize low-end and low-value transactions and focus most of our investment in high-end, high skilled, larger scope projects, a segment that is currently under 15% of our revenues, and we think it should and it will contribute much more to bring us back to GMV growth. That's what 2026 is all about. It's about making that turn so that 2027 and beyond will be growth years. Now bear in mind that the nature of the transactions that are happening on our platform are changing. And we're running a platform that has been built over 16 years. Some parts of it need to be rebuilt. Some parts of it need to be reinvested in and aligned into this new reality.

As I've said in my opening comments, the matching portion of it has to deal with the more nuanced needs of businesses today. So we need to calibrate this and make sure that we maximize the usage of our very deep data assets in order to do this. The same goes for product. When we think about the entire matching and fulfillment management, we need to understand that larger projects require more sophisticated fulfillment and collaboration and matching capabilities, also understanding that on the demand side, we may not just see companies and human beings, but also AI agents that can actually take care or find benefit from using our platform.

When we think about go-to-market, again, in this case, we are expanding our go-to-market flywheel. So first of all, is the high-value projects of work. And then there's the aspect of AI native use cases. And I've given a few examples in my opening comments in our letter to shareholders where we see more and more businesses and foundational companies that are building more AI models and more agents and all of them require human in the loop to continue calibrating, ensuring their security, their integrity and overall being able to make them customer ready. And we're seeing more businesses that are coming to Fiverr because we probably have the largest and widest scale talent in the world on our platform. And so adjusting for all of these new needs will help us accelerate that portion or that segment of the market, which we feel is the most durable and most sustainable to ensure that we can continue growing for many years ahead.

Ron, on the M&A front, I will start by saying that we have $300 million, but the amount of cash is growing and expected to grow throughout the quarters. And then we continue to be highly disciplined in the way we utilize this cash, looking for tuck-ins and then large transaction at the same time. Of course, we're looking to going up market and any M&A should support this high end and flywheel, as Micha mentioned earlier.

The next question comes from Eric Sheridan with Goldman Sachs.

Just want to come back to the theme of deprioritization of the lower end as you realign the platform. How should we think that manifests itself in financials as we move deeper into the year? Are there any elements of things that will impact the OpEx line as you sort of pare back investments in the lower end of the market or elements where there could be more volatility than usual as we think about either the first half versus second half dynamic as you sort of reposition the business for the longer term?

Eric, thanks for the question. So when we think about the deprioritization, it's really to ensure that the majority of our resources are directed in growing the segment that, as we demonstrated, have grown significantly over last year and make sure that it becomes a much larger portion of our market. As a reminder, when you look at the low skills and small scope, a lot of that is being replaced with AI solutions. And still, that is a large portion that contributes to Fiverr growth. In that segment, we are seeing a decline. And we've been talking about this, and we don't foresee that, that decline is going to slow down. The assumption is that with the newer developments around AI, this will continue to be the case.

And so that centralization in our business has to change. And therefore, we're shifting those resources into ensuring that the high-end portion of our business that has been growing will become a larger portion of our overall GMV contribution. That is extremely important, and we want to make sure that we put every available resource towards that. But we're very committed to execute this transformation with very high degree of financial discipline. So we're going to protect the core business to continue generating healthy cash flow. And we talked about the structural profitability of core business to stay north of 20%. I hope this answers your question.

The next question comes from Bernie McTernan with Needham & Company.

Maybe just 2 for me. How should we expect the margin profile of the company to look after Fiverr Forward is done or completed or some progress on? Is this going to be a higher-margin company or a lower-margin company than before? And then how does Fiverr Go fit into this? Are you seeing Fiverr Go help higher-value transactions already? Or just interested in terms of -- if this is still a key product going forward?

I think on the margin profile, so we are going to see some lower margin in terms of EBITDA in the short term. We anticipate the long-term EBITDA to go back to the 25% long-term EBITDA shortly after. In terms of gross margin, I think it's going to remain the same. It's all about the profile of investments, investing a little bit more into R&D, which is why we anticipate some pressure on the margin in short term.

Bernie, on your question about Go, so essentially, a lot of what we built into Go has already been integrated into several aspects of our product. And when we talk about the investments that we're doing in the product side as one of the pillars of this year, a lot of what we've taken from Go and how it can help buyers and sellers communicate more effectively, scope their work, the nuanced understanding of exactly what they need and what is the most suitable talent for the task is going to be integrated further into the product. So we're not we're not focusing on Go as a product by itself, but actually taking the assets that we've developed for that project and integrating them into the customer experience.

The next question comes from Jason Helfstein with Oppenheimer.

This is Chad on for Jason. It seems like you're taking one step back for kind of 2 steps forward. You're cutting a lot of costs out of the business with the restructuring. Is that having a bigger impact on revenue in '26 than maybe you previously thought? And then how should we think about OpEx growth in 2026? Do you have to invest more in the business or -- that's it.

Thanks for the question, Chad. So on the first question, the answer is no. The revenue is not impacted by restructuring. It reflects the ongoing trends on the marketplace, meaning lower end versus higher end, lower end seeing a decrease, higher end seeing an increase. And again, going back to our strategy, the entire idea is to double down on high end and make it grow faster and make it become a more meaningful contributor to our GMV growth.

We've been talking about this for a couple of quarters. What we've seen is an elevated sense of urgency to move faster, which is why a lot of the strategic -- the multiyear strategic plan that we have is all about that. And we said that once the high end is going to become a more meaningful contributor to GMV, GMV will go back to growth. And again, we're seeing this with double-digit percentage growth in transaction over $1,000. And so we're doubling down on that. So again, the reflection is just ongoing trends of what we're seeing in the low scale versus high end.

Then on the second part on the OpEx, I think that -- we think that the core business will continue to deliver a 20% plus margin. The portion of what we will reinvest into the business on the transformational work.

The impact of that is going to be around 2 percentage points. And I think it's also worth noting that due to recent appreciation of Israeli shekel to U.S. dollar, FX have added over $10 million of headwind on EBITDA -- on EBITDA guidance for the year.

The next question comes from Marvin Fong with BTIG.

Congrats to Jinjin and Esti. My first question, you talked about returning to growth in 2027. And I just wanted to understand that commentary a bit better. So are you expecting the high-value work to reach a majority of the marketplace by 2027, even in maybe a single quarter of 2027? Or when do you actually expect the majority of the marketplace to be high-end work? And then I have a follow-up.

Thanks for the question. So there's a GMV mix shift. High-value growth will continue to grow and become a bigger portion of the total market base. And this will lead to GMV inflection. And as we said both in the letter to shareholders and in the opening comments, that change is also going to allow us potentially over the year to start giving the market, the signals that we're seeing. Right now, the metrics that we're reporting is going to remain intact. But over time, we want to put more emphasis on what's strategic and what we feel is going to drive the sustainable growth of the business over the coming years. We haven't guided specifically for that, and we're not talking about percentage. And mathematically, even before it gets to the majority, it will drive GMV growth. But this is the plan, and we expect to see signals over the coming quarters to let us know that the investment there is actually accelerating the growth of that segment.

Got it. Okay. That's great. And then my second question, I would just like to double-click more. I think you mentioned or we've been talking about go-to-market and distribution channels. And so I'd like to talk about both enterprise a little bit more. Is there anything structurally you're doing to either the offering or the way you intake enterprises or approach enterprises that's going to change maybe a more formalized enterprise segment? And then in the shareholder letter, you talked about one of the measurable signs of progress would be one -- at least one AI native distribution channel contributing to GMV. I just would love to understand, is that a specific partnership you're developing there? Or do you just kind of expect the growth of those channels for at least one of them, presumably ChatGPT or Gemini to just naturally become a large distribution channel for you?

Thank you. The reason why we didn't call out specifics was deliberate. But it's -- the expectation is based on existing proof of concepts that we're having with AI model companies and enterprises, and we believe that when the product can deliver their needs at scale, this could be a very strong contributor for the growth. And so when we think about in the same aspect on our enterprise. And we've given some qualitative examples, again, both in the shareholder letter and in the opening remarks, and I've called out the example of the partnership that we have with an AI model safety company to provide domain experts who help identify vulnerabilities in foundational models.

And in another partnership, enabling enterprise to build AI workflow automation through white label solutions that allow them to deploy AI agents quickly and cost effectively. And again, this is all a part of the fact that AI enables many more businesses to build more and to build more ambitiously. Along with that building, there's a lot of support that they need. There's a lot of calibration and fine-tuning. There's very specific types of expertise that are required to take those products and those solutions and make sure that their integrity is high. They're coming to us with it, and we believe that we can create meaningful flywheel around these opportunities.

The next question comes from Nat Schindler with Scotiabank.

Yes, Micha and Ofer, can you help me understand -- I'm a little confused. I understand deprioritizing the lower end of the market. But I'm trying to figure out, is there any products you're not going to actually even be selling, any services from your service business that you're just not going to sell to them anymore? Because I'm trying to understand why you think revenue will get worse as the year progresses. So your revenue declines exceed as you go down as you go forward. I understand investing it might take a little while to turn -- to really build the enterprise business further. But I'm trying to understand what you're seeing in quarters 2, 3 and 4 that make them worse than quarter 1 on revenue.

Essentially, we're not killing any of our products. It's -- what we're mostly deprioritizing is the continuing optimization of these products in favor of developing the types of product experiences and the underlying technologies and infrastructure to address the higher-end project. So this in and of itself should not be a driver for decline. And it's not about the types of services or products that we're discontinuing because we're not discontinuing anything. We're just making sure that after the restructure, we have the vast majority of our resources to invest in the higher end and higher skilled types of services for the larger types of customers that are spending more with us and accelerate the growth that we're seeing there even further.

Okay. I still am a little unclear on them what signal you're reading that things are going to decline more in the back half of the year than in the front half.

So we've seen some decline in simple services across the board. There are areas where we're seeing slightly higher decrease than others. For example, within programming, we're seeing the simple side of programming, things like simple website building, accelerating the decline as a result of vibe coding and simplistic types of coding-related solution. But on the same side, we are seeing areas where we're seeing growth, like digital marketing is one of them where we're seeing nice growth in services. So the idea is not to discontinue anything. And by the way, we've called out these changes over the past few quarters.

As an example, writing and translation was heavily impacted by AI, down 20% over the year. And we've seen this for a while. We've seen the same under the vertical music and audio is also impacted, but slightly less in the teens range, primarily because voice over is a meaningful portion of the music and audio vertical. So there's anecdotal areas. Again, the decline that we're seeing there is mostly in the very simplistic low skill related types of services. And this is not our focus now. That transformation is going to continue happening, and that's fine. The function that we're focusing on is the one-of-a-kind way for us to deal with high-end, high-value transactions, utilizing all the data that we've collected over the past 16 years and the incredible talent bench that we have with us today.

The next question comes from Josh Chan with UBS.

Sorry about that. Apologies for that. I guess maybe following up on the prior question a little. I guess if you take your full year guidance, it's less than 4x your Q1 revenue guidance. And so I guess what's the -- conceptually, why does the rest of the year kind of step down from Q1, I guess? What are you seeing that's worse? And then my second question is, is there a way for you to frame for us what free cash flow can be in 2026, maybe from a conversion perspective versus EBITDA, something like that?

I think on the first part, I think the combination of the trends we are seeing in Q4, together with the transformation that has been discussed is creating some uncertainty in terms of the 2026. And under those circumstances, we are guiding for a wider range.

Yes. And on the second one, the free cash flow, it largely follows EBITDA. And we've guided for a midpoint EBITDA of 18%, 20% plus on the core business and 200 basis point impact from the investment that we're doing in the restructuring.

The next question comes from Matt Condon with Citizens.

First is just on, I think you said in the shareholder letter, you're building the marketplace for more reoccurring work. Can you just talk about the products and functionalities that you need to launch to enable this reoccurring nature of work? And then I wanted to ask a follow-up on an earlier question is just given where the stock is trading, just can you talk about the prioritization of buybacks versus M&A?

Thanks for the question. So essentially, it's all about putting trust and quality at the forefront. And we're meaningfully upgrade our data infrastructure, matching algo product in order to achieve that. And those are the most important components of being able to optimize for recurring work. And also the fact that we're modernizing our platform allows the usage, as I said, of not just human customers, but also agents within the platform. A lot of it is about how we build this infrastructure and how we ensure the quality and the happiness of the entire fulfillment cycle, which have been one of our biggest moats.

The fact that the work actually happens on the platform is being documented and being tracked, allows us to understand the right path or route in which work is done to be able to actively intervene in cases where it's less than great. All of these are indicators from our data for recurring usage of our platform. The second one was on buyback. We have a continued disciplined and balanced capital allocation, investing growth while continuing to utilize our buyback authorization to return capital to shareholders. As of December, there's $67.5 million left on our authorization. And as Ofer mentioned earlier, we'll continue to be optimistic on M&A -- opportunistic, I'm sorry.

The next question comes from Brad Erickson with RBC.

This is Audrey, on for Brad. First, new business formations has been growing pretty solidly, but that doesn't seem to be lining up with parts of your business. Is that just because there's really no connection there? Or what is the disconnect you would say if there is one? And then second, in the new world of changes at the top of funnel, how big should SMB as a percentage of revenue or marketplace GMV relative to where it's been in the past? Any reasons why it should be structurally higher or lower?

Thank you for the question. Business formation only impacts a small part of our catalog that's focused on the very early-stage companies. So I wouldn't read too much into that aspect or the correlation between the 2.

I think that regarding the second part -- don't anticipate any change -- SMB is part of the GMV.

The next question comes from Rohit Kulkarni with ROTH Capital Partners.

A couple of questions. One is just on this doubling down on high-value things that you'll be doing going forward. Where do you see the heaviest lift next 12, 18 months? Is it you need to attract more supply where that is capable of doing these high-value things? Do you think you already have the supply? Or is this a function of building the product and then getting more and more high-value buyers?

And then kind of as you do this transition into more higher value and better matching, is there a scenario where the services revenue or the attach rate of services to marketplace has a different algorithm given higher-value gigs may not need as much ads or there may not be any need for as many subscriptions to sellers who are trying to get signed up for more long-term contracts as such. So how do you feel longer term that mix between the core marketplace and value-added services could look like versus where we are today?

Thanks for the questions. So on the first one, it's really very much around the data infrastructure, the matching algorithm that prioritizes quality and trust. And it's really all about the customer satisfaction and retention. In terms of talent, it differs between categories, and it changes over time because we see more and more types of skills coming in demand. In most cases, it's very easy for us because we've been doing this for 16 years to make sure that we have the right talent. We haven't seen any pockets of shortage in talent. But in any case, we're always equipped to fill any shortage in a very short amount of time.

As the go-to-market, we see that an opportunity, expanding channels from existing channels into AI native channels, building the enterprise partnerships, as we've talked earlier in the call and targeted growth loops for specific use cases. As to your second part of the question, services revenue will continue to be a growth driver for us this year. That said, the pace of growth will be more moderated this year because most of the efforts this year will be foundational to improve the marketplace moat and enable high-end flywheel. That said, service revenue has a long growth runway long term as we enable every aspect of the talent need and lots of service expansion opportunities down the road.

This concludes our question-and-answer session. I would like to turn the conference back over to Micha Kaufman for any closing remarks.

Thanks, Megan, for moderating the call today and for everyone who has joined us this morning. Have a great day.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.

Thank you for standing by. At this time, I would like to welcome everyone to the Fiverr Third Quarter 2025 Earnings Conference Call. [Operator Instructions]

I would like to now turn the conference over to Jinjin Qian. Please go ahead.

Thank you, operator, and good morning, everyone. Thank you for joining us on Fiverr's earnings conference call for the third quarter that ended September 30, 2025. Joining me on the call today are Micha Kaufman, Founder and CEO; and Esti Levy Dadon, EVP, Finance.

Before we start, I'd like to remind you that during this call, we may make forward-looking statements and that these statements are based on our current expectations and assumptions as of today, and Fiverr assumes no obligation to update or revise them. A discussion of some of the important risk factors that could cause actual results to differ materially from any forward-looking statements can be found under the Risk Factors section in Fiverr's most recent Form 20-F and other filings with the SEC.

During this call, we will be referring to some key performance metrics and non-GAAP financial measures, including adjusted EBITDA, adjusted EBITDA margin, and free cash flow. Further explanation and a reconciliation of each of the non-GAAP financial measures to the most directly comparable GAAP measure is provided in the earnings release we issued today in our shareholder letter, each of which is available on our website at investors.fiverr.com.

And now I'll turn the call over to Micha.

Thank you, Jinjin. Good morning, everyone, and thank you for joining us. We delivered another strong quarter with solid performance across the business. In Q3 '25, revenue grew 8% year-over-year, and we achieved a record high adjusted EBITDA margin of 22%. This is a clear reflection of our disciplined execution and the inherent leverage in our market-based model. Over the past several years, we've consistently prioritizing moving upmarket and investing in product innovation to support more complex use cases and larger customers.

Q3 results clearly demonstrate our success on both fronts. Spend per buyer increased 12% year-over-year, our strongest growth rate since the COVID era and off a much higher base. Not only are we seeing wallet share expansion across the broader buyer base, but more importantly, thanks to the adoption of dynamic matching and managed services, we are witnessing strong growth among projects that are significantly larger than the average marketplace transaction.

In Q3, GMV for Dynamic Matching grew 22% year-over-year with 15% of job briefs having a budget of over $1,000 and an average order value of $2,200. Managed Services is capturing even larger and more sophisticated engagements with a minimum budget of $3,000. In Q3, Managed Services GMV grew 65% year-over-year with average product size reaching $17,000. The success of these offerings marks a meaningful evolution in Fiverr's value proposition. We're no longer just a platform for fast, lightweight freelance tasks. We are increasingly becoming a trusted partner for businesses executing highly specialized multistage projects that often require depth of talent and orchestration.

Another area where we are seeing tremendous growth is AI-related services. As AI is increasingly reshaping how work is delivered and being implemented across industries, demand continues to surge in areas such as AI agents, workflow automation, and vibe coding. Fiverr freelancers have become an essential partner for SMBs looking to turn AI, from potential into performance. This demand is directly reflected in the Programming and Tech vertical, which grew 14% year-over-year in Q3. We believe that this AI transformation cycle mirrors an early stage of the digital transformation and could provide a multiyear tailwind for broader tech investment. To lean into this secular tailwind, we are doubling down on our investment in AI-related categories from growing specialized talent communities and launching tailored AI solutions to expanding our go-to-market channels through strategic partnerships. Our ambition is to position Fiverr as the go-to destination for finding top-tier AI talent and deploying applied AI solutions.

Despite a macro environment that remains uneven, we're seeing positive signals and gaining market share. Labor markets continue to show mixed trends and broader hiring recovery remains elusive. However, our growth strategy, which centers around upmarket expansion and AI enablement is built on long-term macro-agnostic trends. We believe these are the right bets to get us back on track for GMV acceleration regardless of macroeconomic scenarios.

In that context, we announced a strategic restructuring in September to streamline our organization, sharpen our product focus, and accelerate our evolution into AI-first company. This means accelerating investments in building an AI-native team, upgrading our tech infrastructure to drive faster AI integration and operational efficiency and reimagining our marketplace with an AI integrated experience. From a product perspective, this transformation is anchored on 4 key pillars:

One, strengthening our go-to-market execution. We're expanding our generative engine optimization, GEO, capabilities, integrating our catalog into native AI channels and building AI-powered catalog management systems. We're also investing in partnerships that drive growth across AI-related verticals.

Two, building the next Gen AI-powered buyer experience. This includes expanding LLM-powered workflows across the buyer journey, advancing our Know Your Customer, KYC, capabilities through data and product innovation and investing in customer success to deepen trust.

Three, evolving our matching technology. As we serve more upmarket clients and more complex projects, we're transitioning from traditional search to agentic matching, delivering a recruiting-like experience that surpasses human performance through deeper data, richer context, and advanced reasoning.

Four, investing in talent and the talent community. Talent is at the heart of the entire marketplace experience. In a world where AI is rapidly transforming how work is done, our priority is to build a high-quality, trusted talent ecosystem. This means supporting human-in-the-loop workflows, creating pathways for professional growth and deepening our commitment to long-term community engagement.

I am truly excited about the opportunities ahead and the strength of the roadmap we have built. As we enter the final stretch of the year, we remain laser-focused on execution. Our momentum in AI and upmarket expansion gives me confidence in the foundation we've built. I look forward to sharing more about our 2026 roadmap in our next call.

With that, I'll turn it over to Esti.

Thank you, Micha, and good morning, everyone.

We delivered a strong third quarter, with both top and bottom lines exceeding the midpoint of our guidance. Revenue for the third quarter was $107.9 million, up 8% year-over-year. We also achieved record adjusted EBITDA and adjusted EBITDA margin. Adjusted EBITDA for Q3 was $24.2 million, representing an adjusted EBITDA margin of 22%, an improvement of 260 basis points from a year earlier. We continue to generate strong cash flow, with free cash flow totaling $29.1 million in Q3. The strategic restructuring, combined with our continued discipline in expense management, contributed to strong profitability and robust cash flow generation. As always, we remain focused on balancing between growth and profitability, while maintaining discipline in capital allocation.

Q3 saw solid performance across both our Marketplace and Services segments. Marketplace revenue was $73.6 million, driven by 3.3 million active buyers, $330 in spend per buyer, and 27.6% marketplace take rate. Within the Marketplace segment, we saw strong momentum driven by the tailwinds in AI-related categories and the success of our expanded Managed Services and Dynamic Matching. These channels continue to fuel higher-value, complex projects, which in turn result in higher average transaction values and increased in share of customer spending. We continue to believe the structural tailwinds within the Marketplace segment, particularly around AI and upmarket adoption, will help offset broader economic headwinds and serve as a sustained growth driver.

Services revenue was $34.3 million, representing a year-over-year growth of 40% and accounting for 32% of our total revenue in Q3. The upside was driven by Fiverr Go increasing adoption of Seller Plus, which saw 20% year-over-year growth. Fiverr Ads maintained double-digit growth as a result of ad load expansion, and AutoDS benefited from enhanced synergies with Fiverr and continued success with the Shopify partnership. Looking ahead, we expect Services revenue growth to moderate as we lap the 1-year anniversary of the acquisition, but to maintain healthy double-digit revenue growth. We continue to expect Services revenue to represent a little over 30% of total revenue for the full year 2025.

Now onto guidance. For the full year 2025, we expect revenue to be in the range of $428 million to $436 million, representing year-over-year growth of 9% to 11%. We are raising our full-year adjusted EBITDA guidance and now expect it to be in the range of $88 million to $93 million, representing an adjusted EBITDA margin of 21% at the midpoint. For the fourth quarter of 2025, revenue is expected to be between $104.3 million to $112.3 million, representing a year-over-year growth of 1% to 8%. The wider-than-normal revenue guidance for the fourth quarter reflects the elevated uncertainty in macro environment with mixed signals. Adjusted EBITDA is expected to be $23.9 million to $27.9 million, representing an adjusted EBITDA margin of 24% at the midpoint.

During Q3, we announced a strategic restructuring plan, which resulted in a streamlined headcount and enhanced operational efficiency. These efforts contributed to the increased adjusted EBITDA guidance in Q4. While the pace of EBITDA improvement in Q4 should not be viewed as a steady-state cadence, profitability, margin expansion, and cash flow will remain key priorities for us, even as we redeploy some of our cost base savings into selective, high-impact investments in AI, and upmarket initiatives in 2026. We remain committed to our accelerated schedule to reach the long-term adjusted EBITDA margin of 25% in 2026.

With that, we'll now turn the call over to the operator for questions.

[Operator Instructions] Your first question comes from Ron Josey with Citi.

This is Jake on for Ron. Micha, I wanted to double-click on how you're reimagining the marketplace to be AI first, specifically the pillars around evolving the buyer experience and improving matching. Could you just double-click and help us better understand your vision here and why you believe Fiverr is uniquely positioned to be AI first?

Thank you for the question. I should clarify that Ofer is with us on the line. He's having a sore throat. So this is why Esti was coming with the opening remarks. But in case we have questions, he will be happy to answer them. As to your question, look, I think what AI is giving us is the ability to change the way people express themselves. And you see that in the market, the way search is being augmented or replaced by prompting, it gives us an opportunity to extract more -- a more accurate representation of the actual need from the customer, which in turn gives us better tools to be able to accurately match in a very precise manner the right expert to the right mission.

The same applies for more complex projects that sometimes require multi-talent and sometimes require an orchestration of those multi-talent into the end result. So this is just one example, which is very robust because the primary function that customers are using is the browsing and searching. And this is the most fundamental thing that we can use AI for. The same applies for the rest of our solutions like the dynamic matching, as an example, the project management, and so forth. And I've alluded to it in my opening remarks with everything I said about the matching capabilities, the Know Your Customer aspects. And we're already reaping the benefits or starting to reap the benefits of this being able to deliver much better matching to our customers that, in turn, result in larger types of projects and with higher satisfaction.

Ofer, I hope you feel better. Just one quick follow-up for you or Esti. Kind of given the wider 4Q revenue guidance, could you just touch on the key assumptions that would get you to the low end versus the high end, maybe specifically around GMV trends?

So this is Ofer. Thank you for the concern. The assumptions for the remainder of the year is that the revenue coming from Services will continue to grow. While the revenue coming from Marketplace dependent on GMV trend that currently seem to be flat, might decline by single digit. The assumption is that the trend that we have been seeing in the last quarter will continue into the fourth quarter. And based on the high volatility of the market in the last few quarters, we kept the guidance range wide to take into consideration that.

Your next question comes from Jason Helfstein with Oppenheimer.

Obviously, the spend per buyer increase was nice. I think you highlighted that this really wasn't from like the SMB, but just the ability to move into more advanced projects and does show you more than a company. I guess how does the reorganization tie into the ability to ever get back like the SMB opportunity? Like is the assumption now that you are -- you've reorganized the business to focus on higher-value jobs, and hey, if SMBs ever come back, that's great, but like there's no assumption. Like do you think there's like just -- it's been so long since we've seen SMB demand that we just shouldn't assume that it ever comes back? Or any broad thoughts about that?

Right now, since the dynamics in the macro economy that we've seen hasn't changed materially, then we don't assume those changes. And while we see the Fed has started to lower interest rates, which could be constructive for SMBs, we also continue to see weak job data across full-time and staffing sectors. So the macroeconomic conditions are still highly uncertain, to say the least.

So when we think about our guidance or how it's being made off, it really assumes no change in the macro front. We talked about it in the full year basis that services will be a little over 30% and of the 2025 revenue. So that will give you some idea of how we think about the marketplace versus services revenue. But overall, we continue to expect the marketplace revenue to be flat or low single-digit decline and services revenue to exit the year with double-digit growth. And that assumes no improvement in the SMB side.

That said, when we look at everything that has to do with the upmarket, meaning the larger types of customers, and the larger types of projects, this is where we do see an improvement, absolutely. And that contribution is very noticeable. And as much as that portion of the business becomes larger, the contribution is going to be larger, which means that by definition, the return to growth in active buyers is going to happen, period. We've been saying that for multiple quarters, and we're seeing this. So it converges to that point, okay? That is the assumption. That's the framework. And this is what we're seeing happening in reality.

Your next question comes from Doug Anmuth with JPMorgan.

I have 2, Micha. Can you just talk, I guess, first, just about the key investments you need to make in '26 to transform into an AI-first company? And how should we think about timing as you kind of progress along this shift? And then, I guess, secondly, can you talk more about the drivers of spend per buyer in the 12% growth? Kind of like beneath the hood, what are you seeing in terms of changes in types of projects and how buyers are really engaging with the platform?

So thanks for the questions, Doug. So the investment that we're doing is, one, on talent. And I think that this is true for everyone. Finding AI natives on the talent side within the company is one area of focus. The second is the improvements that we're introducing to our infrastructure. So some of the benefits of being able to use new development, coding, design, marketing allows us not to just put more people on problems or building things on our infrastructure, but actually finding new ways of moving much, much faster without paying the price of working on a 15-year-old infrastructure.

The third is the marketplace experience. On everything we do, and we've highlighted things like dynamic matching and project management and orchestration. And we've highlighted also the aspects of having a much more nuanced and much more accurate matching technology and KYC. And I've mentioned those 4 pillars, which is the go-to-market with the investment in LLM engines, GEO and partnerships, the buyer experience, the matching and the talent. So across all of these areas that I've mentioned, these are the areas that we're putting focus. And the approach is AI-first mentality, meaning we want to make sure that we maximize the new possibilities that AI gives us both internally in how we work and how we develop and how we execute, but both in how we can make AI more involved in our core business to make that core business better.

As to your second question about the spend per buyer, I think that this is -- when we think about spend per buyer, there are catalysts for it. And some of it is the ability to identify categories that are now growing, some of which thanks to our move upmarket and some of which is because of the technological transformation of AI. Now I've given some examples for it. Programming and tech vertical is one of them, which has a much, much higher project size, dynamic matching and its contribution and the fact that 15% of them are from briefs that are above $1,000, managed services that have grown 65% year-over-year with an average size of $17,000. So these are very, very different from our average transaction size of services. And we continue to drive very healthy deal flow on managed services. So the nature of these projects is very strategic, not just tactical. And there are a few examples, if you are interested in more examples in the shareholder letter.

[Operator Instructions] Our next question comes from Matt Condon with Citizens.

Just with these product catalysts across AI and moving upmarket and potentially decoupling you from the macro environment. Just what is your confidence level that these products can actually return the marketplace business to growth in 2026 as you just more deeply integrate them? And then my second question is just on AI displacing some of the commoditized jobs at the lower end of the market. Just have we seen those plateau at this point and become less of a headwind going forward? And just the placement of those types of jobs, is it less today than it was, say, a year ago?

Thank you for the question. So as these AI-driven products or needs grow, they grow much faster than the average. And so -- and the more they become a bigger portion of the total, they are driving us to change direction and go back to growth. We're seeing that. Still, it takes time for those types of customers and those types of projects to become the majority of our business. But as they grow, we are, by definition, going back to growth. And we see that, we see that on a constant basis, and you're seeing the numbers grow every time we meet here every quarter and they become larger and larger.

Today, transactions over $200 is already the majority. They're over 50% of our market base, and they're growing double digit. And transactions over $1,000 is growing in the 20s year-over-year and more than 10% of the marketplace already. So this gives you some idea that they are already meaningful in the marketplace, and they continue growing, and this is where we draw our confidence from.

The second part of your question about the job displacing, I've addressed that many times. The jobs that are being displaced are the very, very simplistic types of jobs. All of us are using AI for 2 years plus. We know its limitations. There's a lot of limitations to what AI can do to its accuracy and to its quality. And customers understand that as well. And they are -- we've seen a lot of our customers and talent, by the way, using AI. But the more they use it, the more they understand its limitations and their ability to trust the outcome to be production-ready and business ready.

And so the things that we have seen being displaced are very low skill types of services. Those types of services have been very small in size anyway. So the fact that they are being displaced is not a big deal. The more we invest in larger projects, the more we grow. That's the bottom line. And this is where AI doesn't replace human beings. It doesn't replace human talent and high skills.

Your next question comes from Josh Chan with UBS.

I just have 2 questions today. The first one is on your comment about macro uncertainty. I was just wondering, the macro has obviously been choppy for a while. So are you seeing anything different now than before that kind of led you to make that comment and that wider guidance? And then the second question is on, could you just talk about the phasing of the restructuring benefits? To what extent some of the benefits are coming into Q4 and how that kind of layers into the rest of 2026?

This is Ofer. On the first question, on the contrary, there is no change in macro, which is why we have kept the guidance pretty wide. And on the second question, definitely, there will be a bigger impact after restructuring into Q4. But as we look into next year, we do plan to fill up the lines with some of the needed talent so that I would expect next year to improve in terms of EBITDA, but not to the same cadence as we are expected to see as of Q4.

Your last question comes from Marvin Fong with BTIG.

I hope you feel better as well. Question, I don't want to be -- cover ground that we previously did, but I think I'd like to ask it just kind of on a category basis, you called out the 14% growth in programming and tech. I just would like to know if other major categories are, how are they benefiting from AI? Are you seeing the same trends? So perhaps you could comment on like design, creative, I think that's another large category for you. But any other major categories you'd like to kind of call out and how AI might be a tailwind for that?

And then second question, just on the move upmarket, obviously, really great traction there. And I was wondering if there were other unlocks that you can do? Are you satisfied with the product suite or in the next 12 months, what are some new features that you might be able to launch to address other parts of the ecosystem. So for example, 1099 versus W-2, anything along those lines would be great.

Thanks for the question. So to highlight some of the areas where we're seeing growth. So programming and tech is growing fast, and it's becoming a very meaningful category with about 20% of our business. Alongside programming and tech, we have digital marketing, video and animation, which are also growing very strong. Some of it is due to AI and the possibility of bringing highly skilled people that know how to extract the most out of AI, which is kind of the case I mentioned where you have customers sometimes trying to use AI, understanding its limitation and their limitations is not being experts in how to make the most out of it, and they come to us in these cases, and we're seeing this as a very prominent case in these verticals.

And also the nature of how customers come to us is very different from 2 years ago or even a year ago. So they -- in many cases, they come more educated. They do a little bit of work on their own. They're not clueless. They can better express their needs, which also changes the basic function of what we do, which is less of explaining them what they need, but more trying to address that need by giving them a really strong and very accurate high-quality match.

As we think about the move up market, there is a lot to do. And I've already highlighted both in my opening remarks and in some of the answers and in the shareholder letter, some of these areas. But just to reiterate some of them, today, many customers are already enjoying dynamic matching and managed services, but many more don't yet know that Fiverr can do these projects that are in the tens of thousands of dollars. So there is a lot of opportunity that we can unlock there.

At the same time, there is the LLM channels, which are another example of areas to invest in where the traditional search or the fact that Google was the Internet up until a few years ago, and now they're not playing there alone. There are also other ways to explore the Internet and a lot of it is going to LLMs, creates a challenge, which we prefer to look as an opportunity. And we're seeing how the LLM channels at the top of funnel is contributing more and more into the top of funnel traffic, which also makes us more motivated to continue investing along those areas.

And lastly, I would say that is maybe to reiterate what I said before, which is customers are more accurate in how they share their needs. They're more explicit, which is good news for us because now we have more to work with and be able to match them with this incredible base of unbelievable talent that we have on our platform so that we can address their needs.

That concludes our Q&A session. I would like to now turn the call back over to Micha Kaufman for closing remarks.

Thank you, Moore, again for moderating this conference. And thank you, everyone, for participating. Wishing you a great day, and talk to all of you soon.

This concludes today's call. Thank you for attending. You may now disconnect. And have a wonderful rest of your day.

Okay. So I think in the interest of time, we're going to get going on our next -- you can go right here -- at our next session, it's my pleasure to have the team from Fiverr here at the conference with a fireside chat with Micha Kaufman. Thank you so much. It's good to see you again, always.

Look, I think to set the stage, and I feel like I ask you this whenever we get the opportunity to do this. Maybe talk about the journey the company has been on and the evolution of what you're trying to build for the company. It has changed a lot over the last 12, 18 months. And I think with the advent of AI, you've been very much at the forefront of trying to move the company to platform, the product in a certain direction.

So why -- just reflect a little bit on the journey you've been on, and then we'll get into the pathway forward as we progress?

So essentially, if you look at the history of the company, we started the marketplace for very distinct discrete types of services. Most of them are micro services or very small services that was a part of our go-to-market strategy.

We started with micro business, SMB. And over the years, we started extending the capabilities of the marketplace and then turn the marketplace into a full-fledged platform. And in recent years, what we've been doing is really taking this model that we pioneered where people can just come online with their teams, go through a route through directory or marketplace and services in just one-click order.

As we were starting to go upmarket, we realized that in some cases, the marketplace is a great solution. In other cases where there's more complex types of projects and needs where you need orchestration, sometimes you need assembly because it's complex [indiscernible], then we needed additional tools. In recent years, we've introduced this business suite. And in recent quarters, we've reported that the portion of the cohort of the larger businesses that are actually coming to do more with the platform is increasing dramatically.

And I think that this went very well with what was happening with AI, starting 2.5, 3 years ago where the micro services. Some of them are no longer needed because they could be automated. But at the same time, you need very high skilled people that know how to us AI. And so we've been focusing more on that and more on ensuring that we build the AI solutions and technologies into our platform so that they are integrated in the workflows of the professionals that are actually delivering the project.

That portion of our business has been growing very nicely. We keep reporting and you see some verticals mostly around technology, and marketing, workflow, agent development and so forth, growing 10x. So this is really changing the function. And we're using AI to build a better platform as we do it so that we can provide better value, better turnaround times for our customers. Everybody wants to integrate AI, every business wants to take these new tools to their maximum. And this is why they are coming.

So from our perspective, it's also changing the type of supply that we have on the platform to bring more of these AI native excellence that can actually help organizations through this confirmation much more.

Okay. So I'm going to fill a buster for a second because I think we're going to swap out your microphone.

Yes. Okay. So I'm going to fill a buster. While we do that. And maybe reflect backwards again before we go forward because there was initially a very negative reaction to AI with respect to how the economy would evolve around freelancers and this type of work. And I think the narrative in many ways has evolved over the last couple of years.

Talk about what you're most excited about to put AI to work across the platform of the products when you think about the incremental drivers of growth going forward?

Yes. I think that there is some kind of awakening from the early thoughts around AI. And so one, there's this realization that while AI gives everybody the superpowers to be more efficient, to do more, to do faster, it's not really creating any competitive edge because everybody has AI.

So if everybody has AI, no one has an advantage. And so the advantage always goes back to the people that actually take use of it. The most incredible things that you see being created with AI is not AI. It's the people that actually use AI in the most creative ways, in the most effective way.

The other part is every organization says that they're doing the transformation to AI. But in most organizations, this means providing ChatGPT to their employees. This is not the usage of AI, okay? And so to really integrate AI into how companies, our organizations work, unique people that can actually orchestrate this. And the I think that what a lot of our organizations realize is that the amount of AI native that every organization has internally is very, very small. And so these types of skills are very solid. So that's where we're concentrating.

So we've done the same with our team, making sure that we're optimizing the way we were across the organization, the way our legal department work, the way our financial department works, the way we do customer care, the way we develop and write code. And we see a tremendous opportunity in actually helping other organizations do this because we have all this talent on the platform.

And the other part is the fact that because of AI, many more people in our organization are building more things. So on the one hand, it creates more issue of signal to noise because everybody is building, but who's going to know about the thing you're building. So they need to market it. But also those who are building stuff at some point, get stuck. So you can do bytecoding, you can build maybe 60%, 70%, 80% of what you want with bytecoding.

At some point, you're going to get a cloud database authentication issue. And if you don't know what that means, you're done. And so even helping companies, helping organizations finish their building is another huge opportunity for us. So these are areas where we've been investing our time.

Okay. I want to follow up on this idea though because you interjected it before as well. This idea of the gigs that you're seeing on the platform, how differently are trends tracking when you think about complex gigs versus non-complex gigs? And how AI might be affecting that mix over time as well?

Yes. So we talked about this before, and I said that what we're seeing from our data is that the areas where we're seeing technology displace manual work is mostly in the very simplistic type of services.

Now we give this example because it's easy to understand, up until, I don't know, 3, 4 years ago, removing background from photos has been something that you have to hire somebody to do it. Today, you don't. But the removal of background has been very low skilled always. And these were services that were mostly sold for $10, $20.

So these are the areas where you see technology replacing stuff much, much faster, but it's creating a host of other professionals that didn't exist 3 years ago. People that are designing specialized AI agents that didn't exist or people that help optimize workflows using AI. That didn't exist as well.

So we're seeing the pace of these more complex, more new ones, more high-skilled services growing much faster than the simple tasks are being displaced, plus the simple task are low AOV anyway. So for us as a business, AI is a positive effect more than it is a negative effect.

Understood. I don't know if you can provide any progress update on FiverGo, an initiative you launched, how we continue to scale? What do you think this might mean in terms of opening up? Why their pockets of addressable market for the platform over the longer term?

Yes. So the idea of FiverGo was to start experimenting with integrating and providing tools that impact the way our freelancers work to make AI more integrated in the way they do work on the platform. .

So just as a reminder, one of the things that makes Fiverr very unique is the fact that it's not just matchmaking, but it's actually the platform on which the work is being done. Anything from finding and matching to contracting, to invoicing, to payment, to communication, to delivering work on the platform, to tracking revisions, all the way up until the service is done and rendered on the platform.

And so we saw this opportunity of actually creating these highly specialized tools for our community so that they don't need to use multiple tools is their workflows because we can integrate those tools into our -- into the platform. And this has to do with the way they create, but also has to do with the way they communicate.

So we take a lot of the back office admin of running, prioritizing work, to all the way to actually how they fulfill the work. And we've seen that once we do this, first and foremost, our customers are getting much better and much faster service and turnaround. And this increases conversion. And so we've introduced these tools that are changing the game for freelancers. They're reducing the type or the portion of work that they don't like doing anyway, which is answering the same questions that they've answered 400 times again to a new customer.

And just the fact that they are always concern about missing the next opportunity. So there they keep their alarm on when they go to sleep, if they have a new notification and no, they don't need to do that. We have an agent that does all of that work for them. So this has been going super well. we've seen more impact actually on the higher quality supply because they get this and they get that essentially at the end of the day, it gives them better ROI. And so they use it.

And we do the same with generative AI into how we can use generative AI to help them deliver much more professional work and do it much faster. So FiverGo has been the beginning of a much -- of a very long journey that we have in plan. And actually, the other part that makes this very interesting is, when you build the company over 15 years, you build this tech stack that becomes your legacy operating system.

At some point, when technologies make this leap, you have the opportunity to actually start building things from the ground up. So instead of building stuff into our existing tech stack, you can actually build new technology and then integrate it into how your work -- into how your product is set up.

And FiverGo was the first real experiment that we've done with doing that. And I think that one of the most exciting things for me and I was -- this was my project, was this allows you to get back into startup mode and really builds up super, super quick.

Yes, it is -- it does require more of an AI-native team to do it. But this has been one of the most exciting things for me, sleeping on the floor of my office for months, which sounds terrible, but it was -- for me, that's kind of -- that's my element. That's the part that I love in my job.

Understood. Okay. I do want to talk a little bit about macro, because you guys have been talking about some of the headwinds that the business has been facing for quite a while. And it seems like what it really translates into is type of client growth and type of client budget.

So maybe just level set for what you're seeing in the macro and what distinctions or elements of that you want to call out depending on type of client, type of spend?

Yes. So first of all, if you look at professional staffing, and this is public indexes, that the U.S. government provides. We see there the professional staffing is a little bit depressed. It's been like minus 15%, sometimes minus 20%. So mid-teens negative for the past 2 years or so. If you look at our GMV, our GMV is around 0, which means that we're offsetting on a minus 15%, which is not easy in it of itself.

This creates 2 opportunities for us. One is it gives us more supply because there's more people that are looking for opportunities. This is great because it allows us to be pickier on who do we want to join the platform. And so this gives us a massive buffet of talent that comes to us.

The other thing that we've done is we really started focusing more aggressively on the larger types of customers. And the reason is that if you look at macroeconomic because of the high interest rate and the inflationary environment, that portion of macro influences SMBs more than established businesses. The reason is SMBs need to borrow money to grow and borrowing money is just more expensive.

And so they're having a harder time. Established businesses have to grow, but they have returned capital. So while they're a little bit more cautious than usual, they need to continue investing. So we've shifted a lot of our attention in acquiring these businesses. These businesses are more expensive to acquire, but yet at the same time, even in their first transaction, they spend much more, their lifetime value is much higher and their ROI is great.

And so what we've done is we've shifted our focus, and you can see this on our active buyers. So active buyers is shrinking a little bit because we're investing in buying more expensive types of customers. But if you look at spend per buyer, spend per buyer is growing really fast. The reason is that, that cohort becomes bigger, becomes larger within the entire cohort of our customers.

So maybe the follow -- the natural follow-up is if you've got this level of engagement in enterprises, and you've got some strategies that are continuing to move you up market, in a better macro environment, how should we be thinking about mix of SMBs and enterprises, n ot to put you on the spot? But just -- what does that mean for the business if you get an actual healthier macro environment overall?

Our assumption, given the 15 years of history is in a better macro, you're just going to see much higher growth. And the reason is that the majority of our customers are coming to us organically. And so if you see some improvement on the SMB side, there -- the majority of them are free. It's just free business for us. .

So you should see increase in active buyers coming from that portion. And we can always go -- I mean, we're generating a lot of free cash flow. So essentially, if we see opportunities to grow, both invest. And on the other segments, the other segments do business even now.

So if there's a better macro, then we should see an increase in that portion of the business as well. I think Fiverr has been super sensitive to macro and we've been calling out macro ahead of most companies because the customer base of Fiverr is so large. We can actually call those trends.

Right now, we're not -- we're still not seeing it. I don't know -- I don't know if the Fed is going to decrease interest rate. If that's going to immediately translate, but give me a call, I'll let you know. And I don't know if it's going to be enough...

We'll talk to you in early November.

Yes. But it is sensitive enough. So if we do see interesting trends that we can call trends that are not just short spikes here and there. I think that we're going to be able to talk about this.

It's going to be a much more convenient environment for us for growth. And it should, again, it should be said that as a company, we are prioritizing growth. This is what we're here for. And because there's a huge market capture and we're doing it very efficiently. I mean we're now growing double digits while providing 20% EBITDA and generating a lot of free cash flow. So we have we have the tools to maximize growth if the opportunity is there.

Got it. So maybe 2 follow-ups here. In terms of that upmarket strategy, if you were to see that type of recovery, do you still think there'll be a more large customer, large enterprise mix dynamic that will play out? Are there lessons learned from moving upmarket that you think sustain for the longer period of time?

Absolutely, yes. And by the way, even if macro doesn't change, the larger that cohort is within our business, we are going to go back to growth because that segment is growing and the larger it becomes the more impactful it is on the overall growth. So absolutely, yes.

And because we're the most highly traffic platform in our space, what we've identified -- and we talked about this before, but we're getting better and better at actually capitalizing on it, is that we have a ton of these businesses coming to us organically. And it's just the ability to identify these customers as fast as possible.

Because we're a platform and still the majority of our business is being generated on the marketplace side. We realize that the faster we do KYC on our customers on traffic that comes to us, the better we can route the right customers to the right solutions that we provide.

And so a lot of the customers that we've called out in letters to shareholders, like publishing company that comes to us because they need to edit hundreds of books. These types of customers, we've been able to entertain because we've identified them very early in the process and diverted them into a much more white-glove service that can actually run a complex project for them. So it was very easy.

So the better we become at identifying these customers early on in the cycle and ensuring that we route them to the right solution, the more we can capitalize on it. So this is definitely -- these are definitely key learnings that we're taking with us regardless of how macro turns out.

Okay. I know there are specific areas -- I know you've talked about a few, but are there specific use cases driving the strongest traction in sort of complex high-value projects today? Is it all AI implementation? Or are there other areas?

I mean, look, largely, if you think of Fiverr as a platform, there is -- you can find experts in every aspect of a business from the very initial stages of a business that could be a new business or an existing business, which is ideation, consulting, whatever. To the second phase, which is building. To the third phase, which is marketing. To the fourth phase, which is growing.

And so you see this across the board. And we've seen specific categories in which that we've seen booming. We've seen that across not just AI related, which is agent design or data or workflows. We've seen that in areas like drop shipping, we've seen that in areas like book publishing.

So there's a there's a mix of different verticals that are experiencing natural growth. And what we're doing with these is we're doubling down on every aspect of the business that is growing, we're just doubling down on.

So sticking with that theme of sort of doubling down. One of the other things we've talked about on public earnings calls would just be the continued scaling of Fiverr Pro. Talk a little bit about what you've learned there? And sort of how we should be thinking about that as a contributor for growth given the scale and size of some of those contracts?

Yes. So essentially, Fiverr Pro was a part of the strategy of going up market and understanding that when you start working with larger types of businesses, their needs is much more bespoke and nuanced. And sometimes they require either super high-skilled people. And in some cases, it needs agencies or the orchestration of multitalent to achieve a certain project.

And so this was a part of this understanding that we need this highly vetted, highly qualified types of customers that we can pair with customers that are maybe slightly less budget constrained, but are very much quality oriented.

So this was really building the supply side of it alongside the fact that this coupled with the more advanced types of tools that we offer to our customers like project management or the ability to work with a professional to actually translate what you're looking for into a well outlined brief or purity was there. And we realize that unlike the gig portion of our business that requires a higher skilled professional.

Okay. Understood. One of the other interesting areas has been the growth in services on the platform. You've built some of that. You've done some things to bring things in inorganically in the company as well. How should we think about where services sits today? What some of the key learnings have been? And where do you think that might go as a percentage of the business over the medium to long term?

Yes. Look, we've identified the fact that because of our relationship with our customers, there is an opportunity not just to provide a solution for specific services, but actually create tools around them. So -- and this goes both to the demand side and the supply side. And this was a way for us to also diversify our business.

And so that could be on the supply side by creating these highly specialized services and tools for our sellers, things like Seller Plus, which are the high-quality sellers love because it helps them generate more business. And I should say that across all of the services that we offer to our community, the idea is it's value creation.

If it doesn't create value, you can maybe convert someone, but they're not going to be retained. And so it's really important that whatever service that our community doesn't matter if it's buyers or sellers, it needs to be ROI positive. It needs to generate more than you're actually spending on it.

And so I've mentioned Seller Plus as an example for these specialized solutions that we have for the supply chain, but we also identified the opportunity in areas where we see growth, like drop shipping is an example. Drop shipping spans across maybe 40 different categories. Because if you're in the drop shipping business, you need anything from the specialized platform knowledge, whether that is Shopify or eBay or Walmart, whatever it is, to product research to then product marketing, UGC, store management, personal assistance, whatever it is. So we have about 40 different categories, and they're all growing very nicely.

But at the same time, you say, okay, great. So if we have that portion of the business, what tools are they using? So for us, it was very obvious to say, okay, so there's -- if there is an opportunity to offer these specialized services maybe there is an opportunity to actually combine them with SaaS solutions for drop shipping, which is why we've acquired AutoDS, which is Shopify's #1 partner. And this is a really high growing business. But they provide the platform for dropshippers to monitor store, but it locks the connection with services. We have the services, but not the SaaS.

So this was -- for us, it was an obvious way of actually combining the 2 businesses and actually growing -- creating something that grows even faster. We actually talked about this, and we began to integrate AutoDS, and we've introduced the Shopify store builder. So now you have this combination of [indiscernible] services. And we've seen that portion of our business growing faster than we anticipated, which is awesome.

Great. Okay. I know we have a few minutes left. You referenced earlier the free cash flow you generate, you referenced earlier some of the buyback activity you've done in the last couple of years.

Talk a little bit about the priorities for investing back in the business rather than other uses of capital like returning capital to shareholders and sort of striking a balance between growth and investments and capital returns.

Yes. So for us, -- and I've said that earlier, growth is #1. There's a massive market opportunity. And we think that this is, even for returning value to shareholders. This is the best possible investment. So that's definitely #1.

We do think about -- I mean -- and you said it. We have -- we want to constantly have authorization to continue to buy back when it makes sense and do M&A. So sometimes, you have great ways of actually innovating and creating additional business from within. And sometimes it's inorganic, much like the acquisitions that we've done before. which makes sense. And I think that this is kind of the priority. So it's growth, it's creating value to shareholders, and it's M&A.

Okay. All right. I think we're going to leave it there with a minute or 2 to go. But please join me in thanking Fiverr for being part of the conference this year.

Thank you for standing by, and welcome to Fiverr's Second Quarter 2025 Earnings Conference Call. [Operator Instructions] I would now like to hand the call over to Jinjin Qian, EVP, Strategic Finance. Please go ahead.

Thank you, operator, and good morning, everyone. Thank you for joining us on Fiverr's earnings conference call for the second quarter that ended June 30, 2025. Joining me on the call today are Micha Kaufman, Founder and CEO; and Ofer Katz, President and CFO.

Before we start, I'd like to remind you that during this call, we may make forward-looking statements and that these statements are based on our current expectations and assumptions as of today, and Fiverr assumes no obligation to update or revise them. A discussion of some of the important risk factors that could cause actual results to differ materially from any forward-looking statements can be found under the Risk Factors section of our most recent Form 20-F and other filings with the SEC. During this call, we'll be referring to some key performance metrics and non-GAAP financial measures, including adjusted EBITDA, adjusted EBITDA margin and free cash flow. Further explanation and a reconciliation of each of the non-GAAP financial measures to the most directly comparable GAAP measures is provided in the earnings release we issued today in our shareholder letter, each of which is available on our website at investors.fiverr.com. And now I'll turn the call over to Micha.

Thank you, Jinjin. Good morning, everyone, and thank you for joining us. We delivered another strong quarter, building on a solid start to the year with continued momentum across our business. In Q2 2025, we achieved 15% year-over-year revenue growth and a 20% adjusted EBITDA margin as we continue to drive profitable growth with disciplined execution. Within our platform, we are seeing strong signs of durability and growth, including surging demand for AI-related services and the continued momentum of managed services and dynamic matching products.

While SMBs continue to take a cautious stance on spending and hiring amidst a volatile economic environment, our success in these efforts contributed to the acceleration in spend per buyer, which grew 10% year-over-year alongside robust growth across key verticals such as programming and tech, digital marketing and video animation. Additionally, we continue to expand our value-added services across our ecosystem. AutoDS continues to maintain strong growth momentum, and we've made meaningful progress in integrating our platforms to drive adoption and scale. Notable highlights include the launch of an AI-powered Shopify Store builder as well as the development of a deeper integration to enable seamless upselling and cross-selling between AutoDS and Fiverr's Marketplace.

The strong performance of the Services segment contributed to the revenue upside this quarter. We are also incredibly excited about how AI is positively impacting every dimension of our business, driving demand through AI-related services, improving discovery and conversion for buyers and sellers on the Marketplace and unlocking operational efficiencies through AI agents across functions. Today, I want to delve deeper into each of these areas and demonstrate how we are at the forefront of AI adoption with the speed, conviction and clarity that sets us apart.

First, the rapid development of AI technology is giving rise to numerous new skills, and Fiverr's Marketplace is becoming the go-to destination for accessing and engaging with AI experts. Categories such as AI agents, workflow automation and vide coding have experienced 5 to tenfold growth on our market base over the past 6 months. With AI fundamentally changing how humans and machines interact, it allows many nontechnical entrepreneurs and professionals to build and leverage the technology. At the same time, we are increasingly seeing the gaps between on-the-shelf AI tools and the real-world problems our customers are trying to solve. This is where Fiverr comes in.

Freelancers on our platform are filling the critical first and last mile gaps for our customers. This could involve setting up AI systems, selecting the most efficient AI models, integrating the back end with existing systems, adding functionality, creating custom workflows or simply debugging when the customer encounters issues. Fiverr's freelancers help our customers turn concepts and prototypes into high-impact solutions and tangible business results. These exciting trends underscore our conviction that human expertise is crucial in unlocking the full potential of AI.

With the proliferation of AI tools in the market and the increasingly ubiquitous access to these tools, we believe this represents a long-term tailwind for our business, driven by an increasing number of buyers who are deploying AI and their growing need and budget for tackling AI-related problems. That's why we are not only expanding our catalog to meet this demand but also exploring ways to embed Fiverr's talent network and transaction infrastructure directly into AI-driven workflows. These efforts include several ongoing strategic partnership discussions, the development of a targeted fulfillment capabilities and laying the technical foundations to build scalable AI powered by experts.

Our goal is simple. Fiverr freelancers are addressing critical challenges for businesses adopting AI, and we must strive to meet customers where they are and build an integrated experience that makes it seamless for them to leverage our platform. Second, we are shipping at an incredible pace so that we can leverage AI to strengthen our marketplace flywheel. Every transaction in our marketplace involves 3 core participants: the buyer, the seller and the platform. Our vision is to create an intelligent Agentic experience for each. 2 years ago, Fiverr launched Neo, the first-of-its-kind AI matching agent for buyers in a marketplace environment. Since then, Neo has evolved into a powerful AI engine that drives underlying KYC and matching across all of our front-end products.

As we continue to build out the Agentic experience on the buyer side, we envision a future where each buyer will be accompanied by a recruiting agent who can assist with drafting job briefs, communicating with freelancers, curating candidate list and even managing project execution end-to-end. This is the beginning of a search-less vision for the future, at least in the traditional keyword-based sense of search that makes room for more expressive and nuanced way to address customer needs. One that unleashes the power of the multiple solutions Fiverr has built on the platform to tackle any project from simple tasks to the most complex ones imagined that requires multitask and multi-talent orchestration and assembly.

On the seller side, we introduced Fiverr Go earlier this year, an AI assistant designed to help freelancers with project discovery, client engagement and creative ideation. Following the successful launch in February, Fiverr Go continues to drive strong seller engagement and meaningful conversion uplift across the funnel it touches. Similar to the Agentic AI experience on the buyer side, we have an extensive road map for Fiverr Go that will enable seller agents to provide more sophisticated support and guidance for our sellers, including service listing optimization, lead generation and qualification and other marketing, analytics, operations and production capabilities.

Last but not least, we are deploying Agentic AI across the internal functions to boost platform level efficiency from automating customer support workflows to enhancing marketplace integrity operations from improving job post matching algorithms to empowering customer success managers. These systems are designed to enable faster, more seamless orchestration between buyers and sellers while scaling operational productivity behind the scenes. As these agents become more capable over time, we believe they'll increasingly act autonomously, not only to improve individual workflows, but ultimately to enable agent-to-agent transactions that reduce friction and eliminate the need for their human counterparts to manually navigate the platform. While this is an ambitious long-term vision, the path is clear. Having this road map enables us to make informed architectural and product investments today. And these early bets will position Fiverr at the forefront of market-based innovation, further reinforcing our leadership through AI-powered differentiation.

As we wrap up the first half of 2025, I'm incredibly proud of how our team has delivered, and even more excited about our strategy and road map for the second half of this year. I look forward to updating you on our continued progress in the months to come. And with that, I will turn the call to Ofer.

Thank you, Micha, and good morning, everyone. We delivered a strong second quarter with both top and bottom-line exceeding expectations. Revenue for the second quarter was $108.6 million, up 15% year-over-year. Adjusted EBITDA for Q2 was $21.4 million, representing an adjusted EBITDA margin of 20%, an improvement of 80 basis points from a year earlier. We continue to generate strong cash flow with free cash flow totaling $25 million, up 21% year-over-year. As always, we remain focused on taking a balanced approach between growth and profitability while maintaining discipline with capital allocation.

Q2 saw solid performance across both our Marketplace and Services segments. Marketplace revenue was $74.7 million, driven by 3.4 million active buyers and 318 of spend per buyer and 27.6% of marketplace take rate. Within the Marketplace segment, we saw strong demand for AI-related services and AI category expansion and managed services remains an important channel for upmarket penetration. We are encouraged by the accelerating growth across several of our core verticals and the steady increase in larger, more complex projects on the platform.

In Q2, over 50% of GMV on our marketplace came from transaction over $200 and the higher-value transactions are growing at a double-digit pace year-over-year. This is a strong indicator of our marketplace ongoing evolution towards serving more sophisticated business needs. While the overall macro conditions do not warrant us to revise our assumption going into the second half of the year, we believe the structural tailwind within the marketplace segment, particularly around AI and upmarket adoption will continue to help offset broader economic headwinds and serve as a sustained growth driver.

Services revenue was $34 million, representing year-over-year growth of 84% and 31% of our total revenue in Q2. The upside was driven by several key initiatives, including the launch of AI-powered Shopify Store Builder, streamlined cross-sell execution between AutoDS and the marketplace and continued momentum in Seller Plus. We continue to see strong engagement and positive conversion impact from Fiverr Go, leading to incremental uplift to Seller Plus premium tier subscription in Q2. Looking ahead, we expect services revenue to maintain healthy momentum, and as mentioned previously, expected to represent a little over 30% of total revenue for the full year 2025.

Now on to guidance. We are reiterating our revenue and adjusted EBITDA guidance for the full year 2025. We expect full year 2025 revenue to be in the range of $425 million to $438 million, representing year-over-year growth of 9% to 12%. Adjusted EBITDA is expected to be in the range of $84 million to $90 million, representing an adjusted EBITDA margin of 20% at the midpoint. For the third quarter of 2025, revenue is expected to be $105 million to $110 million, representing year-over-year growth of 5% to 10%. Adjusted EBITDA is expected to be $21.5 million to $23.5 million, representing an adjusted EBITDA margin of 21% at the midpoint.

We continue to operate with the highest level of discipline and efficiency. We believe we are on track towards our long-term target to reach 25% adjusted EBITDA target in 2027 and deliver 40% CAGR in free cash generation for the 3 years ending in 2027.

To close, we continue to execute on the goals and road map we set at the beginning of the year and are looking forward to the second half as we remain well-positioned to capture the enormous opportunity ahead of us. With that, we'll now turn the call over to the operator for questions.

[Operator Instructions] Our first question comes from the line of Eric Sheridan of Goldman Sachs.

When you think about services revenue growth building and momentum deeper into the second half of this year, but more on a multiyear view, what do you guys see as some of the key investments you need to make to sort of unlock continued scaling of that service revenue? And maybe honing in a little bit on Seller Plus, understood the goal you have around getting to a certain rate of penetration. But how do we think about maybe over the longer term where that can go in terms of a percentage of the mix?

Eric, thanks for the question. So, for the services, we expect services revenue to continue growing at a healthy double-digit rate exiting this year after we lap both Fiverr adds and AutoDS acquisition in Q. And we believe that services still have a long growth runway ahead of us as we continue to expand value-added services to productivity, financial and other adjacent tools for our freelancers. In addition, we're seeing opportunities to drive synergies between AutoDS and Fiverr. And as we mentioned this quarter, we unlocked some nice Shopify affiliates revenue through the new AI-powered Shopify Store Builder tool. And we believe that services will continue to be a growth catalyst for our overall revenue growth.

As to Seller Plus, it continues to expand value-added services, to drive adoption among a wider range of audience. Fiverr Go is an example of that. And we keep adding tools and the audience of Seller Plus is growing as they see this as a net positive contributor to the growth of their business. And broadly, we are expanding value-added services to a wider range of services for freelancers. We want to empower the entire freelancer career, including those financial tools and benefits. So, there's a lot of growth runway there.

Our next question comes from the line of Ron Josey of Citi.

Micha, I wanted to talk and ask a little bit more about the managed services offering and the broader mix shift to upmarket. And just talk to us a little bit more about the progress here. I know there were some stats in the letter around 50% of marketplace driven by transactions over $200. And so, as we see greater demand for managed services, as we see the mix shift to upmarket, help us understand just the progress overall and maybe the size. And then from an AI or I guess, Fiverr Go perspective, I think we saw and heard a little bit more about conversion rates. Any more insights there would be great.

Thanks for the question. So essentially, managed services, as we think about it in general, is a part of our going upmarket strategy. And as we said time and time again, this is -- I mean, you can think about the history of the company. A company that started from services of $5 and now more than half of our business is coming from transactions of over $200, which is incredible. And this is thanks to the fact that we've actually extended beyond just being a marketplace into becoming a platform, where we offer a whole suite of different services so that we can mature with our customers and cater to more of their needs. And this is really demonstrated in the fact that these managed services are growing very nicely. And this also has to do with our acquisition strategy, which means we're focusing on customers that have larger wallets, that have more sophisticated and complex needs and making sure that we have the right tools to answer for those needs.

And we're seeing this. And by the way, as this grows, it is actually decreasing our exposure to the segment of the market of SMBs where they're more exposed to macroeconomics than the larger customers. So, this remains very much in the focus of what we're doing, and we're very encouraged by the growth that we're seeing there. In terms of Go, I think we provided color in our letter to our shareholders. As you noted, yes, it is increasing the conversion rates and the speed and efficiency and decreasing the time to convert for our sellers. And this is why we've seen a 50% jump in the amount of sellers that are using it. They love it. It's an incredible tool that actually let them sleep at night quietly knowing that their business is continuing to run.

In addition, the way we're thinking about this, and this is a part of our entire agentic vision for the company, what we're doing is we're building sophisticated tools that allow them to not just deal with the communication, but actually smart AI that will allow them to optimize their offerings and actually take actions for them to make sure that they get the best exposure and they maximize their potential for deal flow. And I think that this is -- I mean, we're pretty early in that phase, but the things that we're building internally are just incredible. And I think the fact that we started with Neo way, way ahead of any company, any marketplace out there, allowed us to build the muscle and build one of the most incredible AI teams that I know in the world. And we're building incredible tools that I'm sure we're going to be able to talk about in the next few quarters.

Our next question comes from the line of Doug Anmuth of JPMorgan.

I have 2. Maybe just first in terms of supply/demand, just balance on the marketplace. I'm just curious how you're thinking about the supply on the freelancer side with AI expertise and whether that's where you need it to be? And then second, just as you think about the marketplace business, what would it take for marketplace to return to growth in your view?

Doug, thanks for the question. As for supply and demand, I think we're in a good place there. The one benefit that we have with freelancers is that they are ahead of most full-time employees in terms of embracing new technology and new tools. A lot of them are what we call AI native. And we're seeing how they're actually elevating the outputs of their work in an incredible way. And obviously, when we open new categories, sometimes it takes a little bit of time, a couple of weeks to build new supply. And as we go into the areas of more sophisticated types of services, obviously, we allow more sophisticated types of sellers to join the marketplace. But I think that we're in a very good place. We don't have a category where we don't have the right supply to entertainment needs of the customers.

As we think about the marketplace business, we believe that the efforts that we're doing around AI and upmarket will allow us to turn to growth even in the current macro. But given the macro uncertainty, we don't think that we're ready to bake this into guidance yet. As we said a number of times before, and I think I mentioned this in my first answer, the fact that we're going upmarket and actually acquiring high-value buyers, by definition, lowers our exposure to the segments of the buyers that are mostly affected by macro-economical conditions. And therefore, even without the macro turning, over time, we will continue -- we will turn into growth. And obviously, if macro turns, it's going to be a further tailwind for us.

Our next question comes from the line of Jason Helfstein of Oppenheimer & Company.

This is Steve Hromin on for Jason. So, 2 questions from us. One is, have you seen any specific categories relating to AI? And second, if more entry-level jobs are going to be replaced with or by AI, how does that affect demand for Fiverr services?

Thanks for the question. Essentially, I mean, if -- we haven't seen any accelerated decline. But as we said before, we made the comment about simple versus complex types of services. And we said that the simple types of services, and usually, they're associated with very, very low prices are going through displacement. But at the same time, there is many new categories and many new needs that we can answer for. And by the way, that's a good thing for us, because we're obviously less focused on the micro types of services and more interested in the more complex, nuanced types of services that our larger customers need.

As to entry-level jobs, look, I think the definition of an entry-level job is changing. I think the expectation of everyone is to do more, thanks to technology. And again, I think that much like freelancers, I think that people that are in their very beginning of their career cycle have the benefit of actually being AI natives, which means that even on entry-level jobs, they can do much, much more than that. I said that many times, I'll repeat it now. I think what AI has done is to elevate the floor for everyone, meaning it's providing incredible tools and capabilities to everyone, but that's exactly the thing. No one has an advantage, thanks to AI because the tools are available for everyone. So essentially, they're changing the starting point, the floor, but it is not changing the ceiling. So, I think that it's moved all jobs up, including entry-level jobs. This is what we're seeing at least.

Our next question comes from the line of Brad Erickson of RBC Capital Markets.

I have 2. First, on the services revenue. Can you just update us on the mix within that line item between what you might regard as sort of recurring revenue versus more transactional and then kind of how you think about the general visibility level? And then I have a follow-up.

So essentially, when you look at the services side, it is mostly made out of Promoted Gigs, Seller Plus and AutoDS. And all of them are continuing their momentum. There is nothing specifically to call out there. Was there a second part to your question?

It just kind of general visibility related to obviously, some of the revenue being sort of what might look like recurring versus purely transactional.

I think that to give some more color. So, I think Promoted Gigs is not recurring by design. It's -- but it's pretty much ongoing, and we see a very high retention of sellers who are using this product. It's aligned with the GMV on one hand, but it has more room to expand as we have more inventory that hasn't been utilized yet. And I think one territory that we haven't expanded to in terms of Promoted Gigs is the Fiverr Pro environment, which is growing faster than the Marketplace.

So, I think that -- but this is one example. Then on the Seller Plus, Seller Plus is subscription by design. This is the offering. There are multi-tier -- there are a few tiers of offering for different services being offered and packaged. And as time goes, we expand those offerings, adding more capabilities and opening more tiers. So, I think on the Seller Plus as well, it has the room to -- for us to offer more services and to charge accordingly and then also to expand in terms of the target audience, approaching sellers that were not entitled for the program before and offering them a different class or different type of tool to improve their business.

And lastly, AutoDS is subscription, has been growing nicely and contribute more than expected on the second quarter. And we see this business building up nicely as we just finalized the integration with Fiverr ecosystem. We see that building nicely looking forward. I don't think we break down each and any of those to line item. But this is, I hope, more color and visibility to this line of revenue.

Yes, it is very helpful color. And then just second one, maybe following on to a few of the questions around Marketplace growth. You've talked for a while about this whole mix shift between complex services and simple services. I guess, do you feel like you have visibility to when that mix shift might become a tailwind? I know you just said you're not ready to guide to that because of macro. So, I guess the question is like do we need to see macro ultimately improve? Or do we reach a point maybe at some point soon where that mix shift is enough to push the Marketplace back to growth? How to think about that?

Yes. Thank you. Look, I said it before, we don't know exactly when this is going to be the case given the volatility of the market. That said, we do believe that by going upmarket and enjoying the tailwinds that AI is giving us, over time, this is going to bring us back into growth on the marketplace side, even without macro economy taking a turn.

Our next question comes from the line of Bernie McTernan of Needham & Company.

Just one for me. I wanted to just follow up with one of the comments you made in your prepared remarks that you were in discussion with several companies on potential partnerships with -- on AI. Just wanted to -- I know it's too early to announce what those are, but what could those partnerships actually look like, either the technology being better put into your marketplace or you -- or Fiverr's Marketplace being on like other AI platforms? Just how we should think about those partnerships potentially playing out?

Bernie, thanks for the question. Look, what I referred to is what we call internally scaled AI using experts. Essentially, what we're seeing with technology is we're seeing technology allowing everyone to take a pretty meaningful step forward. That said, using many of these technologies in a way that would create either a complete product or would provide a competitive edge usually requires either at the beginning or at the end or both using someone who's an expert in using these tools, as friendly as they are. At some point, they become technical. At some point, they reach their limitation. I think this is exactly -- this is exactly the concept of actually working with those who are -- who can give you either the help to get to the finish line or the competitive edge required to build successful things. And so, without getting into specifics, the things that we did with certified could give you a sense of how we're thinking about empowering our customers as they use these AI tools.

Our next question comes from the line of Josh Chan of UBS.

I guess 2 questions for me. One, on your AI capabilities. I know that because they're more complex, you're getting higher ASP from your AI type of services. I guess I'm wondering, are you also attracting new and different customers that weren't part of the Fiverr platform before that are attractive to your AI capabilities? And then I guess my second question is on the margin front. As you think about the path to targets, how dependent on GMV improvement will your margin trajectory be?

Josh, thanks for the question. The answer for your first question is yes. Given the fact that it's more complex types of projects, it allows us to actually get the types of customers that we're very much focused on, which are customers that are able to actually pay and fund these types of expertise. And again, having these experts on our platform is something that we've been working on for a while. And it's definitely paying off. We called out the positive and accelerated growth around programming and tech and video -- and animation verticals as an example. And obviously, we're seeing that also accelerating our growth in the spend per buyer in a very nice way, actually putting us in a great position where I don't think that there is any company in our space that delivers the performance at our level with double-digit growth and a 20% EBITDA. And so, investing in these areas allows us to deliver one of the best performing businesses in our space.

On the second part on the margin and dependent on GMV improvement, I think we have a 2027 margin -- long-term margin that we are committed to the 25%. And I think as you can see, we are moving there step-by-step and have full confidence with the ability to deliver on the margin improvement. And this is taken into consideration multiple scenario of GMV growth or GMV flat. So, I don't think there is a dependency. We have the flexibility, but also the confidence and the ability to deliver despite a different scenario.

Our next question comes from the line of Matt Condon of Citizens.

My first one is just on marketplace revenue in the quarter. It looks like trends on a year-over-year basis deteriorated in the second quarter, and that's just in the backdrop of a stable macro environment and the AI tailwinds that you guys are seeing. Can you just elaborate just on what you're seeing, I guess, real time, if there's anything to call out there? And then my second question is just on marketing spend. Specifically, as you guys move upmarket, are you seeing any changes in your return on marketing spend or any changes in your TROI framework, specifically as you move up market?

Yes. So, I think the answer for both questions is that the quarter behaved as expected. I think we have committed to marketplace being flat, maybe a low single-digit decline, and that's exactly where we are. And honestly, I think that as we look to the second half of the year, we anticipate or predict a similar trend. And as for the marketing trend, I mean, you can track, and we provide the ROI data and return. I think that those are -- those fall into the expectation without much exception.

Our next question comes from the line of Rohit Kulkarni of ROTH Capital Partners.

One question on -- you've mentioned a few times, AI agents, workflow automation in the letter. Maybe talk about what you're seeing from a demand standpoint, what types of customers or what types of use cases you feel Fiverr is well-positioned to address the demand in AI agents and robotic process automation, workflow automation in a new AI world. And then just on this -- based on our calculation, I think GMV seems to have accelerated almost 3 or 4 quarters in a row. Perhaps talk about what your assumptions are heading into second half as the guidance seems to be unchanged but just talk through how you feel the sustainability of GMV acceleration, although from a minus 1% to plus 2%, but that's still encouraging in my opinion.

Rohit, thanks for the questions. I'll start with the first one. Look, AI-related services are booming, and the demand is surging, especially around AI agents, workflow automation and vibe coding. As an example, AI agent, 35% growth quarter-over-quarter, 10x growth in the last 6 months. Workflow automation and vibe coding growing 10x in the last 6 months, as customers are using services like Meg.com or GoHighLevel to create those workflows. AI development, 137% growth year-over-year. Mobile app, 47% year-over-year. AI consulting, 37% year-over-year. And also, we're seeing some prolifheral categories that are benefiting from it, such as data governance and protection with 58% growth and crowd funding with 51% growth, where AI drives -- is now driving millions of start-ups.

And so, AI experts on Fiverr play a critical role in helping customers navigate the rapidly evolving AI landscape and bridge the gap between technology and tangible business impact. So, we're excited about this. We think that this is just the beginning, and we're going to see a lot of more categories that are going to be affected as a result of it and experience accelerated growth.

I think on the second part of the question, what we are seeing is a really stable macro. We don't interrupt small deceleration or acceleration as a trend. And I think this takes us to the assumption on the second half of the year, which is a reiteration of what we said at the beginning of the year. We don't think -- unless something materially changed on the macro environment, we think that's going to be stable between flat and maybe slightly 1-digit decrease in GMV. But that's what we anticipate in the second half of the year.

I would now like to turn the conference back to Micha Kaufman for closing remarks. Sir?

Thank you, Latif, for moderating the call today and to everyone who dialed in this morning. I look forward to speaking to all of you very soon. Have a great day.

This concludes today's conference call. Thank you for participating. You may now disconnect.