CarGurus, Inc. Class A Aktie

Good day, and welcome to the CarGurus' Second Quarter 2026 Earnings Conference Call. Please note, this event is being recorded.

I would now like to turn the call over to Javier Zamora, General Counsel & Corporate Secretary. Please go ahead.

Good afternoon and thank you for joining us. With me on the call today are Jason Trevisan, Chief Executive Officer; and Sam Zales, President and Chief Operating Officer.

We will be making forward-looking statements, which are based on our current expectations and beliefs. These statements are subject to risks and uncertainties, and our actual results may differ materially. Information concerning those risks and uncertainties is discussed in our SEC filings. We undertake no obligation to update forward-looking statements, except as required by law.

Please refer to our press release and our investor presentation on the Investor Relations section of our website for a reconciliation of GAAP to non-GAAP measures.

I'll now turn the call over to Jason.

Good afternoon. We delivered strong second quarter results with revenue growing 13% year-over-year to $251 million, above the midpoint of our guidance range, including another robust quarter in our international business, which grew 28% year-over-year. Adjusted EBITDA increased 7% year-over-year to $85 million, at the high end of the guidance range at a 34% margin, reflecting disciplined investment. We also generated strong cash flow, converting 103% of our adjusted EBITDA or $88 million to free cash flow in the quarter.

In the first half of this year, dealers have taken a more deliberate approach to incremental spending decisions. We believe market trends like dealer margin pressure and fewer days on lot, coupled with onetime developments like recent FTC mandated all-in price transparency requirements made dealers more cautious about spending. We view these factors as temporary, not structural. Despite these pressures, dealer engagement continued to grow on our platform.

Average sessions per dealer on our platform were up 28% year-over-year in Q2, driven by increasing AI functionality in our products and expanded capabilities across more of the dealer workflow. We continue to grow wallet share, which we believe is driven by our strong ROI and new dealer product introductions. We believe we remain well-positioned to continue capturing a disproportionate share of incremental spend as the environment improves. Our full year revenue growth guidance is unchanged, and we remain confident in our long-term opportunity.

We've also begun to realize meaningful benefits from our investments in AI, which is helping us accelerate the pace of innovation, enabling our teams to bring new products to market faster while improving engineering efficiency and operating productivity across the business. Combined with our focus on operating discipline and organizational efficiencies, these productivity gains have generated greater leverage than we anticipated at the start of the year. As a result, we are raising our full year profitability outlook and now expect full year non-GAAP adjusted EBITDA margins to compress approximately 50 to 150 basis points in 2026 relative to 2025.

We believe our ability to accelerate innovation thoughtfully and operate more efficiently positions us well for all market environments. More importantly, the investments we've made are reinforcing a virtuous cycle across our platform. As we have used AI to innovate faster and bring new products to market more quickly, we are seeing deeper engagement with our platform from both dealers and consumers.

In addition, we continue to capture rich first-party shopper signals, nearly 0.5 billion signals each day across demand, pricing, inventory and shopper behavior that inform and continually improve our dealer software and analytics and consumer experience. This creates a differentiated data advantage that we believe helps dealers make better decisions while enabling a more personalized and trusted consumer experience.

We believe faster innovation, leveraging our proprietary marketplace data makes CarGurus increasingly valuable to both dealers and consumers and strengthens our competitive position by providing a better car shopping experience for both consumers and dealers.

As we continue to expand beyond our leading marketplace business, our strategy is built around 3 value creation drivers. First, we're expanding CarGurus' offerings into integral parts of the dealer workflow, connecting inventory, marketing, lead conversion and data pillars through mutually reinforcing products. Second, we're focused on transforming car shopping into a trusted AI-led journey from research through consideration and purchase, giving consumers greater confidence and increasing the value they get from CarGurus. And third, we are deploying capital with discipline with the aim of growing long-term earnings power and stockholder value.

I will now walk through our second quarter progress across each of those drivers. Driver #1, expanding CarGurus' offerings into integral parts of the dealer workflow, connecting inventory, marketing, lead conversion and data pillars through mutually reinforcing products. We have increased dealer engagement with our platform, reflecting our focus on embedding data and insights into more of the dealers' daily decision-making.

We're leveraging our differentiated data on retail dynamics, pricing and inventory trends and deep consumer insights to inform smarter and more predictive dealer decisioning, which we believe results in an ultimately stronger ROI for our dealer customers. In fact, among subscribing independent dealers, those in the top quintile of engagement with our platform had a 78% higher leads per unit than those in the bottom quintile, suggesting that use of our expanding dealer product suite not only drives adoption among more people at the dealership, but also drives performance on our platform and ultimately, dealer profitability.

Within the inventory pillar, our focus remains on helping dealers source, stock, appraise, merchandise and price inventory more effectively. In the second quarter, PriceVantage bookings grew more than 50% sequentially with a higher AOS as the product has continued to prove its value. Dealers that adopted PriceVantage saw a median 15% lift in VDPs and 9% lift in leads per listing after adoption.

In our marketing pillar, we recently introduced VINMax, our newest AI-powered product that helps dealers improve merchandising performance and achieve turn time goals without dropping price and sacrificing margin. VINMax identifies high potential, but underperforming VINs and dynamically boosts them across organic sort, highlight and audience targeting.

Since we began rolling out VINMax to early access dealers in February, promoted listings have sold 23% faster and received 34% more leads per day than comparable non-promoted listings.

In the conversion pillar, we introduced a new competition filter to shopper signals that helps dealers understand how many other dealers a shopper has submitted leads to, allowing them to better assess urgency, allocate sales resources more effectively and convert customers to sales at a better rate.

Within the data pillar, we are building on our data advantage by equipping dealers with deeper competitive insights, helping them benchmark their performance to competitors on metrics like leads per vehicle or recently sold vehicles on competitor lots, so they can make more informed pricing and inventory decisions. We've seen a greater than 80% open rate on the weekly competitive digest e-mail, illustrating how our data has become a critical input into dealers' daily operations.

Collectively, these new products have extended CarGurus beyond our marketplace and into daily dealer decision-making, which we believe will fuel growing dealer engagement, greater product adoption, stronger retention and more dealer wallet share over time.

Driver #2, transforming car shopping into a trusted AI-led journey from research through consideration and purchase, giving consumers greater confidence and increasing the value they get from CarGurus. Buying a vehicle remains one of the highest consideration purchases a consumer makes, a process that often lasts several months as they research, explore, select and negotiate their final deal.

Consumers want confidence in their search, transparency in their results and trusted guidance throughout that process. In July, we launched Guru, our end-to-end consumer-facing brand for all of our AI capabilities across the CarGurus platform. Guru currently exists in 2 forms: as an AI native experience, formerly known as Discover, and as a seamless overlay to our existing site.

Together, fueled by real-time dealer feeds and a robust methodology and ontology, they serve as a trusted guide that helps shoppers research better, understand and compare their options faster, make more informed shopping decisions and complete their purchase in the dealership with greater confidence.

Engagement with Guru continued to grow rapidly with Guru-driven leads up 60% sequentially in the U.S. In addition to better shopper to lead conversion rates, Guru provides much richer signals about user preferences and intent that bolster the depth and quality of the lead we send to dealers. Given the strong usage in the U.S., we recently launched our conversational AI experience in the U.K. and Canada, allowing shoppers to shop by describing their needs and use cases, which is not solved easily with traditional one-size-fits-all filters rather than searching by vehicle specs.

In the consideration step of car shopping, Sell My Car continued to gain traction, improving the selling experience for consumers as we made enhancements to the consumer experience that increased funnel conversion and drove significant incremental leads. These improvements also gave dealers more efficient access to sourced inventory at a time when acquiring used vehicles remains a top dealer challenge.

At the point of purchase, dealership mode extends the CarGurus experience into participating U.S. dealer lots by putting pricing transparency, deal ratings, payment estimates and Guru-powered vehicle comparisons directly in shoppers' hands. Consumer engagement with dealership mode in our app more than doubled in the second quarter, further cementing us as a trusted tool for both consumers and dealers, which we believe will deepen engagement and facilitate more closed deals and cars sold on our expanding platform.

Driver #3, deploying capital with discipline with the aim of growing long-term earnings power and stockholder value. We believe our subscription business model is resilient, and we expect it will continue to generate strong free cash flow that will fund our long-term growth initiatives while also allowing us to return capital to stockholders.

In 2026, we have increased our investment in product, technology and development to continue accelerating AI expansion on our platform and supporting a faster pace of product innovation. We've also increased our investment in sales and marketing to support the launch and adoption of new dealer products and create consumer awareness of our Guru and other AI user experiences. In addition to organic investment, we plan to maintain the flexibility to pursue disciplined M&A for compelling strategic opportunities.

Finally, we remain committed to returning capital to stockholders through share repurchases. In the second quarter, we bought back $29 million in shares, increasing our year-to-date repurchases to $204 million of the $250 million available under the 2026 share repurchase program. Since 2022, we have repurchased approximately $925 million in shares, representing more than 30% of shares outstanding, reflecting confidence in our long-term strategy, strong financial position and commitment to disciplined capital allocation.

We are excited about our accelerated innovation velocity and how that has translated into progress across all 3 value creation drivers. By introducing new AI-driven products and features that help dealers operate more effectively and give consumers greater clarity and control throughout their shopping process, we believe we will continue to deepen our role in the car shopping journey.

We have embedded our data and AI more deeply into dealer workflows, driving deeper engagement from more people at the dealerships. Our expansion into software and data, all unified by our underlying data layer is demonstrably improving the marketplace performance of adopting dealers and growing our TAM and wallet share potential.

And among consumers, we are creating more personalized, trusted and transparent experiences across more steps of the shopping journey. Backed by strong free cash flow and disciplined capital allocation, we will continue investing in the product innovation and AI capabilities that we believe will generate the greatest long-term returns and create durable long-term value for our customers and our stockholders.

Now let me walk through our financial results, followed by our guidance for the third quarter and full year 2026. Second quarter revenue grew 13% year-over-year to $251 million, above the midpoint of our guidance range, driven by adoption of add-on products. Our OEM business outperformed our expectations, benefiting from consumer interest in certified pre-owned vehicles amid ongoing new car affordability challenges.

In the second quarter, U.S. QARSD grew 8% year-over-year, and we added 673 paying U.S. dealers year-over-year. We continue to increase our dealer base while taking greater wallet share. In the second quarter, adoption of add-on products was the largest driver of year-over-year QARSD growth, followed by listings upgrades, like-for-like price increases and higher lead quantity and quality. For the third consecutive quarter, add-on product adoption was the largest driver of the sequential increase in QARSD.

Our international business had another robust quarter with second quarter revenue up 28% year-over-year, driven by strength in listings and Sell My Car in Canada and OEM advertising in the U.K.

I'll now discuss our profitability and expenses on a non-GAAP basis. Second quarter non-GAAP gross profit grew 12% year-over-year to $231 million. Second quarter non-GAAP gross margin was 92%, down about 90 basis points year-over-year. Second quarter non-GAAP adjusted EBITDA grew 7% year-over-year to $85 million, toward the high end of our guidance range, and adjusted EBITDA margin was roughly 34%, down about 200 basis points year-over-year.

Second quarter non-GAAP operating expenses totaled $154 million, up 16% year-over-year, reflecting higher sales and marketing expense and increased investment in product, technology and development expense to continue the accelerated pace of AI product introductions.

Second quarter non-GAAP net income per diluted share attributable to common stockholders was $0.66, up 16% year-over-year. We generated strong cash flow, converting 103% of our adjusted EBITDA or $88 million to free cash flow in the quarter. We ended the quarter with $122 million in cash and cash equivalents, an increase of $50 million from the end of the first quarter as $29 million in share repurchases was more than offset by our cash generation.

Since 2022, we've now repurchased approximately 30% of our shares outstanding, while we continue to grow revenue and profitability. As of the end of Q2, we have $46 million remaining on our 2026 authorization, and we will continue to repurchase shares when we believe it is an attractive investment and consistent with our capital deployment priorities.

I will now turn to our guidance for the third quarter and full year 2026. We expect third quarter revenue to be in the range of $253.5 million to $258.5 million, up between 9% and 12% year-over-year, respectively. Our guidance reflects a more measured pace of dealer decision-making, which we view as temporary, not structural. We believe dealer engagement and retention remain healthy, and we expect contracted new product revenue to layer in through the second half, supporting our full year outlook.

For the third quarter, we expect our non-GAAP adjusted EBITDA to be in the range of $82 million to $90 million. We expect third quarter non-GAAP earnings per share to be in the range of $0.63 to $0.69 and diluted weighted average common shares outstanding to be approximately 90 million.

Turning to the full year. We are reiterating that we expect 2026 revenue to grow in the range of 10% to 13% year-over-year. We are raising our full year profitability outlook. We now expect full year non-GAAP adjusted EBITDA margins to compress approximately 50 to 150 basis points in 2026 relative to 2025. Our updated guidance reflects more efficient execution within our investment year plan while preserving our ability to invest in the most attractive growth opportunities.

With that, let's open up the call for Q&A.

[Operator Instructions] First question, Andrew Boone with JMP Securities.

I wanted to double-click in terms of the macro. Is there anything else that you can help us understand about what dealers are seeing and what those conversations are kind of looking like? And then secondly, how do we think about the duration in terms of kind of this downturn? Are you guys seeing any improvement of late? Or has it been more of the same?

And then just a bigger picture question. As we think about users and the greater adoption of AI and your guys' kind of transition to more of a natural language search, can you just provide us an update in terms of what that is doing to the consumer experience? Is that improving conversion? What else are you guys seeing as you guys unlock more of an AI experience for consumers?

Andrew, Jason here. So on macro, what we saw in the first half was -- and the data would support all this, at dealers, units tended to be up, prices tended to be up and days on lot tended to be down. And so what inventory they did have, especially lower-priced inventory, tended to sell well. And when that happens, they tend to be a little less reliant on marketing. At the same time, there's been margin pressure at dealer groups, and there's been GPU pressure at dealer groups. And so they have been looking for opportunities to save on OpEx.

And then you throw in some of the FTC activity, and that's what's led to them just being a little slower in decision-making. A lot of the trends that I just mentioned in terms of units and prices and days on lot, those -- they fluctuate. They change by the month and by the quarter. We have started to -- I would say some of those factors improved over the course of that first half of the year. And so as we look at how we're executing, we still are introducing a lot of products. We are still #1 from a listings perspective, still #1 in leadership in ROI, lead quality, lead quantity. We're gaining market share. And then we're growing outside of listings, too.

So you heard us talk about some of the products that we're introducing in inventory and conversion. And so as a result, you may have heard the stat about engagement is up significantly on our platform because more people at the dealership are using us more frequently.

How long will it last? I mean, like I said, it improved over the course of the first half. And so I don't know if there's a binary, it's sort of happening and it's not, but it did start to improve in the first half.

What is AI natural language search doing to our search. We've given some stats that those that are engaging deeply with Discover tend to convert a lot higher. I think the bigger and more exciting thing is that we're gaining just so much more information on the users, and we're able to leverage that, that they're giving us in the course of the conversation. We're leveraging that to; a, give them a better search result and experience on our site; and then b, pass that information along to the dealers so that the dealer can give them a better experience as well. And we're seeing that the dealers that are engaging with that like in shopper signals and things like that, are converting those leads much better.

And so the beauty of that is that the dealers and our customers get a lot more value out of the platform, like the same platform and the same users as before without having to pay more and without us having to invest more because we invested in a better search experience.

Next question, Rajat Gupta with JPMorgan Chase.

This is Jash Patwa on for Rajat Gupta. I was just curious about the profile of the dealers signing up for PriceVantage. Are these primarily dealers already on the higher-priced premium tiers? Or is adoption more broad-based? Just trying to gauge whether the uptake suggests a potential stand-alone market for some of your other products that are currently being bundled. And I have a quick follow-up.

Thanks, Jash. It's Sam Zales. I appreciate the question. It is a mix, and that's really exciting for us on PriceVantage. You saw the 50-plus percent growth in bookings, the breadth of our opportunity there. Remember what the value proposition here is, it's the first product in market that provides predictive analytics to dealers and has them look at the difference between wholesale and retail pricing. It uses consumer trends and demand to say to a dealer, where is there more demand for specific vehicles in your market? What should you be stocking up? And so the opportunity is broad-based. It's independent to franchise dealers across all segments. And those that are looking for a way to grow their profitability. Jason talked about the challenges and the economics in the market.

So thinking that you're offering a product that combines with our listings package, we're now able to help you provide a profitability increase to your business is working phenomenally. And that's what is part of the engagement story that Jason shared. So it's really broad-based. And it says to us, this is a product that gets us to a broader set of decision-makers within the dealership. And if it compels both sides of the marketplace, independents and franchises, we feel like that's a value proposition we're excited to take forward.

Awesome. That's very helpful. And I appreciate the initial color on AI leads Jason provided to the prior question. But just as a follow-up to that, with the traction you're seeing with Guru, are you at a point where you're starting to monetize AI search results or AI search placements?

Well, I mean, we've woven AI into many of the search paths right now and search features. And so it's not about monetizing an AI search result separately from how we monetize other user activity on our site. It's instead creating a better user experience on our site so that they are ultimately have a better chance of finding a car that's great for them and converting to connecting with a dealer. So unlike, say, Google traditional search results and Gemini being 2 different things, we've woven AI into our core user experience, which allows it to get leveraged by and exposed to a broader portion of our audience and just improve any search that's done.

So the volume that you heard sequentially was up significantly. So more users are using sort of what we call the deep AI search features, but it's really affecting all searches there. It's not separate.

[Operator Instructions] The next question comes from John Babcock with Barclays.

Just first, I was wondering, you did talk about more efficient execution on the plan. I was just wondering if you could elaborate more on that and then what the drivers were.

Sure. So we have, I would say, pretty aggressively looked at how we can infuse AI and agents into workflow across our business. And it's largely -- and it's working well. I would say it's largely been focused on how we increase our product velocity and our productivity. But we've also found it to help with efficiency in other parts of the org outside of product and engineering. We've found that we're able to save in software versus where we expected we would. And now some of that's offset by AI by token usage and other AI software. But it's largely coming through productivity and efficiency. There are some areas that -- including in go-to-market efficiency as well.

There are some areas like in customer support where it is -- this is a form of software replacement, but where we're able to get a lot more efficient than what we were able to do with third-party software, third-party partners or services with our own internally built agents. So there are some concrete ones there. And the thing that we're really excited about is we've been able to do that without, in any way, sacrificing on the product velocity and accelerated product introductions that we talked about at the beginning of the year. In fact, if you look at all the new products we've introduced in the last 12 months, it's a very long list, and that has helped fuel on both the dealer and consumer side. And that has helped fuel that add-ons is our biggest driver of QARSD, and we have such great traction with things like PriceVantage where bookings were up 50% sequentially.

Got you. And then just a quick follow-on here. Just on the M&A side, you did talk about that very briefly, but I'm just kind of curious if you could remind us what sort of deals you might be looking at, what sort of skill sets and capabilities you might be looking to add?

So we always have our radar up for good opportunities that fit into our strategic plan, and our strategic plan is pretty ambitious on both the dealer and consumer side. But I would say from a or like a richness of opportunities mingled with our strategic focus, I would say the more likely opportunities are on the dealer side. And so it would be technology solutions, software that help serve the dealer that likely tend to be smaller, more point solutions, but that would help us accelerate our expansion into our 4 dealer pillars faster. We have a really exciting organic road map in all 4 of those, and we have products in all 4 of our pillars already, but we want to go faster.

And we also want to make sure that we're tying these 4 pillars together through our common data set and through integrations with other systems. And so there is quite a bit of work to get to the robust platform that we see ourselves getting to. And if we can get there faster with an acquisition, we will. This isn't anything too novel, but the important things to us are great technology and a great team.

Next question comes from Naved Khan with B. Riley Securities.

This is Ryan Powell on for Naved. I wanted to kind of drill into the sales and marketing investments. You framed them as both focused on product or focused on dealer and consumer reach. So I guess, first, any difference in traction from franchise versus independent dealers? And then also good to see some growth in the unique users and sessions in the quarter. So how much of that was driven by organic traffic versus paid?

Thanks, Ryan. It's Jason. No, I would not -- on your first question, I would not say there was -- there is a difference on our traction between franchise and IND. We've always served both of those segments and top to bottom, large to small within each of those segments really well. That's one of the benefits of our platform. And as Sam mentioned or answered earlier, PriceVantage has been broadly embraced. And so even a product like that, that's in a new pillar for us, is not all that segment specific at all. So I would say it continues to be market-wide, the appeal of our products.

In terms of users and sessions, just as a reminder, we don't focus on driving upper funnel-traffic. We focus on the value we deliver to dealers. If we achieve that by having more traffic and more sessions, that's fine, but we can also achieve that without doing it because it's really a small percentage of our total traffic that converts to connecting with the dealer. But we did grow our investment in marketing as well as in sales. I would say more of that growth has been on brand and brand building. Our direct and owned and app and organic traffic continue to be our fastest growth. And yes, beyond that, I would say, at the channel level, we don't get into that.

[Operator Instructions] Next question, Chris Pierce with Needham & Company.

I just wanted to get a sense, as you've been introducing these new products, could you -- for a lot of us, maybe we don't specifically know how many products or the level of detail that dealers are really taking when you sort of get down to except we hear that they take lots of different products that they don't really talk to each other. Are you guys like -- I think I know the answer, but I just love to hear you guys talk about this.

PriceVantage, is it a novel solution? Or is it the uniqueness is that you're able to leverage the data you have to offer a much more powerful solution versus what's on the market now? Because if it's door #2, I guess, does that imply that there's a long tail of products you can sort of take share from, from like legacy one-off like software products that dealers are taking? Like what are some other areas you might be able to push into if that assumption is correct?

Sure. So yes, I mean -- well, let me back up to just offer a reminder, which is the 4 pillars, the 4 categories of dealer products that we see huge opportunity in our inventory, marketing, lead conversion and then market and competitive data. Most of our products historically have been in marketing. All of our products across -- and we have products in each of those pillars, and I'll give some examples. All of those products benefit from the data that we have in our marketplace. And the specific data in the marketplace that is so helpful are the retail trends, the inventory trends and the pricing trends. And embedded in retail is demand.

And so the reason that that's so important to call out is because every decision a dealer could make in any of those pillars is going to be much smarter if they're able to say how much that car will sell for and how to best sell it. So the example is if a dealer is able to know that a particular car can sell in 10 days for $32,000, then they'll know that they're willing to spend $30,000 for it at wholesale. Because they're looking for a $2,000 GPU. That's an oversimplified example, but hopefully, you get the point.

And so any products that we build related to inventory are informed by this like golden piece of data, which is how much that car will sell for 30 days out when they're ready to retail it.

Similarly, in conversion, if we're able to tell them how to convert interested parties to buy that car better than anyone else because we see the demand trends and we have the deep information on the consumer because they engage with our conversational AI, then they're going to convert those much better and make much more money on their same investment in our marketplace.

So we're bringing this data set that's borne out of the $0.5 billion data points a day that we garner in our marketplace to each of the pillars. So if you look at inventory, as an example, we talk about 5 or 6 key things that need to be done in inventory, which are sourcing, stocking, appraising, pricing, merchandising and syndicating. PriceVantage is just the first step in that, and we believe we have that same data advantage in all of those categories within inventory and inventory is north of $1 billion business in the U.S. alone.

And so when you look at the how is PriceVantage value prop today, it uses real-time data and recommendations. It's predictive, not reactive, and the results are proven in real time, which is another benefit. So you make a pricing decision, you make a pricing change, we tell you, we think this will grow your leads by 30% and you see those leads grow by x percent the following day. And so VINMax, another product that we introduced in marketing. VINMax says, we can see the retail demand trends, and we know it's going to be hard for you to sell this VIN if you don't market it more aggressively. So VINMax says, we don't need you to make the decisions at the dealership anymore. We will make the decisions for you and prioritize the toughest to sell VINs based on the demand data that we have.

So when you look at the new products we've introduced to dealers in the last 12 months, PriceVantage, New Car Exposure, again, New Car Exposure is based on knowing the demand trends for new cars. VINMax, which I just mentioned, Shopper Signals and Performance Insights, those all are based on having a better sense of what's going to sell at what price and to whom than anyone else. And that's why dealers, more people at the dealers are tapping into our system, 28% more activity on our system versus a year ago because it's giving them the intelligence to make these smarter decisions.

Okay. Can you just talk about then dealers' perception of data and dealers' trust in data versus 3, 5 years ago and sort of how things have kind of come to where they are now? And is there room for dealers to even become more data dependent? Like on scale 1 to 10, are dealers at like a 6 now? Are they at a 2? Like where have they been and where are they going to and where are they now?

I think it depends on the dealer, maybe to state the obvious. I think the dealers have -- dealers tend to be data-driven. I think the challenge that many of them would say they face is they get disparate data from disparate systems and it doesn't always fit. It doesn't tie together. And so that's one of the key things that we think a platform like ours that has this singular data layer across all of it helps solve.

So I think sophisticated dealers have been a 7, 8, 9 out of 10 user of data. But the challenge they would say is this source data says one thing and this source data says something else. And so in our products, when there's a recommendation made in PriceVantage, it's the same recommendation that would be made in our marketing -- in our listings product or if it is making a suggestion to do something to a price on a particular VIN because that VIN is not moving well, that is coordinated with the VINMax system, which says, actually, you know what, don't move price on this. It's more economical for you to market it more.

And so -- and again, this is where the quick feedback loop comes back in because with a lot of other disconnected systems, if something says you should do that, you merchandise a car this way and they make a change, they may not see the results, whereas on our platform and with our products, any change they make is directly observable in their leads and listings performance.

Thank you. I would like to turn the floor over to Jason for closing remarks.

Thank you. Thanks, everyone, for joining the call this evening. Thank you very much for the questions. And as always, we always want to thank our customers for their trust in us and also our employees that work so hard to build and deliver all of these wonderful products and the great results. Thanks, everyone. Have a great evening.

This concludes today's teleconference. You may disconnect your lines at this time, and we thank you for your participation.

Jason, thank you so much for your time.

Thank you.

We're going to get into it. And so Jason, when we see the stock trading at these levels, it kind of pops out on the screen as one of the most attractive valuations across the Internet sector, not only within online autos, but all of Internet. What do you think has been most misunderstood about the trajectory, whether it be your recent results, guidance or the broader AI narrative?

Thanks for the question, and we'll try and talk over our neighbors here. So hopefully, everyone can hear us. So it's hard to say what may be misunderstood. But I mean, if I take the different elements of what you just talked about, from a results perspective, we're very proud of our growth, most recently grew 15% in Q1 year-over-year, have had multiple years of double-digit growth now, and that's largely through product innovation and product expansion.

Our EBITDA is growing just as nicely. We have really strong margins in the 30s. And I think maybe actually something that might be missed is just the strength of our EPS growth and free cash flow per share growth over the last several years because we bought back so many of our shares. So EPS from '23 to '25 grew at a CAGR of north of 50%. On the guide, we guided to, again, double-digit growth in '26.

We also talked about a little bit of margin compression that is really intended to maintain our product innovation velocity especially as we move into software and data, serving dealer -- our dealer customers with software and data as well as marketplace, which that requires significant product builds, but that's also helping support our growth of wallet share with our customers. And then on AI, we think of it, and I'm sure we'll talk more about it today, but we think about it in a few different dimensions.

One, we're positioning ourselves really well with the LLM. So we are winning in our category for representing well in the AI piece of the funnel. Two, the trust of our experience, shopping for a car is a multi-month experience. We have extraordinary proprietary data and integrations, which we'll talk more about.

But we have the trust of dealers and consumers to go through that multi-month journey, which is important in AI. And then lastly, and maybe the most exciting is that we're building AI native tools ourselves for both consumers and dealers and AI products that are leading the industry. So we're not just sitting idly while AI advances quickly, we're actually, we believe, at the front of that curve.

And when you think about how AI changes the car buying experience over the next 3- to 5-year period, where does CarGurus fit into that change? And how does the consumer kind of change their experience?

Yes. We've defined the consumer shopping journey in 3 segments. There's research, consideration and then purchase. So research is determining what type of car one might like. Consideration is which of those used cars, typically, could be a new car, too, though, which of those cars is best for you and then purchase is your experience with the dealer to actually complete the transaction. And so we're focused on advancing each of those -- there's certainly threads that tie across all of them, but those tend to be 3 pretty discrete experiences.

And we're transforming that experience into a full AI modal experience for our consumers. And so I think the 3 themes of things that are going to change most with that AI evolution for us is, number one, it's moving from curated information to an expert guide that makes recommendations. That's really powerful and useful to the consumer, and we're seeing them adopt that and really embrace it and really engage with it. Number two is it will reduce the human effort.

So shopping for a car will still take a couple of months. Today, on average, it takes 3 to 4 months. It's still going to take time because consumers need to line things up. They need to do their research, they need to feel confident. But the time that they invest during that 2- or 3-month period, we think will be less because agents can do a lot of that work. Agents can do the comparison. Agents can seek out alternative options for cars.

Agents can start to engage with dealers in useful ways for both sides. And then the third theme is it will optimize the result for both sides. So there's a lot of information asymmetry and a lot of confusion in car shopping. AI and agents can help reduce that. So it gets to a better outcome that is a positive outcome for both the dealer and the consumer. I don't think that will necessarily have people buying more cars than they otherwise would, but we do think that it will lead to better outcomes and better deals and transactions that are in total, better for the car industry.

Got it. And you have an early ChatGPT app integration. Are there any learnings from that so far? We hear about higher conversion from these types of LLM experiences. Is that something you could speak to or any other learnings initially?

Yes. So just to briefly put it in context, as I said, we're doing well in positioning ourselves with the LLMs. We are the #1 traffic destination in the auto category from the LLMs. We are #1 in visibility on the LLM search engines in our category. And then we were the first -- as Michael just mentioned, we're the first to have an app in the ChatGPT app marketplace.

The traffic that we get from that channel is still about 1% of our total traffic. So just to put that in perspective. It does, though, as you said, convert well. It converts about 2x what our normal traffic does. A lot of that traffic comes into a product we have, which is our AI virtual assistant called Discover. And the consumer engagement with Discover, while it's a small percent of our total audience, their engagement is deep.

They have approaching double-digit prompts through a search process and a recommendation process that is now seamless between the research phase and actually recommending cars. And the amount of information that they give us is extraordinary. They're not searching for a make, model, trim. They're searching for what their family situation is and the weather and what type of car and how they wanted to drive, and we're then recommending the make, model, trim out of that. And that's really important because we're then parlaying that information into packaging a better consumer profile for the dealer.

So then when that consumer ends up connecting with the dealer through us and walks into the dealership, if that dealership has been able to get their salespeople to really leverage our platform as much as they can, they tap into something called Shopper Signals, and they'll then see that Jason has come in and he's looking for a particular car, but it's because Jason has a family with 3 kids and a dog and lives in New England and is looking for this type of performance, and they can use that to convert those leads much better to sales.

And so conversion is double through LLMs, which is a big leap. How are dealers reacting to learning about this and changing how they might maybe monetization or just how they're operating their ad spend?

So we have -- I talked about the 4 dealer pillars, and one of them is conversion, converting leads into sales. And we have a number of features and products that we offer to dealers to help them convert better. So Shopper Signals, which I just mentioned is one. Another one is a product called Digital Deal, which allows consumers to do a number of elements of the transaction on our site before they walk into the dealership. They can get a trade-in value. They can put down a deposit. They can set up an appointment.

They can buy other F&I products from the dealer. So then when they walk in, they're much further down funnel. Couple that with Shopper Signals and you are much closer to a sold car and really understanding that customer. The challenge -- there's no shortage of data. The challenge is actually behavior at the dealership. It's dealerships facilitating and training their staff, which typically has high turnover to actually leverage these tools so that they can have a higher conversion rate.

We have a group called Dealer Performance Partners, and they go in and work with hundreds of dealers a year for a day or 2 to make sure that they're getting the most out of our platform. They're using all the data and tools that we have to offer. They're leveraging best practices and they're understanding their competitive set better than they currently are. And it's not uncommon for that group to double the conversion rate of that dealer customer, which is extraordinary. So AI is a part of how dealers can do that, but there's actually -- it's not just AI, it's more around behavior change.

Yes. And we've been hearing -- we just had an e-commerce panel where we're hearing about different conversion-related data points and talking through categories, and it seems like the conversational search experience through LLMs lends itself to more complex categories, less consumables, higher velocity, more kind of complex. So would you go so far to say that autos is probably one of the better use cases for conversational search across categories?

Autos is very high consideration, second typically to homes, and it's very complex because literally no 2 used cars are alike, make, model, trim, mileage options, condition, location, et cetera. So -- and hence, that's why it's a multi-month process. The much higher upper funnel LLM experience to begin that process, I can definitely understand why that would be a good use case for it.

But I actually think that there probably comes a point, I haven't thought about it this way, where it becomes so complex that you actually need that period of research, you need that confidence building exercise that you have to go through. And so we don't believe that car shopping is going to be a zero-click experience. It just is far too important of a decision.

And you saw this years ago when Google introduced what are called VLAs, Vehicle Listing Ads. It went from just links to sites to a carousel of cars. And that has become a really wonderful marketing channel for us. We perform very well there, and consumers will sometimes select a car from the carousel, but then they come through to us to do the research. And we believe that LLMs are similar that it's hard to imagine that a consumer will gain enough confidence in what has been a multi-month process to make a decision at that superficial level. And furthermore, the LLMs don't have access to all the data that we do.

Right. And that data advantage, CarGurus being having the most data out of any platform online or offline, how does that lend itself in an AI ecosystem? I know there's a lot of questions about data ownership and if there's a disintermediation risk there, web scraping and things like that. How do you protect your moat?

So yes, so data is certainly a part of the moat answer, but it's not the only part. Data alone is not enough. But we do have proprietary data. So we -- again, just to set some context, there are about 42,000 to 45,000 dealers in the U.S., about 65,000 dealers total in the 3 countries in which we operate, U.S., U.K. and Canada. If I talk about the U.S. for a second, 42,000 to 45,000, we have about 26,000 who are paying us, and we have over 30,000 on our site because we have a freemium model.

So we have the most dealers, the most inventory, the most paying dealers. We also then have the largest consumer audience with, by far, the most sessions. And so on any given day, we're collecting about 0.5 billion data points around pricing, inventory and consumer demand. We also get feeds from all of those dealers, and we are integrated with many systems at those dealers. A lot of those feeds are unstructured data.

They will describe a car. We then turn that into an ontology that helps us understand exactly what make, model, trim options are in that car and allow us to calibrate and compare that to other cars. We do pricing validation. We do deduplication. And so all of that, plus all of the trends that occur over time, demand trends, a consumer profile that builds over time, those are all things that cannot be gleaned from scraping.

And so that is all data. It's -- but it's data over time, some of which is proprietary, and it's what we do with that, that makes that a moat. You then build that into a trusted experience, and we think that is the moat actually. It's the trusted experience that you need both sides, dealers and consumers to commit to in order for people to have confidence that, that's where they want to transact.

I want to go back to your recent Q1 print. How should investors think about the key drivers behind your Q2 and full year guidance that was updated? What are the main puts and takes from here?

We guided to the year in Q1 and we -- at the end of Q4, and we kept that -- we didn't change that guide. So we've guided to double-digit growth this year again. And we -- and the underpinnings of that, and I think you may -- we may be getting into sort of the elements of 2 of our important KPIs in what drive our revenue are number of paying dealers and then a metric called QARSD. QARSD is quarterly average revenue per subscribing dealer, so how much dealers pay us. There are a handful of key drivers in QARSD that have long runways and are really cranking right now nicely.

So the underpinnings of the revenue guide are double digit of primarily QARSD but also rooftop growth. On the margin side, we guided to a little bit of margin compression. That's a temporal, not a structural thing. It is a function of as we move into software and data and as we want to maintain our product innovation velocity that we've achieved, we've introduced more products in the last 18 months than we probably had in the last 3 or 4 years.

So that velocity and expansion requires investment, and we're really leaning into that. We are definitely getting efficiency, workflow efficiency, a ton of engineering efficiency from AI and agents internally. We're parlaying that or we're focusing that though on productivity enhancements rather than on focusing on margin in the near term because we've seen as those -- as that product velocity accelerates, we're getting more engagement from dealers, and it's giving us the license to introduce products in these other pillars.

Can you go into some of the specifics of the products that you're investing more heavily into in the second half?

Yes. So on the dealer side, we continue to innovate in our marketing category, which has been our bread and butter. So one key example there is new car exposure, which allows us -- allows dealers rather to market specific new cars in addition to or in more sophisticated ways than they have in the past. That's a very timely product right now because new car inventory is building up on dealers' lots. New car affordability is an issue, is a concern with consumers and dealers are trying to find ways to move that inventory more. So marketing continues to see innovation.

Inventory is a category that I'm incredibly -- we're incredibly excited about. It's a big category. It's between $1 billion and $2 billion of spend for dealers relative to a $3.5 billion U.S. marketplace spend. So it's significant. We had introduced some free products there in the past and dealers just absolutely devoured them. And so that gave us the positive signals we wanted to introduce a pricing product. We introduced in Q4 of last year, it's called PriceVantage. We shared that between PriceVantage and New Car Exposure, both of which launched in Q4.

We expect those to grow 15x this year and be 8-figure revenue stream, the 2 of them. So that has hundreds of dealers, and we're seeing really high engagement. On conversion, I talked about Shopper Signals, Digital Deal, there's huge opportunity there. We are still pre stand-alone commercialized product. And then data, what's really nice about data and market intelligence is that makes every other product smarter. Dealers are very competitive. They need to be. It's a competitive arena. And so we are giving them insights and intelligence around their competition and around the market that they literally just can't get anywhere else. And so that is -- I think of that as sort of an umbrella or a layer over all the other products that we're introducing that help those products become more effective.

Got it. Can you unpack the dealer ads portion of your growth and the mix between kind of macro tailwinds versus more execution-driven gains? And where do we stand today in terms of penetration of the overall dealer TAM?

So in the -- as I mentioned, in the 3 countries in which we operate, there's about 65,000 dealers. We have about 35,000 paying dealers. So we're just over 50% penetrated. We're the market leader in the U.S., and we are the market-share-gaining #2 in both the U.K. and Canada. Canada is, I believe, at a tipping point where we are generating lead quality or lead volume rather, quantity that is, in many cases, on par with the incumbent, Auto Trader Canada, such that the largest dealer group in Canada called AutoCanada recently announced that they fully switched from Trader to us.

So we're a little over 50% penetrated. We're gaining share in all 3 markets, both in rooftops and in spend. And we think that the momentum that we are building by focusing on productivity rather than short-term margin expansion is a winning formula. In all 3 markets, we're considered the best ROI that helps us all sleep very well at night because at the core, before we even add any of these other features and products, we know that we're delivering significant value to them.

And can you unpack some of your QARSD growth algorithm recently? It's been very strong. Are new products contributing to that? And what are the biggest components there?

So QARSD, as I mentioned, is effectively how much a dealer pays us. Our QARSD in the U.S. is around $7,500 a quarter. So dealers -- the average dealer pays us $2,500 a month. That's the sort of baseline way to think about it. That's been growing at high single-digit, low double-digits for many, many quarters now. The drivers of that, there are 2 top drivers. One is upselling to higher package tiers and the other is cross-selling.

So if you look over the last year, upselling is our #1 driver. If you look over the last quarter, cross-selling new products is our #1 driver. We love both those drivers because those are a function of our innovation and our product expansion. Upselling is typically now a function of we are adding more and more features and value to higher tiers, and that is causing dealers to upsell into higher package tiers.

Cross-selling would be PriceVantage, Sell My Car, New Car Exposure, all of the ones that -- Digital Deal that we've talked about that are discrete, monetizable, add-on products. So those are the top 2. Lead quality and quantity is a driver. So we deliver a lot of customers to dealers, and those are in the form of traditional leads, which are e-mail, phone call, text chat. We also are sending a lot of consumers to walk into the dealership. That proxies for that are that we're sending them link -- clicks to their website. We're sending them people who have clicked on the map and directions.

And then a recent very high-growth channel is in our app, we have something called Dealership Mode, which gives consumers a ton of value when they're in the dealer. It helps them compare cars. It helps them understand financing, gives them a lot of information in a particularly anxious part of the process. And the number of consumers who are checking in at the dealership in Dealership Mode has grown very quickly recently.

And the most compelling stat is that about 80% of those who are checking in did not submit a lead. So that's showing this is another avenue of value that we're delivering to dealers that we have historically not been getting credit for. So the audience quantity that we're delivering and quality is paramount. And then we do have unit price to pull on. So we are considered the highest ROI.

For most dealers, when you look at the survey data, we are the highest volume of leads, highest quality leads and highest ROI. Pricing in this industry is not standard. It depends on where the dealership is, how big the dealership is, the package tier that they're on a variety of things. But there is a unit, which is a unit of currency, which is cost per connection or cost per lead. And in many cases, we are still below on that metric, our smaller, less innovative competitors. And so unit prices have fueled only a couple of points of growth for us per year over the last handful of years. We don't want to get greedy. We think it's still there in the long term for us to pull if and when we choose, but that's not a focus for us.

When you think about Dealership Mode specifically, and you gave a really interesting stat there about the 80% of customers not submitting a lead. I guess you talked about that trust advantage that you bring for consumers. What are you unlocking for those consumers? And is there, like, a path to monetization of that over time, a greater monetization?

It's a good question. So today, we don't monetize the consumer at all. And well, we don't monetize the consumer at all, and we don't have near-term plans to monetize the consumer. The relationship that we have with the consumer is long term over the duration of their search. And it's increasingly as we build more of these AI features and elements and then pull them together into a more cohesive AI mode, we really are becoming their expert adviser soup to nuts, start to finish.

And our Discover has memory, for instance, and that memory feeds into their sort order, and that feeds into Dealership Mode, and that feeds into Search A that they have and Search B that they have with 2 different dealers. And so we are really becoming their guide through the whole process. And we don't think that we need to monetize that because if we're their guide for the whole process, then that becomes infinitely valuable to the dealer. And if we can win over the consumer, then we will win the dealer has always been our philosophy.

Got it. And how do you assess your evolving competitive landscape with AI accelerating development when we think about, especially, increasing focus online from more traditional offline dealers that are investing there.

And is there a specific competitive angle that you're thinking about with that or competition?

I think when we look at where the consumer will be in 3 to 5 years with their shopping journey, I think everyone is investing towards that. How do you feel about your place in that evolution versus competitors?

We -- I've said the word a lot today, so maybe I'm beating the drum too much, but it comes down to trust and confidence. And we feel that if we can -- by creating the smartest, most efficient, most trustworthy, most data-intensive experience with the most inventory, the most dealers, the most seamless connectivity that, that is increasingly what consumers are expecting.

I think you're -- I'm sure we're all reading a lot of the same content that is talking about how when the novelty of AI starts to wear off, the scrutiny of what can I actually trust starts to set in. And we're seeing that with dealers and consumers. But with dealers, when we make a recommendation on how they should price a car, increasingly, they're starting to ask, well, why?

How did you come to that, that recommendation. And I think consumers are starting to ask those questions as well. And so by having the ability to start from a position of trust because we have a contractual relationship with the dealer or all of the dealers, most of the dealers and then reinforcing and substantiating why we're making the recommendations that we are to consumers, we're going to continue to earn that trust.

The early adopters of AI are incredibly informative test cases. They are still very early, and it is still very small. And so for all of the AI features that I've talked about here, and we've seen this in a lot of our e-commerce peers, they're fantastic use cases and there's really deep engagement, but it's on a very small percent of our total user base. The average user for us is still using the drop-down menus to search for make, model, trim. So we're trying to push that because we do think it's a better shopping experience, but it is still the tip of the spear.

Are there any questions from the audience for Jason? I want to ask quickly, as we wrap up about capital allocation, you've repurchased a large portion of your share base since 2022. How do you think about it from here as we evolve into this more AI-driven ecosystem and just how you allocate, whether or not you invest in product development or elsewhere?

Our 3-tier approach is always how much should we invest in the business largely for innovation. And we've talked about that a little bit, but I would say we continue to invest intelligently, but we think heavily there, and we're not easing up on that despite all the efficiency gains we're getting in the business.

We like how we're expanding into different pillars. We like how we're really evolving or revolutionizing the consumer experience on our site. So we're going to stay with our foot on the pedal there. Next is M&A. We have acquired a few companies in our tenure. We hope to acquire more. So that's the second use. And then the third use is returning capital to shareholders. And in the last, I think, 3.5 or 4 -- 3 to 4 years, we've bought back almost 30% of the company, almost $900 million worth of shares.

The way we think about it is on a free cash flow yield. And when we were at 7% free cash flow yield, we thought that was attractive. And now I think we're closer to 10% free cash flow yield. So we think in these times when we might be misunderstood, as you said, or we think we have a long runway of solid durable growth, and we think it's a good price, then we're going to get more aggressive. So in Q1, we bought back $175 million worth of shares. We have a $250 million share repurchase approval for this calendar year.

All right. Any other questions for Jason?

Who owns the inventory systems? The dealers, they manage the inventory themselves or is there one dominant I hear that, that provides an inventory management system?

There's not one dominant. There is -- there's 2 layers. There are inventory -- true inventory management systems, some -- one that skews toward price, which is one of the larger ones is called vAuto owned by Cox. So there are inventory management systems. And then there are syndication systems that pull from hundreds literally of small inventory management systems and then aggregate and syndicate those out. But I would say vAuto is probably the largest in that category. And then there are a dozen-or-so other sizable ones, and then there's a long tail of a couple of hundred.

And do you have all of the data for all of the [indiscernible] all.

We interact with -- yes. We have -- I believe it's 85-or-so percent of the inventory in the U.S. on our site, which is the largest of anyone. And that's by working with virtually all of those hundreds of feeds. And the information that comes in from those feeds, as I think I mentioned before, is very unstructured. The way you describe the same car that he may -- if you 2 had the same car in theory, you would describe them very differently. We take that unstructured information and turn it into something that can then be compared to other cars for the consumer, which drives our deal ratings and our instant market value and all of those things.

Do the dealers more willing to deliver that data to you [indiscernible] 5 years ago? Is there any pushback about -- no?

No. No. That's the underpinning of how they market their cars, and they know that we do a good job reflecting their cars. We give them advice then on how to merchandise them better, how to market them better, how to think about pricing them. We really try to empower them to be as successful as they can with that inventory that we're showing for them.

And the 15% that you don't have that is because large automobile dealers do themselves or...?

No, we have all the largest dealers -- the top -- we have all the large dealers. No, it's the -- it's probably the opposite. It's the very small ones or it might be dealers who have a very high-priced strategy because they might be in a rural location and they know they're high priced. But by putting their inventory on our site, we have a deal rating, and we would say that's an overpriced car.

And so their mentality is why would I have you say my car is overpriced? I'd rather just not have you say anything about my cars. So that's -- there's also a segment of buy here, pay here, where we think the price they're showing is not an actual fair validated price. And so we don't accept those. So there's a few segments like that.

All right. We are out of time. Thank you so much, Jason. Really appreciate it.

Thank you.

Thanks, everyone, for joining. My name is Rajat Gupta, member of the Automotive Equity Research team at JPMorgan. Very pleased to have with us today, CEO of CarGurus, Jason Trevisan. I'll go through like a few quick questions with Jason. We'll keep opening it up to the audience in between? Or you could also just send your questions through the online portal, and I'll ask anonymously. So with that, Jason, thank you for being here.

Thanks for having me.

We'll get right into it here. So start with like a softball. Maybe just start with some of your new product rollouts. Can you walk us through some of the new and existing products? You've rolled out a ton over the last few months, quarters? What are you most excited about? Let's just start there.

Sure. So we have described the -- through our customer lenses, the dealer workflow, we serve both the auto dealers as well as car shopping consumers.

On the auto dealer side, we've defined four pillars of workflow areas that they have. And we've historically operated in one, which is marketing, but there are three others that are very important to dealers. One is inventory another is converting the leads to sales and another is marketing competitive intelligence or data associated with that.

So we, over the past few years, have been preparing for expansion and introduced a number of free products in each of those pillars. And we've now started to turn that activity that we've had there into monetizable products. And inventories where we're most excited we're furthest along. It's probably the biggest category of those.

In inventory represents a dealers' ability to stock, figure out what types of cars they should stock, source, where they can get those cars, appraisal, how much they should pay for those cars, pricing, how much they should charge for those cars and then merchandising, how well they merchandise and market them.

So end of last year, we introduced PriceVantage. It's a pricing software solution in that category. We've announced recently that we have several hundred or many hundreds of paying customers there. We said that, that and another product in the marketing category that helps customers market their new cars. Together those products were going to go from starting or launching in Q4 to north of $10 million combined in revenue this year. So they're starting 0 to 1, but they're growing very fast. So I would say we're very excited about inventory category, and we're excited about the others, but they're earlier.

On the consumer side, I would say we're very excited at just how AI is transforming the whole consumer shopping experience. And so historically, it was a consumer comes in, they need to know what type of car make-model trim they want, they would use drop-down filters and do what is now considered a fairly manual search. Now AI is helping them determine in a conversational way and in one that travels with them over time and remembers them and has memory and personalization, helping them determine which make model trim they should want based on what they're looking for. We then have AI that's helping with sort and personalization throughout the shopping process. And then lastly, when they get into the dealership, we have a product called dealership mode that uses AI that helps them navigate that whole experience, which historically has been really clunky and one that they don't enjoy. That's helping them understand what other cars at that dealer they should consider or what their financing options are that they should consider. And really helping them make sense of that through an AI companion.

So we've accelerated how quickly and how frequently we're introducing new products on both the consumer and dealer side, and that has really helped drive engagement from our customers as well as maintain the double-digit growth that we have.

Understood. That's a helpful rundown. Just quickly shifting gears to AI. Has there been any consideration for partnering with LLMs. And how do you think this could improve your competitive position in the market? Where do you think you are? Where do you think you could go with some of these partnerships eventually, if at all?

Yes. So specifically, the partnerships, we, in a measurable way, are doing exceptionally well. As measured by visibility in the organic results of LLM search results as well as in traffic generated from the organic search results of LLM and auto, we're #1 in both. We were the first to introduce an app in the Chat GPT app ecosystem. And we're talking with both and working with both Chat GPT and Google on their paid models as those start to emerge.

So we are positioned very well in all things, LLM search related. It's a journey, not an end process. And so we're going to continue to work at it. It still represents a very small percent of our traffic, like low, low double-digit percent of our traffic. But it's growing quickly, and we want to make sure that we stay at the front of that, just like we say -- have stayed at the front of Google and anything that's happened with search over time and mobile and apps, and it's a new channel, and we are determined to be the winner in it.

And could you break down how investment headcount is currently allocated across these key cost buckets and just how the integration of AI and automation, could you just reshape this over the next several years?

We have talked about our biggest areas of incremental investment this year being product and engineering, #1 and go-to-market, number two. And then international is an area where there's just a lot of momentum for us. And so we're, I would say, keeping our foot on the gas there.

But in terms of the first two products and engineering, we have accelerated the -- as I said before, the pace and volume with which we're introducing new products, and we want to maintain that accelerated pace. And so we are investing in the ability to build products and get them into the market faster, build better products, build more differentiated products and continue that pace because we do not take our leadership position for granted, number one. And number two, we're moving into new categories. We're not just adding on a feature or a product in an existing category, like marketing. And so moving into these new areas takes investment. We're going from a standing start in some of these.

And so that takes people, that takes technology and we're committed to them because we see that these categories represent a more than doubling of the TAM that we have in our current product set, and we're going to be aggressive in going after it.

On the go-to-market side, it sort of follows suit, which is these are new products. They are sometimes sold to a different person at the dealership. They do require some behavior change at the dealership. They do require dealers thinking about us differently than their marketing partner or lead provider. And so we need to educate them more extensively than we have in the past. We need to onboard them, and we need to change their behaviors in some cases. And so we have to invest in that.

We are getting material and growing efficiency gains from AI. And those two things, efficiency gains and growing investment to keep an accelerated pace are both happening at the same time right now. So a lot of the efficiency gains that we're getting, which we're seeing through higher, faster, more frequent code releases, more PRs per developer, more lines of code written more use of code creation and code assistance. I mean all the metrics that we're looking at, we're seeing is more productive. But -- and we're also seeing token use drives and token expense rise. And so we're in the very early stages of that. So we're seeing both more efficiency, more productivity but it's not translating into we need fewer people or we're going to spend less money. We're in this period where we're looking to accelerate with new technology. And so we're going to lean into that from an investment standpoint.

Outside of engineering, we have stood up an AI architect or an AI solutions team that's made up of AI architects that are building agents for all the functions across our company. And that is leading to some areas where we're already seeing dramatic efficiency gains. Most other areas, we're seeing early signs of efficiency gains, most of which were parlaying into productivity, but eventually, that can lead to efficiency. And so as I describe it, our headcount forecast for 2028 is lower today than it was 3 months ago. But we do not have plans to reduce our headcount in the near term because of the productivity gains we're seeing from AI.

Got it. That seems like a good sweet spot. Maybe you can build Agentic platform for the CFO role at some point.

That would be good.

Any other broad strategic organizational structural changes you foresee?

I mean if you go back a couple of years, we reoriented to our customers much more a couple of years ago. And so we restructured our product and engineering teams to mimic the workflow and the life cycle of our customers. That, coupled with a strong focus, again, for the last couple of years on speed. And we are just running far more tests getting far more signals much faster, and that's helped lead to faster product introduction.

The other thing that we've been doing for the past few years, which is now paying a lot of fruit in addition to the AI efficiencies that I talked about, our investments in platform. And so it is much easier for developers to work in our platform now. It's much more modularized, it's much more bite sized. They can run better tests, safer tests, better QA, faster QA than they ever could before. So those have been two key structural changes historically.

Going forward, what we're working on is how do we structure the org in a way that aligns incentives and speed with, say, on the dealer side, a multi-pillar environment. So we need to and are building products that help dealers with -- I'll just take this one example I've been using marketing and inventory. In some cases, that's a different person at the dealership. We hope that's a different wallet. There's a different onboarding process. One is marketing technology, one is software. It's a different product at the end of the day. There's different security needed for both.

And so we have to figure out how we can do both of those well and efficiently in a way that is tied together because they do mutually reinforce each other. We gave a lot of data points on our -- in our script that showed the dealers that are using PriceVantage and embracing it are seeing immediately faster turn times, more VDP views, they're seeing better performance in our marketplace and getting better ROI out of it. So we need to lean into the benefits of those two things being on the same platform, but also recognize that delivering them is different, and we don't want to create a heavily matrixed organization. We don't want to create more touch points than we have to, but we also want to enjoy the benefits of the distinctions of them because it does double our TAM. It does help us tap into other wallets. So we're thinking a lot about org structure for an expanding product suite.

Understood. That makes a lot of sense. Going to the 2026 comments and guidance, are you able to break down the drivers of the 10% to 13% revenue growth in guidance across QARSD dealer growth? How should we think about that breakup? And then maybe you can touch on individual items in more detail.

We don't break it down more than that. And in fact, there's a few reasons that we don't, but one is that we incentivize our team to net MRR growth, Monthly Recurring Revenue growth. And sometimes that comes in the form of more rooftops and sometimes that comes in the form of expansion of existing. Typically, when you look at our QARSD growth rate, which is the -- QARSD is a quarterly average revenue per customer. So what they spend with us and the rooftop growth rate and you add those two together, you typically get roughly our revenue growth rate.

The bulk of our growth for the past handful of years has been QARSD growth. And -- but we've been growing rooftops as well, even in tougher market environments. When it's harder to grow rooftops, we still grow faster than our competitors. So we continue to gain share.

I think you can expect that the bulk of our revenue growth will continue to come from QARSD. It's not that maverick of a supposition. And within QARSD, what gives us confidence and comfort is that there are several drivers, most of which have long runways ahead of them. So our drivers of QARSD growth historically have been we've upsold to higher packages, and we do that by adding in new features and value to higher packages. We have added on products. We have a growing product suite, so it's easier to add on products when you have more products to choose from. And I'll come back and give a little more detail on each of these. Lead quality and quantity, again, that's got significant runway. And then lastly is unit pricing, which we've been very measured on, very steady and, I would say, not aggressive on unit pricing increases.

So if you look at each of those, and I'll give some examples, upselling to higher package shares, we just introduced something called shopper signals. That is a very rapidly adopted, widely adopted tool that gives dealers a robust 360 on each consumer that we're sending them. When dealers use that and tap into that content, they convert much better. We offer that to only premium tier packages for free. And any dealer that uses that who's spending the same with us that they sent before using it is getting materially more value out of our platform because they're converting better. That's a strong incentive for dealers to upsell the higher packages.

Add-on products, we've talked about PriceVantage, so that can help them move into the inventory category. Even if we stay in just marketing, we've introduced new car exposure. So franchise dealers who, frankly, right now, are having a hard time selling new cars, for a variety of reasons. Prices are up, affordability is challenged, gas prices are high, interest rates haven't come down. And so them being able to more aggressively market new cars is very compelling to them. So we're always introducing new products, and we're getting smarter, back to the AI and go-to-market point, we're getting smarter at which products are -- which customers are ripe for which products. So we're being much more solution sale oriented with that.

And then lead quality and quantity, of all of the users who come to our site, only a low single-digit percentage of those users convert to a lead. Today, we know that also a low double-digit percent of those users click on a link to the dealer's website, click on map and directions, and now we're seeing a growing number of consumers who are opening dealership mode in the CarGurus app on the dealer's lot, and we know that 80% of those people never submitted a lead.

So we now have multiple channels that we're sending consumers to dealers for and dealers are, frankly, really paying us for only traditional leads. But all of those are growing the lead quantity. As we add in things like shopper signals, it's growing the lead quality and that adds up to the number of cars we're helping them sell.

At the end of the day, we want to help dealers sell more cars, and so if we can do that through a variety of channels and increasingly prove the strong role we played in that, then that is a natural rising tide for something like QARSD.

You mentioned earlier you touched -- you just mentioned like doubling the TAM or something about it. So how should we think about just the upside to QARSD over time? Any way to like just contextualize that how many cars does the dealer need to sell to make up for their cost? I mean, where are we in that equation today to think about just how much -- how long can this rise continue.

I'll take it from the angle of top-down rather than bottoms up, just dealer segments there's a lot of different types of dealers out there. But from a top-down perspective, so the doubling TAM comment is dealers today in the U.S. only, we operate in the U.K. and Canada, as most of you know. Dealers today spend about $3.5 billion on marketplaces. That's in the context of spending about north of $20 billion in marketing. So they're spending only 15-or-so percent of their total marketing budget on marketplaces. And marketplaces tend to be the best ROI of any of their marketing channels. So I think there's room to grow for marketplaces within their overall marketing spend.

We, CarGurus are about 27%, maybe 30% of their total marketplace spend. We're the largest audience, were the best ROI or the highest volume of leads. And by different measures, we capture about half, 50% of consumers' mind share and time spent on marketplaces. So that to me says we have headroom just to capture our commensurate share of marketplace spend and marketplaces have had room to capture more of marketing spend because of the relative ROI -- its strong relative ROI compared to billboards and Facebook and banner ads and other things that they're doing.

The doubling the TAM comment is then when you start to think about these three other pillars, inventory conversion and data. Today, they spend over $4 billion -- $4.5 billion in those categories. We have less than 1% of market share or wallet share in those categories. But I think in each of those, we have a prime position and a right to win to enter each of them because today, we're already giving two dealers free products in each of those, and we've started to monetize inventory. And so that represents a bigger opportunity set than what we have in marketplace today. And so top down, I think Marketplace has a runway for growth. Marketplace category is a run rate for growth. And then we're only just getting started in the other pillars.

No that's very helpful. And maybe just touching on international a little bit. Obviously, the QARSD differences are pretty stark. What's -- how should we think about the ramp there over time? And what's your vision to ramp that up?

Our vision is for the foreseeable future to continue running the playbook that's working so well for us. So in each of those markets, there's a large incumbent in each of those markets, we're the fast growing #2, who in Canada, we believe we're a viable challenger to the market leader. And we've had many groups and some of the largest groups in the country, say that they're transitioning exclusively to us. So we think there's a tipping point happening in Canada. We continue to grow our lead volume. We're always focusing on lead quality and now we're introducing more products there. We just introduced Sell My Car in Canada. We're taking a lot of the learnings and innovation that we've done for the U.S. and started to introduce that in Canada.

And in the U.K., it's a similar story, but earlier stage. We are the further #2 in that market, but we also continue to gain share there in terms of traffic and paying dealers and wallet share and so forth.

And we think that in those markets where there has been just a single dominant provider, bringing in a transparent, ROI-friendly unbiased consumer experience is extremely welcome to the industry. We proved that that's the case in the U.S., and we're proving it in those markets as well.

Understood. Before I just touch on a few more points, I just want to see if there were any questions in the audience.

I don't see any in the portal as well. Okay. I'll just go on. From a dealer, I mean, obviously, a lot of good detail on QARSD. From a dealer perspective, we've seen some pressure on profitability recently on the used car side and some on the new car side. Are you seeing any sense of caution from the dealers maybe more near-term question on how they're managing their spend.

We have seen some. The trends, as I touched on earlier, consumer affordability is a challenge right now for a variety of reasons. New car prices remain really high. And so a lot of new car buyers are being pushed into used car just for affordability purposes. Used cars also have been a tough spot for dealers to get their hands on to source. There's a few factors for that. And so if you look at -- so used car in the used car segment, used car inventory is down a little bit and demand is high. And so they have an easy time selling what few used cars they have. Margins are also down a little bit. which is an interesting dynamic that you wouldn't expect necessarily from that. And so when dealers have had an easy time selling cars, they are less inclined to want to invest much more aggressively in marketing, for instance.

And so these things are finite. I mean, they change and they evolve over time. Inventory will be easier to get more off-lease more cars are coming off lease. So that might help. Affordability could change, gas prices could change a number of things can change. It's a very dynamic industry. Pricing can change. We help them change pricing. And so we're focused, what gives us confidence is that we're focused on optimizing for dealers to run a great dealership. And they can do that and protect their margins and turn a lot more volume, even when being more transparent with the consumer, which we help them to do as well.

Got it. And just last point on just '26 guidance. Maybe help us like break down the 150 to 250 EBITDA margin compression. I mean you talked a lot about continuing to invest, but also seeing a lot of efficiency. How should we think about -- are you able to like double-click on where that compression is coming from? Which specific areas? And is this just like a onetime reset that we should think about from a margin perspective? .

Yes. The double click is what I mentioned earlier. So it's product and engineering. Number one, go to market. Number two, to onboard and educate customers. And then number three, keep the momentum in international. And yes, to your second question, I would think of it as a onetime push in investment that is to maintain this elevated frequency and velocity of product introduction and not a structural reset in any way.

Understood. Okay. Going back to just some of the -- the supply/demand dynamics that you mentioned, which is hurting dealers right now, but there is a lot of used car supply expected to come to the market, especially from the off-lease side later this year. And presumably, dealers we'll get a look at a lot more inventory from these maturities. Are you doing anything different from a company perspective to position dealers for that? Or you just think these tools will ultimately take care of that -- or I'm just trying to think like, is there anything you're doing geared towards that supply recovery that we're going to see from off-lease over the next, I think, 2, 3 years combined?

Yes. We don't -- yes, but maybe for a different reason. We don't try to build things for a moment in time as you would think. But everything in the inventory vertical is going to help with everything that you just described.

I mean if you think about the five areas that I talked about in inventory, it's telling them which cars they should have on their lot. And it's not based on what cars are selling in wholesale in the past. It's based on what the supply dynamics in their region will look like in 60 days or in 30 days or in 90 days. and based on how effective they've been at selling X, Y, Z types of cars over time or even more recently.

So as those off-lease cars come in, we are telling them which ones they should be more aggressive, more or less aggressive on which ones they should stock, how their inventory mix compares to what it should look like in 60 days, if they want to achieve their goals of turn time or margin per unit or volume or whatever the case may be. How much they should pay for those cars, where that is price advantage has that aspect in it embedded. But as we build out more appraisal capabilities, it will tell them how much they should pay for each specific car.

Merch pricing is telling them what they should charge for those. And again, it tells them, if you want to turn the car in fewer days, 5 fewer days, 10 fewer days here is where you should price it. If not, if you want to maximize the margin and depending on what your holding cost per day is, I mean they can set their goals and our software then executes on the right set of actions in order to optimize to that goal. And then same with merchandising and then all of our marketing products help them then optimize their odds of marketing that car most effectively.

But you're not necessarily assuming some sort of cyclical upside from a market perspective to your business this year necessarily right? Is there any of that -- because in the past, when we're coming off of '23, off of the inventory shortage. There was this massive uptick in new car inventory across the dealership ecosystem and that benefited marketplaces. We saw that with new cars, but this is the first time we're going to see that with the used cars. So next 2, 3 years. But is that something that you're preparing for as like some sort of cyclical upside to your numbers? Is it embedded in your guidance?

Well, our guidance for this year was based on everything that we knew and believed would happen when we set the guidance, and we haven't updated that. So I would say, yes, to the extent of what we believe would be the case then. A lot of the trends that I just talked about have like amplified a bit in the last few months since guidance. And so as a result, you've heard us talk about Sell My Car is now one of the more interesting products in our portfolio for dealers, and we wouldn't have anticipated that.

And so what makes us really excited though is that as we grow our portfolio of products, we start to have products that are more or less compelling no matter what the dynamics in the market are. And it becomes a much more stable and universally appreciated platform.

I mean, given like the space of -- you've seen some efficiency gains, you the space of product rollout really get better. dealerships tend to have a lot of inefficiency in many other areas, many other systems. Are there one or two areas you could point to where it's like an easy adjacent or not easy, but just an adjacent opportunity that your platform can expand into going forward?

At times, we think that expanding into four pillars is too ambitious. So we don't have thoughts of turning that into five or six. I think keying off of some of the word choice in that question. I mean conversion is a key area where we see night and day between dealers who are good at it and dealers who are not good at it.

And an example I'll give is we have a team called Dealer Performance Partners. They work with dealers on using our platform to the highest extent possible, making sure that our system is integrated with the right other systems in the dealership and executing best practices on lead handling or customer communication or lead conversion, a variety of things. It is very common for that group to work with a dealer over a couple of days and double their conversion rate, double their conversion rate. So dealer that goes through that spends the same amount on marketing and sells twice as many cars. That's game-changing for the dealer and shows the sort of latent value in our platform. And that's a natural extension for us because these are the customers that have come from us. And so shopper signals is a step in that direction. Digital is a step in that direction.

So think about a consumer that comes through digital deal, Consumer A connects to a dealer via an e-mail, that's it. Consumer B on our platform puts down a deposit, sets up an appointment, get to trade and value, gets financing and buys three other products online from that dealership on our platform. They interacted with our Discover, so they exchanged a variety of information about themselves to help them navigate to the best car possible. That turns into a shopper signal. That consumer walks into the dealership. The salesperson looks at that. they walk in and they say, "Hey, Rajat, I've got all your information here, your deposit, et cetera, et cetera. I know this about you. I know this is what you're looking for. And this is why you landed on that Blue Subaru Outback. Congratulations". Let's get this closed. That conversion rate is multiples of what it would be otherwise and that's all content that came from us.

Understood. That's helpful. One last one before we run out of time, just capital allocation. You obviously have a very strong free cash flow profile provides the optionality on M&A buyback. Clearly, you've been a big -- you bought back a lot of stock. But -- are there some areas you'd be willing to consider on the M&A side? Just given the focus on data intelligence, any core competencies or data modes that you like to in-source.

M&A targets or in-sourcing data modes? What was -- can you say the second piece again?

Yes. M&A targets in a way to like just get more of that some of the...

I mean I think look, our -- on the dealer side, our strategy of moving into those three new areas to have four areas total, we think is ambitious and robust. And so anything that can help us in that move faster in that or more intelligently in that is interesting to us. And the good news is that there are a lot of technology and data companies serving auto dealers in those areas. And so -- and many of whom we're currently partners with and have wonderful relationships with. And so those are all possible areas for M&A.

The consumer side is, of course, an area for M&A, but it's less obvious that it could be integrated as well. And so M&A is always something that we've looked at. We've always said, first, we invest in our business. Second, we consider M&A and look at it aggressively. And third, if we have excess capital and think the share is a good investment, share price of good investment, then we'll do that. And so I think we're going to keep that hierarchy. And we hope that we can conduct M&A for acceleration or to get us into areas that we otherwise wouldn't get into.

Clearly, AI is creating the ability to build organic software faster. But there are definitely companies out there that have proprietary data that's interesting. There are companies out there that have done things through integration and building capabilities that would take a long time or that have garnered dealer trust that would help us to accelerate by buying.

Understood. Any other questions from the audience? You have a minute left.

No. It looks like we should end it there. So thanks, Jason, for joining this. And thanks, everyone, for joining.

Thank you very much.

Good day, and welcome to the CarGurus First Quarter 2026 Earnings Conference Call. Please note, this event is being recorded.

I would now like to turn the call over to Kirndeep Singh, Vice President Head, Investor Relations. Please go ahead.

Good afternoon, and thank you for joining us. With me on the call today are Jason Trevisan, Chief Executive Officer; and Sam Zales, President and Chief Operating Officer. We will be making forward-looking statements, which are based on our current expectations and beliefs. Those statements are subject to risks and uncertainties, and our actual results may differ materially.

Information concerning those risks and uncertainties is discussed in our SEC filings. We undertake no obligation to update forward-looking statements, except as required by law. Please refer to our press release and our investor presentation on the Investor Relations section of our website for a reconciliation of GAAP to non-GAAP measures.

I'll now turn the call over to Jason.

Good afternoon, and thank you for joining us. We delivered strong financial results in the first quarter with 15% year-over-year revenue growth to $244 million and adjusted EBITDA up 17% year-over-year with a margin of 33% as our product investments helped drive sustained growth while maintaining healthy profitability. This performance was driven by premium tier adoption greater usage of our AI-powered products, lead growth and net dealer additions. That strength was especially evident internationally in our U.K. and Canada markets where revenue grew 39% year-over-year, reinforcing our ROI advantage driving share gains in both markets. At the foundation of our product innovation and increasing customer engagement is the data layer of our marketplace.

We ingest roughly 0.5 billion first-party consumer shopping signals each day across demand, pricing, inventory and engagement. Today, we apply these proprietary marketplace signals in our AI-enabled analytics platform to build products that enable dealers to make better informed decisions and help consumers shop with more confidence and achieve better outcomes. That work shows up and our 2026 strategy through 3 value-creation drivers.

First, we are expanding CarGuru's offerings into integral parts of the dealer workflow, connecting inventory, marketing, lead conversion and data pillars through mutually reinforcing products. Second, we've begun transforming car shopping into a trusted AI-led journey from research through consideration and purchase, giving consumers greater confidence and more benefits from using CarGurus.

And third, we are disciplined in our capital deployment with the aim of growing long-term earnings power and stockholder value. I will now walk through our first quarter progress across each of these drivers. Driver One, expanding CarGuru's offerings into integral parts of the dealer workflow connecting inventory, marketing, lead conversion and data pillars through mutually reinforcing products.

Our marketplace has long held dealers market inventory and generate high-quality customers from our largest and most engaged car shopping audience. and dealers ranked CarGurus #1 in ROI among listing sites, which was recently reaffirmed in a survey of a select group of dealers. We're building on that marketing foundation by expanding across additional key dealer pillars. Inventory, lead conversion and data and have begun embedding predictive intelligence more directly into those dealer workflows to support better informed decisions. That is showing up in greater wallet share gains and higher engagement as dealers use CarGurus for more of their day-to-day work.

In our inventory pillar, we're focused on helping dealers make better decisions about which vehicles to stock and how to price them to optimize their margin and turn time goals. PriceVantage, our first specialized software solution sold ala carte, has already reached several hundred paying dealers since its October launch. It uses our first-party demand signals, market data and machine learning to generate win level pricing recommendations aligned to dealer objectives.

Our top engage dealers using PriceVantage saw a 117% improvement in turn time relative to their top 5 competitors on CarGurus and a 47% median increase in daily VDP views as they adjusted pricing faster and with greater precision in response to live market conditions. We're also making our data available wherever dealers make daily inventory decisions. Our browser extension gives paying customers access to CarGurus' pricing signals directly within inventory management systems and auction sites, putting our data in the dealer's workflow at the point of action.

The aim is to make CarGurus intelligence a more embedded input in how dealers source and price vehicles. Usage of the extension tripled quarter-over-quarter, demonstrating that more daily inventory decisions are happening with CarGurus data in the -- in the conversion pillar, we're focused on how dealers turn interest into sales. In April, we launched shopper signals in our premium tiers. Shopper signals brings together first-party shopper behavior across cards, including browsing activity, vehicle preferences, dealer engagement and digital deal actions.

Leveraging AI, shopper signals gives dealers a richer view of each shopper's intent, preferences, and activity so they can take a more customer-centric approach to follow-up. That can include prioritizing high-intent leads, understanding which vehicles best match the shoppers' needs, and suggesting similar inventory when the original vehicle is no longer the best fit or available.

Driven by our leading inventory and consumer demand data, we're able to help dealers better prioritize and personalize their engagement with buyers. Improving lead conversion and driving better ROI for dealers. This value is already resonating with our dealer base with over 8,000 dealers engaging with the feature since its mid-April launch.

In our data pillar, we're focusing on giving dealers a clearer view of their performance relative to their competitors and the market broadly, so they can make more profitable decisions. We launched performance insights a monthly report that gives paying dealers a more actionable view of marketplace performance. The report benchmarks dealers on leads and VDP views per vehicle as well as average turn time relative to comparable dealers in their area, then pairs those benchmarks with win level recommendations to help dealers improve merchandising, optimize inventory mix, price more intelligently and drive stronger performance.

Dealers receiving performance insights had a 76% open rate and made an average 59% more price updates than the period prior. By providing data that contextualizes a dealer's performance predicts the outcomes of recommended actions and allows dealers to see the results in our marketplace, we believe we are gaining reliance on our data and engagement with our platform.

Leveraging our marketplace is the foundation of our dealer value proposition, we're expanding our platform by embedding intelligence more directly in the day-to-day decisions dealers make across these 4 pillars. That has shown up in stronger dealer engagement with our platform, growth in products per dealer and U.S. car SID up 9% year-over-year.

Driver number two, transforming car shopping into a trusted AI-led journey from research through consideration and purchase giving consumers greater confidence and more benefits from using cargoes. AI is reshaping how consumers discover and research vehicles. But for the second largest purchase most people make we believe confidence still comes from trust in a shopping process.

In Q1, we deepened our role across that full journey, engaging shoppers earlier through AI-native discovery, giving them personalized tools to evaluate options and building more confidence at the point of purchase. That has driven stronger engagement across cards with monthly uniques, sessions, time spent and steps taken on the platform all growing year-over-year.

Our app reflects that momentum as well. It remained our fastest-growing traffic source, and we are #1 in the auto category in active users and time spent. We categorize the consumer journey into 3 stages. Research, consideration and purchase. In research, shoppers often start broad before narrowing to a model and then a specific vehicle.

A very small but growing percentage of shopping journeys now begin in AI native environments, and we want Carter's inventory and data to be present where that discovery starts. In Q1, we launched a CarGurus' app inside ChatGPT, becoming the first automotive marketplace in the U.S. to integrate live local inventory directly into the platform. When shoppers express purchase intent and ChatGPT, they can see local vehicles, vehicle details and deal ratings without leaving the conversation, then move directly to CarGurus to contact the dealer or submit a lead. AI-driven search traffic remains small, but is growing quickly, and CarGurus has shown up well in traffic because shoppers can access broad inventory, rich vehicle data and trusted deal ratings directly at the point of discovery.

That matters because conversion from this traffic remains meaningfully higher than traditional channels. We also continue to evolve our on-site generative AI search experience discover from a research feature into a more effective shopping guide. We added new car inventory, improved relevance using shopper profile data and made it easier for shoppers to move from conversation to specific vehicles and dealer contact all within the experience.

That helped move shoppers from exploration to evaluating real inventory and connecting with dealers more quickly while generating richer demand signals across our platform. Discover users continued to grow quarter-over-quarter and more notably, leads grew 52% quarter-over-quarter. In consideration, shoppers are deciding whether a specific vehicle fits their needs and budget. Comparing options and understanding the trade-in value for their current car.

We advanced Sell My Car by introducing a conversational AI flow that makes it easier for consumers to get valuations and offers with less friction. Sell My Car continued to gain traction in the U.S., improving the selling experience for consumers and giving dealers efficient access to sourced inventory.

In the first quarter, we launched Sell My Car in Canada where early engagement and adoption were off to a strong start. In the purchase step, we're focused on the moment when a shopper is physically on the lot and making a final decision. Dealership mode is built for that moment. When a CarGurus' app user arrives at a participating dealer, the app brings pricing, deal ratings, payment estimates and AI-powered vehicle comparisons directly into the in-store experience.

Recently, we began rolling out capabilities that give dealers more visibility into verified online activity. Those signals are more predictive of purchase intent and a traditional lead and help us close the loop between online activity and in-store outcomes. Since Q4, daily lot visits have grown 67%. We believe dealership mode has also become a leading driver of app downloads expanding our owned audience at the point of purchase and strengthening the connection between shopper engagement and dealer value.

Across research, consideration and purchase, we're using AI and first-party signals to make CarGurus more relevant effective and personalized across the consumer journey. As consumers engage with us across more steps, we have more opportunities to help them move from research to action while improving the signals we translate into dealer value. We believe that it's driving more leads and more down-funnel shoppers while giving consumers clear information and greater trust in the outcome.

Driver number three, disciplined capital deployment with the aim of growing long-term earnings power and stockholder value. We generated strong free cash flow and that cash generation fuels our capital allocation strategy. We will continue investing internally where we see the strongest long-term returns, preserving flexibility for targeted M&A and returning capital to stockholders.

2026 is an investment year by design, with increased product, technology and development spend, including AI-driven innovation as well as higher sales and marketing investment to support the introduction and adoption of new dealer products and growing our consumer brand and audience. We expect these investments to modestly weigh on margins in the near term, but we believe they will drive more durable long-term growth and a healthy margin profile. In the first quarter, those investments funded the launches and expansions I described earlier across our dealer pillars and consumer journey.

They also funded the infrastructure and internal capabilities needed to launch more products and features faster. As well as the security, governance, standardized systems and dedicated teams required to scale AI responsibly across the business. Today, a majority of employees use AI in their daily workflow. And in Q1, we standardized AI systems across engineering and product and established a dedicated AI solutions team to identify, prioritize and transform high-value workflows with measurable ROI.

We measure engineering productivity using a mix of internal metrics and external benchmarks. Those indicators show that AI solutions have contributed to a 20% year-over-year productivity lift and a 50% lift quarter-over-quarter among AI laggards, expanding our overall product development capacity. Beyond engineering, AI-powered content creation helped drive a roughly 30% increase in unpaid leads year-over-year and AI has also helped our sales team spend more time with customers, reach more dealers and better educate and engage them on how our solutions can help them make better informed decisions. These investments are improving speed, capacity and execution across the platform as we move from AI-assisted workflows toward a genic AI capabilities that we believe will support more complex end-to-end work in the future.

We've applied the same discipline to capital returns. Since 2022, we have repurchased approximately $896 million worth of shares or about 29% of our shares outstanding while continuing to grow revenues and profitability. In 2026, our Board authorized a new $250 million share repurchase program through year-end. And in the first quarter, we deployed approximately $175 million under that authorization. We will continue to repurchase shares when we believe it is an attractive investment after funding product and technology investment and maintaining balance sheet strength.

We are allocating capital with discipline toward product, technology and AI-driven innovation while continuing to return capital to stockholders. We believe this approach will expand our capacity to innovate, strengthen profitability, increase cash generation and create long-term stockholder value.

Q1 showed progress across our dealer and consumer value creation drivers. On the dealer side, we expanded further into the dealer workflow across inventory, marketing, lead conversion and data. On the consumer side, we extended our role across more of the shopping journey from AI native research through purchase. What connects these efforts is the intelligence layer of our marketplace. The real-time demand, pricing, inventory and engagement signals across CarGurus help us build better products for dealers and better experiences for consumers. Over time, we expect this to increase dealer reliance on our platform, grow our share of dealer wallet and deepen consumer engagement across the journey. We are supporting that strategy with disciplined capital allocation. investing in the product, technology and AI-driven innovation, we believe can drive the strongest long-term returns while continuing to return capital to stockholders.

Now let me walk through our financial results. followed by our guidance for the second quarter and full year 2026. First quarter revenue grew 15% year-over-year to $244 million, above the midpoint of our guidance range driven by strength from our strong year-end 2025 bookings as well as continued momentum in our international business, with slight moderation in OEM advertising reflecting the typical first quarter step down.

In the first quarter, U.S. CarGurus grew 9% year-over-year, and we added 963 paying U.S. dealers year-over-year. We continue to increase our dealer base while taking greater wallet share driven primarily by upgrades and broader adoption of add-on products with modest contribution from like-for-like price increases and higher lead quantity and quality.

For the second quarter in a row, new product adoption was the largest driver of the sequential increase in CarGurus. Our international business outperformed in the first quarter with revenue up 39% year-over-year driven by favorable FX and overperformance in U.K. advertising revenue. I will now discuss our profitability and expenses on a non-GAAP basis.

First quarter non-GAAP gross profit grew 14% year-over-year to $225 million. First quarter non-GAAP gross margin was 92%, down about 80 basis points year-over-year. First quarter non-GAAP adjusted EBITDA grew 17% year-over-year to $80 million, above the high end of our guidance range and adjusted EBITDA margin was 33%, up about 60 basis points year-over-year. due in part to a favorable item related to a retroactive change in Canadian tax law.

First quarter non-GAAP operating expenses totaled $152 million, up 13% year-over-year reflecting higher sales and marketing expense and higher product, technology and development expense versus prior year as we invest to continue the accelerated pace of AI product introductions.

First quarter non-GAAP net income per diluted share attributable to common stockholders was $0.58, up 21% year-over-year. We ended the quarter with $72 million in cash and cash equivalents, a decrease of $118 million from the end of the fourth quarter, primarily driven by $175 million in share repurchases in the quarter, partly offset by adjusted EBITDA.

As of the end of Q1, we had $75 million remaining on our 2026 authorization. I will now turn to our guidance for the second quarter and full year 2026. We expect our second quarter revenue to be in the range of $247 million to $252 million, up between 11% and 14% year-over-year, respectively.

For the second quarter, we expect our non-GAAP adjusted EBITDA to be in the range of $77.5 million to $85.5 million. We expect second quarter non-GAAP earnings per share to be in the range of $0.57 to $0.64 and diluted weighted average common shares outstanding to be approximately $91 million. Turning to the full year. We are reiterating that we expect 2026 revenue to grow in the range of 10% to 13% year-over-year. We still expect full year non-GAAP adjusted EBITDA margins to compress approximately 1.5 to 2.5 percentage points in 2026 relative to 2025.

With that, let's open the call for Q&A.

[Operator Instructions]. Our first question is from Chris Pearce with Needham.

I think if I look at the midpoint of the guide for 1Q and where things landed, I guess I'd just love to get some color around, was it planned spending on product development that was pushed out or spending coming less than you expected on that line item because of the AI tools you're talking about? I guess big picture, margins came in a lot better than I think we were all expecting. So I just kind of want to get a sense of why how we should think about it going forward? And just kind of the full year guide you still have intact?

Thanks very much, Chris, it's Jason. Thanks for the question. Yes, so as we called out one thing in particular, which is the retroactive change in Canadian tax law. There were a couple of other timing items or smaller onetime items that if you were to aggregate all of those, it certainly would have put us closer to the midpoint. And as a result, I would say, no sort of structural changes or surprises which leads us to having reiterated our guide. Timing of spend could be -- that's typically marketing and brand as an example, and midpoint is where we aim.

Okay. Perfect. And then just on -- if I think about the new tools and you're talking about premium care adoption it seems like these are mostly data-related tools. I'm just kind of curious, are dealers still sort of holding off on -- holding off -- but on digital deal, are they still sort of they don't want to go full Carvana mode here. They're still afraid they want to get the customer in the room for the loan side of the world. I guess because that seems like an underutilized that's probably my term, but I guess I'd just love to hear how you're winning so much on data, but customer dealers are still sort of afraid to rip the band it off on full digital.

Thanks, Chris. It's Sam Zales. The digital deal continued to grow in the quarter. We didn't put any specific metrics on it. It continues to be a very important tool for more than half of our customers. So we're really proud of what that product has done. Reminder that consumers are still saying, I want to do much of the purchase process online, but then I'd like to still go to the store.

The huge majority are still saying, I want to touch, I want to feel the product in the store. So only a few percentage points in the market are still buying fully online. So we're still gaining more adoption in digital deal. And when we think about it, we think about high-value actions. That's what we talk about here in our digital process. So a consumer who says, I'm putting some trade-in information in. I'm going to share information on my budget. I may get financing as part of that.

You heard about the launch of Shopper Signals in April, which has taken off dramatically with our customers with 8,000 customers using it now either daily or it comes out either in a daily process, a real-time process or integrated to the CRM. Again, we're providing that information as well. So through multiple products now we're helping dealers engage with high-value actions that brings the consumer further down the funnel. And they're more ready to buy, and that's why we're achieving the ROI #1 status according to dealers in the market.

So I hope without giving you the information, it is still growing. That's still a big part of our business, but the dealer knows the consumer still wants to be in store, and that's why we're building multiple ways for dealers to interact with that high-value action, if you will.

Our next question is from Rajat Gupta with JPMorgan.

You mentioned in your prepared remarks, moving from AI-assisted workflows towards agentic AI capabilities. Was that a comment around just your own internal product development and engineering or was that a comment on from a customer standpoint and helping them move in that direction? Just wanted to clarify that, then I have a follow-on.

It's Jason. The comment, I'm pretty certain was about internally, and it's both engineering and brought outside engineering product as well as outside for workflow purposes. But if I'm remembering the comment correctly. But absolutely, you can assume that, that same evolution is happening with our products already.

Got it. Got it. Are there 1 or 2 use cases around agentic that you can talk about from a customer standpoint that's in your pipeline, like dealer operations or consumer support, pricing workflows, any hint you can give us there?

Yes. I think the best examples of AI in general today in our products right now are with pricing and inventory where we are reading the market in real time and making real-time recommendations to dealers and then using that to predict and then hold accountable to those predictions, the implications of those changes. And so that is a -- in terms of the definition of AI versus agentic, I would say that's on the margin.

Beyond that, I would say -- it is a broad-based evolution from AI to agentic that is going to manifest in a variety of ways, and we'll be sure to give more detail as there's more sort of pronounced and obvious examples at the customer level.

A quick follow-up around the U.K. We noticed or we read some reports around potential dealer churn or customer churn at one of the larger players in that market. I'm curious, is that in any way benefited your share growth in the region? Or was that an opportunity at all for you to engage with customers more.

Thanks, Rajat. It's Sam Zales, and thank you for the question. I would say our general success in both the U.K. and Canada just continues on a very, very aggressive and successful path. You saw the numbers in terms of growth in both dealer adds and Card there was probably a positive impact. We heard about it too connected into that market very deeply with our team in Europe. And I think Trader did stub it so a little bit but that didn't have a massive effect on our business.

What we're doing is we're following our playbook from the U.S. We're listed as the ROI #1 in that market, and we have been for a long time with those dealers. And so that is a an element of their saying, how does that compare to the big market leader in that market. They're charging a lot more, and you're producing more from an ROI perspective. So I want to test and use your product, and that's why we're new customers and building cars because we're also building new products in those markets and following the playbook here from the U.S.

So were our visitor base is growing faster than our competitors we're just making the investments to make a product better standing for our customers and hopefully delivers better return and makes them come to us. So I wouldn't put too much emphasis on the incident or the communication in the marketplace that was caught up in November. We're just continuing to seek to outperform, and that's our push.

Our next question is from Andrew Boone with Citizens.

I wanted to ask a big picture question in terms of data. It seems like dealership mode, it seems like Discover are both kind of data plays. Can you speak to just new ways that you guys are trying to interact with consumers and maybe the additional amount of information that you're gaining from more AI-type interfaces? And how you guys think about the basically the deeper integration with customers and what that unlocks for you? And then secondly, if I look at traffic for the quarter, maybe it was down in the U.S., maybe that's weather. Can you just speak to the weather impact and what you guys saw in 1Q how that kind of ran through the model.

Thanks, Andrew. Yes, I mean, it's -- data is certainly a key enabler for us, but I also want to emphasize that it's not just data. It's a lot of things around the data and that we do with the data that allows us to create the products that we're creating that we think are certainly unique in the market. So from a consumer journey perspective, historically, we've been very good once the consumer knows which car they want.

Discover is the first example that we had that is growing certainly in its use cases very rapidly and improving quite a bit, helped upstream, help consumers determine which type of car would be good for them. But as that's expanded, it's done much more than that. It not only does that better, but it also then helps them find and help them navigate through what we would have traditionally called this sort of consideration and then connect with the dealer. And then dealership mode once they've connected with the dealer, helps them understand the full totality of information around that dealer that they would need and would benefit from in buying the car.

And so in doing that, we learn an extraordinary amount about our customers. They, in a conversational exchange, they'll share a lot of information because the more information they share the better the response will be, the better the more informed it will be. And so we're able to take that and help them not only through empowering them, but also through guidance and recommendations help them find a better car, better for them, better match.

And then when they're into the dealership mode, we understand that dealer well. We understand things about inventory and pricing. And demand trends and merchandising and car comparisons and financing. And these are all things that we know better than anyone we would argue because we have the most retail data and confuses the heck out of consumers because it's sort of overwhelming.

And so we try to distill all that down. We try to be there companion throughout and it learns and our AI mode tools have memory and personalization and they travel with the consumer. So it really is changing the game, we think, from a filter-based drop-down episodic hunting and packing, not to be too pejorative of our historical business into a guided tour that answers questions and helps them get to the best answer.

On your second question, you had asked if -- was that a weather impact on dealers?

Yes.

I'm happy to jump in, Andrew, it's Sam Zales. I think if I got your question correctly, we saw visitor statistics rise year-over-year. So we didn't see that impact on our global business and our business in the U.S. continue to grow. I think you're referencing there were storms that did impact dealers' ability to sell a vehicle and have foot traffic walk in during that period of time, just a general phase. And there's a lot going on in our market today with gas prices with consumer sentiment with inventory acquisition hard to come by.

And so in both uniques and sessions, though, for us, we were up year-over-year and quarter-over-quarter. I think that's a sign of our sustenance of a best offering in the market, what we believe is the best offering in the marketplace. Did that answer your question?

Yes. That's great. Thank you, guys.

Our next question is from Marvin Fong with BTIG.

Would love to dig in just a little more on price advantage, not a first -- product is very exciting here. And I appreciate the update here on the number of dealers that are now paying. I just wanted a little more color on the type of dealers that are signing up for this? Are these franchise dealers that are pretty sophisticated. And secondly, where these new installed competitive displacement of an incumbent solution or where -- what's the advantage like the first time using a sophisticated inventory management system. Just as some color to that.

Marvin, it's Sam Zales. Thanks for the question. We're really proud of the growth of the Price Vantage product. here's why it's so different than anything else in the marketplace. It is a profit maximization predictive tool. I think when we see and compare it to others in the marketplace, -- those are risk mitigation look back at book prices to figure out how to stem losses and make those as careful as possible. And I think what we're doing is trying to help dealers find that future looking with our consumer demand data what price should I -- buy the vehicle for what price should I set the vehicle for retail so I can maximize my profitability. It's truly a differentiator.

We're selling to both independent and franchise dealers. We've seen success on both of those customer segments, which we're really proud about. Remember, there are 2 things that happened with this product. turn times, how to improve them for dealers, and you saw the 47% VDP views and 117% faster closes of those sale of those vehicles versus their top competition.

That's truly happening. And then gross profit per unit happens because you find the right balance between pricing and selling those vehicles. So profit is a huge outcome of our PriceVantage program. I think to your question of how we can continue to win share that, in some cases, is a product that we're selling that a customer never had. So they didn't have a product that's in there today and they don't have a pricing tool.

In many cases, it's saying, I have a tool, but this one does something different, as I talked about the predictability of that vehicle. We're tending to complement something that's already there but make it better. And I think the key to that as you're thinking about the chrome extension of this product. So you may have a tool that you're using today, but you put the CarGurus PriceVantage Chrome extension in -- you're looking at a different view of pricing and predictability with consumer demand, and it helps dealers say, I might want to switch because this has given me something that's totally different in the market.

So growth was tremendous there. It tripled quarter-over-quarter, and I think we're going to continue to push for that to be that product that is changing the way dealers are solving their #1 problem in the marketplace. How do I acquire inventory at the right price and sell it at the right price.

Got it. And if I could do a follow-up here. Just on stock buybacks, you guys have done a great job returning capital to shareholders in 2022. I believe you only have $75 million left in the remaining authorization and you guys really -- out even in the first quarter. How should we kind of think about the path going forward, is the plan to just exhaust the remaining authorization over the balance of the year? Or do you have an appetite to reload on even more an authorization?

We're -- I mean, yes, we have that plan in place, and we put plans in place that we hope we can use. At the same time, we as we've shared, don't have an indiscriminate approach to it. We have an approach that says when we think it is a good investment, then we will get aggressive. And so we've never hinted at what we might do in the future in terms of expanding or introducing a new program. So what we've shared is the $250 million. We're through $175 million. So yes, we have $75 million left. And that 75 will adhere to the same philosophy, which is at prices that we think are more compelling, we're going to get more aggressive.

[Operator Instructions]. Our next question is from Joe Spak with UBS.

Maybe just to follow up on that last topic. The cash balance you finished the quarter at models right, I think that's the lowest since 2020. And obviously, you generate some healthy cash flow every quarter, but I would just be curious if you could give us some indication of sort of minimum cash levels, you sort of feel like you want to operate the business at.

Thanks, Joe. We clearly do think about what are right cash levels, what are minimum cash levels. Just to back up quickly, we think about 3 categories of how to use our cash or our cash production power. The first is investing back in the business. So that goes into our operating margin calculus. The second is M&A and powder for M&A and then the third is returning capital to shareholders. As you said, we do generate nice free cash. We have very proud of the free cash flow conversion rate that we have. From a minimum perspective, we think about it as cash we have on hand as well as cash we have ready access to.

And so we have had a line of credit, which is very easily accessible, and that's we think, a prudent thing to do. In terms of minimums, we've never shared what sort of a hard floor is here with a line of credit that makes a hard floor much more sort fluid or almost theoretical threshold because you've got access to so much more. And also, it relies on timing of working capital.

And so the -- we can have pretty large cash swings. We control it. But if it dips down artificially because of timing, then it comes back very quickly because of timing. So we think about risk mitigation. As a balance to us being aggressive when we think shares are underpriced.

Fair enough. And then just -- I appreciate all the commentary on AI and the integration and the apps. But -- and please, if this is sort of unfair because it's sort of too early and please feel free to respond to so. But I am just curious like -- and I know you sort of said adoption is still pretty low, but I think you did say it's sort of growing fast. But are you able to see yet anything about like conversion rates or acquisition costs for dealers as a result of these tools? Or it's just not there yet, and that's not sort of a fair question to answer yet.

I don't know if I can speak much to what dealers are experiencing. And I think even if I did, then you'd see a huge, huge range. But what I would say about us is that it remains a very small percent of our traffic and a very small -- the LLMs remain a very small percent of our traffic. They remain a very small percent of our leads. It does tend to be high-quality traffic. We show very well in the LLM in terms of visibility and traffic receipt of traffic, our share of the traffic that comes from there.

But they're very top funnel, top of funnel. And it is not really a substitute for a marketplace. So whether it's us showing up in an LLM or our app in ChatGPT at Marketplace or some of the or testing paid search now, and we're going to be right there front and center because we tend to be at the leading or bleeding edge of these things. We're showing up well, but they're small. And so in any of those situations, it's top of funnel and it's not a substitute. They need to come to a site where they're going to go through the workflow. They're going to go through the process where they have the trust and confidence where they can compare and look at pricing and deal ratings and we have vehicle history and ontology and so forth.

And so the data can be scrapped. We provide access to some data but that's very different than a marketplace that has inventory, has contracts with dealers. We're normalizing data. We're validating pricing. We're deduplicating listings and managing real-time availability and things like that. And so I think the people who are using it who are the early adopters are getting smarter on it and then they're coming to us to really go through the process. And that's why we're both embracing it in terms of our presence in LLM, but also building our own capabilities and applying AI across our own user experience and dealer workflow so that we stay ahead of the horizontal LLMs as well.

There are no further questions at this time. I would like to turn the floor back over to Jason Trevisan for closing comments.

Thanks very much. I would just like to thank everyone who tuned in this evening, and thanks to everyone who ask great questions. As always, we like to really show particular appreciation to not only our shareholders but also our customers and especially our employees and their passion and hard work that is helping us execute as well as we are today. So thank you very much, everyone. Have a great evening.

This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.

Good day, and welcome to the CarGurus Fourth Quarter and Full Year 2025 Earnings Conference Call. Please note, this event is being recorded. I would now like to turn the call over to Kirndeep Singh, Vice President and Head of Investor Relations. Please go ahead.

Good afternoon, and thank you for joining us. With me on the call today are Jason Trevisan, Chief Executive Officer; and Sam Zales, President and Chief Operating Officer. We will be making forward-looking statements, which are based on our current expectations and beliefs. These statements are subject to risks and uncertainties, and our actual results may differ materially.

Information concerning those risks and uncertainties is discussed in our SEC filings. We undertake no obligation to update forward-looking statements, except as required by law. Please refer to our press release and our investor presentation on the Investor Relations section of our website for a reconciliation of GAAP to non-GAAP measures. I'll now turn the call over to Jason.

Good afternoon, and thank you for joining us. 2025 was a pivotal year for CarGurus. We accelerated product innovation, expanding how we serve dealers and consumers as a marketplace, software and data company, drove significant growth and profitability and returned capital through disciplined share repurchases.

Revenue from continuing operations grew 14% for the full year, our second consecutive year of mid-teens growth and adjusted EBITDA from continuing operations grew 25% year-over-year.

Wallet share expanded, retention reached its highest level in 3 years, new dealer additions accelerated and consumer traffic growth translated into increased lead volume. Internationally, we delivered 27% year-over-year revenue growth, driven by accelerated dealer acquisition, wallet share expansion and strong lead growth.

Our performance was supported by faster, more prolific AI-driven innovation, which increased product velocity and strengthened our differentiation among both dealers and consumers.

We launched more new products in 2025 than in any prior year, embedding data and intelligence directly into dealer workflows and consumer decision-making.

Key launches included dealer solutions like PriceVantage and new car exposure and consumer features like CG Discover and Dealership Mode. We continued bolstering our support of dealer workflows across inventory, marketing, conversion and data.

This product innovation further cements our role beyond lead generation into daily operating workflows, expanding our addressable market into adjacent software and data categories. Early traction validates demand, engagement and differentiation.

Our consumer launches expanded our role across the shopping journey from research through in-dealership decision-making and purchase, driving higher intent engagement and higher converting leads. Our broader portfolio of consumer functionality, combined with the largest inventory selection and a foundation of trust and transparency reinforces our leadership position as the most visited automotive shopping platform in the U.S.

We also executed with discipline. We made the prudent decision to wind down CarOffer while retaining its sourcing technology and data to strengthen our inventory products. Those learnings have informed better pricing and inventory technology solutions that both drive high incremental margins and reinforce the value and performance of our marketplace.

Now I'll walk through our progress across our 2025 Drivers of Value Creation. Driver number one, expanding our suite of data-driven solutions across dealers' workflows to help them drive more profitable businesses. In 2025, our most critical dealer metrics achieved impressive growth.

Consolidated QARSD grew 8% year-over-year. Global paying dealer count increased by 2,399, Add-on product adoption rose nearly 25% year-over-year, engagement and platform usage grew and retention strengthened. Underpinning the products and intelligence we deliver to dealers is a growing body of data that we're translating into higher fidelity insights through AI.

On average, last year, we ingested approximately 0.5 billion first-party shopper signals each day, translating them into real-time consumer demand, pricing and inventory insights our dealer customers are leveraging for measurably improved performance.

In 2025, dealer data insights became central to dealer workflow with 60% of global paying dealers using these insights across their operations. What began as validation of the value dealers ascribe to our data is now embedded in how they make more profitable data-driven decisions.

Building on this foundation, we launched PriceVantage in October, our first specialized software product designed to move dealers from passive data consumption to action-oriented pricing decisions. Early results demonstrated accelerating inventory turnover and increasing VDP views.

Engagement has been strong with nearly 80% of adopting dealers active weekly on the PriceVantage suite of products. Compared with their prior usage of our free pricing tool, PriceVantage users execute 66% more price changes and log more sessions per day, all clear indicators of deeper reliance on data-driven pricing workflows.

Collectively, these efforts embed CarGurus more deeply into dealer operations and decision-making, driving stronger adoption, more consistent action and clear evidence that our intelligence-led solutions improve efficiency and profitability.

Driver number two, meeting the evolving needs of car shoppers by powering a more intelligent and seamless journey. In 2025, we strengthened the consumer side of our marketplace by increasing our reach and the quality of shopper engagement. Traffic grew faster than our primary competitors year-over-year, reinforcing our position as the #1 most visited automotive marketplace.

Nearly half of monthly visitors shopped exclusively on CarGurus, indicating a high degree of reliance on the platform. That behavior translated into results, fueled by lead growth in the U.S., CarGurus-led sales grew year-over-year. And according to a 2025 Clarivoy study, CarGurus influenced 55% of all attributed vehicle sales.

We believe that scale and influence create a stronger foundation to introduce new consumer experiences that deepen engagement and generate richer signals. Our generative AI search experience, CG Discover, continued to scale.

Unlike other tools that simply repackage search results, Discover responds in real time and acts as the decisioning copilot using live marketplace inventory, deep automotive expertise and demand signals to quickly and flexibly answer consumer requests.

Discover traffic grew 3.5x and leads grew 10x quarter-over-quarter. Depth of engagement also strengthened with average session time up nearly 20% and Discover users spending 4.4x more time than regular visitors.

Each interaction generates richer demand and pricing signals, strengthening our data and intelligence layer. We also extended our trusted user support into later stages of the purchase process.

CarGurus was the #1 car shopping app in 2025 by downloads, monthly active users and time spent, giving us scale at the point of purchase.

Dealership Mode now live across all consumer app users moves our role beyond discovery and further into the transaction funnel by assisting consumers on dealer lots. In just the first few months, thousands of shoppers on average open Dealership Mode on dealer lots each day.

80% of app users who visit a dealership lot have not submitted a lead in advance. That means 4 out of 5 high-intent shoppers using our app on dealer lots are not attributable to us. We believe that Dealership Mode creates a clear opportunity to increase previsit lead submissions and drive measurable traffic to dealers.

These investments delivered greater transparency and broader support to consumers for a more seamless shopping experience. We entered 2026 with a more differentiated consumer experience and a stronger foundation to meet shoppers where they are in their journey.

Driver number three, enabling dealers and consumers to complete more of the transaction online, streamlining the final steps of the deal. We scaled digital deal to 13,500 dealers globally, adding nearly 3,800 dealers year-over-year. This growth reflects growing consumer demand for and dealer reliance on workflows that move more of the transaction online and generate higher intent prospects.

We embedded high-value transactional capabilities earlier in the shopping journey, including expanded financing, trade-ins, deposits and appointments. Digital Deal leads with high-value actions increased 78% year-over-year in 2025 and represented approximately 70% of Digital Deal leads, while financing-related leads grew 86% year-over-year, reflecting deeper shopper progression into the transaction and stronger purchase intent.

Overall, Digital Deal leads convert up to 4.7x higher than standard marketplace leads with even greater lift for shoppers located farther from the dealership, delivering higher quality and higher converting shoppers to dealers.

Collectively, these changes are shifting more of the transaction online while preserving the in-person experience that the vast majority of consumers still want. While 86% of buyers ultimately see the car in-person, 83% say they want to complete more of the shopping process from home according to our 2025 Consumer Insights Report.

By meeting that demand, we improve consumer engagement and offer more transaction support for dealers. That's especially important given the nature of the car buying journey, which remains a high consideration decision and often the second largest purchase a consumer makes in their lifetime.

Shoppers want to research and compare options, understand pricing, availability and trade-offs and still negotiate price and test drive vehicles in- person before committing. Confidence and trust matter at every step of the journey. That reality continues to shape how we invest in our platform, brand and own channels and how we show up as discovery paths evolve.

Following the performance of the 2025 Big Deal campaign, we are extending the campaign into 2026 with product-led spots highlighting dealership mode and CG Discover. We believe these experiences bring clarity and confidence to the car shopping process and reinforce the trust consumers place in CarGurus as the #1 most visited automotive marketplace.

We expanded our impressions by 50% year-over-year and direct and owned channels are our fastest-growing traffic sources with direct visits up 16% year-over-year and the app contributing 34% of leads.

We're also leading automotive marketplace competitors in the emerging AI-driven discovery landscape. While AI remains a small share of overall traffic today, in Q4, CarGurus generated more AI-driven traffic than our closest competitors, and these users convert at higher rates, submitting leads at nearly 50% higher rates than traditional SEO in Q4.

To date, AI traffic has been additive to our overall acquisition mix, increasing visibility rather than displacing existing channels. AI is reshaping discovery across many categories. In automotive, the shift has been more measured. However, we are not waiting for it to accelerate. We are expanding AI-driven traffic across both paid and nonpaid channels. On the paid side, we've been early adopters of new AI-powered tools with Google, Bing and Meta with promising initial performance results.

We plan to test emerging AI search ad formats, including those introduced by OpenAI as they become available. On the nonpaid side, we have strengthened and will continue to evolve technical platform best practices and scaling proprietary content so it is discoverable across LLM environments, not just traditional organic search.

We believe our depth of experience and success in audience acquisition across many channels, combined with a disciplined test-and-learn approach positions us well as this landscape evolves.

While AI may shape how shoppers begin their journey, it does not change what they need in a major purchase, clarity and confidence. Even when journeys start in AI environments, consumers still come to CarGurus to validate listings, confirm availability and make data-driven decisions. As discovery paths evolve, platforms with the deepest inventory, broadest dealer coverage, most comprehensive retail data and highest ROI will be best positioned to remain central to the transaction.

We believe our market leadership, data depth and dealer integrations position us to continue serving that role. Stepping back, 2025 was an outstanding year for CarGurus. We delivered strong financial results while deepening dealer and consumer reliance on our products across more steps in the car buying and selling process.

More decisions are informed by our data and AI, more workflows run through our platform and more of the car shopping journey now takes place with CarGurus involved. It was also a year of strong product innovation. The many products we launched in 2025 have shown promising signs of engagement and scaling, giving us confidence in our investments in new product innovation.

For example, we expect the monetized dealer products we launched in 2025 to grow approximately 15x in 2026 and achieve 8-figure revenue levels and exciting exit rates. Entering 2026, our platform is firmly embedded in dealer operations and now serves a larger TAM than a year ago.

With disciplined execution continued investment in AI-driven innovation and proven products to support both our customers, our focus remains straightforward. We intend to execute with rigor, build products to strengthen the dealer workflow and consumer journey while further differentiating CarGurus.

Before turning to our results, I want to address 2 reporting updates. First, we completed the CarOffer wind down in the fourth quarter of 2025. We have presented CarOffer's financial results as discontinued operations in our consolidated financial statements for all periods presented, except for the statements of comprehensive income, redeemable noncontrolling interest and stockholders' equity and cash flows.

Unless indicated, the fourth quarter and full year 2025 results we will be discussing on this call relate to our continuing operations.

Second, beginning in the fourth quarter of 2025, we report our financial results as a single segment following the CarOffer wind down.

Now let me walk through our financial results, followed by our guidance for the first quarter and full year 2026. Fourth quarter revenue grew 15% year-over-year to $241 million at the high end of our guidance range, driven by strength in our subscription-based listings revenue as well as overperformance in advertising and strength in our international business.

Full year 2025 revenue was $907 million, up 14% year-over-year, our second straight year of mid-teens revenue growth. In the fourth quarter, U.S. QARSD grew 8% year-over-year, and we added 1,357 paying U.S. dealers year-over-year.

We continue to expand our dealer base while taking greater wallet share, driven primarily by upgrades and broader adoption of add-on products with modest contribution from like-for-like price increases and higher lead quantity and quality. Robust revenue growth continued in our international business with fourth quarter revenue up 32% year-over-year and full year revenue up 27% year-over-year, driven by new dealer adds along with a modest tailwind from favorable FX.

International QARSD grew 16% year-over-year in the fourth quarter. International dealer count growth surged 14% year-over-year to 8,360 dealers. I will now discuss our profitability and expenses on a non-GAAP basis. Fourth quarter and full year non-GAAP gross profit was $223 million and $842 million, respectively, representing 14% year-over-year growth in each period.

Fourth quarter non-GAAP gross margin was 92%, down about 90 basis points year-over-year. For the full year, non-GAAP gross margin increased by about 40 basis points to 93%. Fourth quarter non-GAAP adjusted EBITDA grew 13% year-over-year to $88 million, above the midpoint of our fourth quarter guidance range. Adjusted EBITDA margin was about 60 basis points lower year-over-year at 37%. Full year 2025 non-GAAP adjusted EBITDA grew 25% year-over-year to $319 million, and adjusted EBITDA margin rose by approximately 310 basis points year-over-year to 35%.

Fourth quarter and full year non-GAAP operating expenses totaled $141 million and $547 million, up 15% and 10% year-over-year, respectively. The increase in the fourth quarter reflected higher sales and marketing expense and investment in new product innovation, as mentioned earlier.

As a result of the CarOffer wind down, we incurred and paid $13.3 million in total expenditures, of which we incurred and paid $5.4 million in onetime cash restructuring charges attributable to discontinued operations. This was at the low end of our previous $13 million to $15 million wind-down estimate.

Fourth quarter and full year GAAP net income per diluted share attributable to common stockholders was $0.56 and $1.96, up 24% and 62% year-over-year, respectively. From 2023 to 2025, this measure has grown at a 56% 2-year CAGR. Fourth quarter and full year non-GAAP net income per diluted share attributable to common stockholders was $0.63 and $2.28, up 17% and 31% year-over-year, respectively.

We ended the year with $191 million in cash and cash equivalents, an increase of $12 million from the end of the third quarter, primarily driven by higher adjusted EBITDA, partly offset by $57 million in share repurchases in the quarter.

In 2025, we repurchased about $350 million in shares, completing our 2025 share repurchase program. Since the fourth quarter of 2022, we have repurchased about $721 million in shares or about 25% of our outstanding shares.

Additionally, I'm pleased to share that our Board has authorized a new $250 million share repurchase program, which will be available through December 31st, 2026, highlighting our commitment to return value to shareholders.

I will now turn to our guidance for the first quarter and full year 2026. We expect our first quarter revenue to be in the range of $240.5 million to $245.5 million, up between 13% and 16% year-over-year, respectively. For the first quarter, we expect our non-GAAP adjusted EBITDA to be in the range of $72 million to $80 million, up between 5% and 16% year-over-year, respectively.

As a reminder, our guide reflects our continuing operations absorbing approximately $1 million in ongoing quarterly CarOffer expenses following the wind down. We expect first quarter non-GAAP earnings per share to be in the range of $0.52 to $0.58 and diluted weighted average common shares outstanding to be approximately 94 million.

Turning to the full year. We expect 2026 revenue to grow in the range of 10% to 13% year-over-year. We expect full year non-GAAP adjusted EBITDA margins to compress approximately 1.5 to 2.5 percentage points in 2026 relative to 2025. We believe that the adjusted EBITDA margin implied by the midpoint of first quarter guidance is a reasonable proxy for the first 3 quarters of the year, with fourth quarter margins expected to be higher due to seasonality. This reflects increased investment in product, technology and development as we plan to continue our accelerated pace of AI product introductions for both dealers and consumers. With that, let's open the call for Q&A.

[Operator Instructions] Our first question comes from the line of Ralph Schackart with William Blair.

Jason, with the strong guide outlook of, I think, 10% to 13%, maybe talk about sort of the visibility and durability of that growth rate. I think it's sort of a key investor question, sustaining these high growth rates in the marketplace business. And then maybe if you could add some perspective as you layer on new products with this reinvestment cycle that you'll go through in 2026, maybe provide some perspective on how those new products be additive to growth over the long term.

Sure. Ralph, so the strong growth, as you described it, is a function of a lot of things. I would say at the beginning or sort of at the core, it's a function of the fact that on our 2 growth levers of QARSD and new dealers across all our markets. So globally, we have good visibility. We have no customer concentration.

And so we can look at the levers, and we have a number of levers that we have been executing really well on that have continued to drive QARSD and are also now driving pretty consistent and nice dealer adds. And so whether it's new products or upsells or growing lead volume and quality, growing connections, you heard us talk about how Dealership Mode is now showing dealers that we're sending them even more consumers that were not submitting leads.

And so sort of at the core, our value prop just continues to strengthen. And it also strengthens in absolute terms. It also strengthens in relative terms compared to our competitors who are not growing at those paces -- yes, at those rates rather. But then on top of that, we also have new products. And so you -- we mentioned a couple, and we called out some of the products that were launched in 2025 and gave some perspective and parameters around how those are going to contribute.

But we also have a number of products before we even get to that, that have continued to grow and differentiate us, some of which we monetize, but many of which we don't. But what they do is they drive retention and they drive engagement, and that allows us to grow our dealer base and have more pricing power.

And so all of that is healthy and strong. And between U.S. and international, we have long runways in both those areas. In terms of new products, so the big thing for us is moving into software and data across the dealer workflow verticals of inventory, conversion and data in addition to marketing. And that's opening up -- that's about doubling our TAM. And you heard us, hopefully, from our remarks, you glean the fact that while they start small, they're growing quickly and they're showing really strong engagement.

One of the levers of QARSD is new products, and that has continued to strengthen as a lever for us. And as we invest more in technology and product, we see that as a nice long-term lever for us.

Our next question comes from the line of Chris Pierce with Needham & Company.

I know you had some stats in there about new products and 8 figures and -- but I just want to understand, is PriceVantage part of the full year revenue guide? Or is that still such a small portion? but I guess I just want to understand if the new TAM is in the revenue guide or something that is further down the road or what's the best way to think about it?

Yes. No, it's certainly part of it because we've launched it. And so we've in our operating plan. And new car exposure was also launched in '25. So that's also included, and that's part of our marketing vertical. So the new TAM that we talk about, the doubling is in the non-marketing verticals, and we're obviously just scratching the surface of it, but we think there's huge potential because, again, as a reminder, all of the products that we build in those areas are reinforced by and reinforce our marketplace.

And so it's not like with PriceVantage d, we're not just another start-up offering a pricing tool to dealers. We are obviously established, but we're a company that has a relationship with the dealer who's coming to them and saying, we have the most data, we have the most intelligence. Here's pricing software and here's what it will do to your performance on our marketplace.

Nobody else can say that. As the largest marketplace, that carries a lot of weight. And so the same applies with conversion, the same applies with data.

Okay. And then just as a follow-up, how should we think about this new margin range? Is this a year of investment specifically as you land in this new TAM and kind of want to have momentum in new product as you kind of try to penetrate this new TAM? Or is this sort of a new normal and we should think about you guys consistently kind of wanting to plow money back into the business to leverage your leadership position?

I don't know if it's binary. I mean I think we absolutely want to do the latter that you said. We want to continue investing in product. In marketplace models, it tends to be a winner take most. Scale matters. Scale helps a deeper, broader platform, drive stickiness and engagement.

And so we're not in it to -- and I know you didn't say this, we're not in it to milk for margin. We're in it to sustain long-term growth and drive healthy margin. And so the comment or the guide rather on '26 and the range that's specifically not guiding beyond that. That's not saying it's a new normal nor is it saying it's onetime.

But I would say that the most helpful perspective maybe to hear from us on it is we're trying to optimize for long-term growth, customer stickiness and a healthy margin business. And if you want to look at Rule of 40 type metrics, we're performing well there the last couple of years and expect to next year as well -- this year as well.

Our next question comes from the line of Naved Khan with B. Riley Securities.

Yes. So I had a question on just this -- the margin outlook for 2026. And I get the -- your thinking behind investing for growth even as you deliver on a least rule of 40 or higher than that. But curious in terms of the areas you're going to invest in between product and marketing, is it going to be more skewed towards marketing? And maybe within that, you kind of mentioned about investing in paid agentic channels. Is that a part of it? And how should we expect that expense to grow?

And then the second question I have is just around the price advantage. Can you remind us on the pricing for this product? And if you're kind of marketing it in a tier in a higher tier? Or is this a separate add-on? And what's the recent -- what's the latest kind of uptake on this?

Sure. In terms of where we're investing, so some of this will be a repeat from last quarter. But -- so in 2025, a lot of our investment growth, let's say, was around sales and marketing and account management. As we look at '26 and even in the back half of '25 in our marketplace business, but into '26, we are investing much more behind the momentum in building new products. And so that's coming in the form of investing more in product, technology and development. We are -- and I'll come back to that in a second. So that's sort of bucket 1. In '26, bucket 2 is international and bucket 3 is account management.

And so really focusing on the first bucket, though, we have become much more prolific in introducing new products. We have gotten a lot of efficiency from AI. And so we're taking that efficiency and translating it into productivity.

And if you think about the fact that we are moving from not only introducing products and marketing, but moving into entire other verticals at the dealer, inventory conversion, how they convert the leads that we send them and the customers that we send them and then market and data intelligence, we're standing up new software products and new data products, and we're going into new areas of the dealer.

And that is not a small lift that to compete well there, you have to build robust products. And so -- it's a very product growth-driven mindset that we have and early measurement is through adoption and engagement and then pretty rapidly it's through things like retention and revenue.

In terms of PriceVantage price point, we haven't given specifics on it. What we've shared last quarter is still the -- what we're sharing now, which is we're testing different price points.

It is sold a la carte to your second part of your question. And so it's not bundled in. It is a separate price point product. You do need to be a marketplace customer to get it, but it's sold a la carte. You also mentioned paid AI LLM-based marketing.

That's well, AI sourced audience is still very small for us. We are positioned the best there by third-party measurement. So we're there as that channel grows. But most of them don't have paid models yet. And so virtually all of the traffic that we're getting from LLMs today is organic traffic.

Our next question comes from the line of John Colantuoni with Jefferies .

I have 2. First, when thinking about your outlook for revenue for the quarter and the year, can you talk to your expectations for the relative contribution of Dealer Count and QARSD? And second, I believe you mentioned that retention reached a 3-year high.

Can you talk about what's helped drive improved dealer satisfaction? And how does that elevated retention fit into your outlook for customer gains and monetization?

Sure. I'll start, John, and then Sam will talk about the retention piece. We don't break out the revenue outlook between QARSD and rooftops. I think if you look at the last couple of years, you'll see pretty consistent trends. What I will say is what we've talked about quite a bit in the past is the relation -- and you can see this in our investor deck, the relationship between those 2 metrics.

When we grow rooftops more quickly, that has a dampening effect on QARSD even if all of the underlying levers of QARSD are just as strong as they otherwise would have been. And so we don't -- we obviously have a lot of models and benchmarks and internal metrics that look at those 2 metrics.

But from an external perspective, we optimize to and communicate about MRR, monthly recurring revenue and then actual revenue. Sam, do you want to talk. And so yes, and as part of all of that, retention is a key ingredient. One of the best ways to grow rooftops is to retain as much of your installed base as you can. And so Sam will talk about some of those factors.

Thanks, John, for the question. Really proud of that retention. You know that our focus with our sales and account management teams is new business growth from the existing base and then retention. It's all part of what we call net monthly recurring revenue.

It's happening because, number one, the ROI is clear to our customers. When you look at lead growth, both quantity and quality, we're in an advantaged position, and we are producing a tremendous set of results. And you've seen lead growth in both our global markets, U.S. and Canadian and U.K. markets have been phenomenal.

We've added, as Jason said, this account management function. What we've done specifically on that one is added this dealer performance partner group. What they do is come from an industry background. They're retail professionals who join the CarGurus team and then go in and share best practices across our leading national providers all the way down to best practices at a small independent dealer.

That kind of investment has made massive changes to the lead management practices our dealers are following, and they're, in turn, growing their close rates even further than where they were previously. And then finally, you embed in that the new software and data tools we're providing and you suddenly become a profit maximization engine for our dealers.

So you look at that and say, as dealers have said, PriceVantage is advantaged in the marketplace because you are #1 in consumer demand signals for me. And number two, you have more inventory in your sites, you can help me price to maximize my profitability.

That, plus the data we're providing in dealer data insights literally makes you a part of the operating system at dealerships. And I think that's the biggest reason you're seeing us today with the record retention of our customer base.

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

I guess 2 for me. One, when you think about kind of the content provided to the LLMs, your content, I mean, how should we think about kind of the moat there and maintaining the direct relationship with the customer just philosophically?

And then second, the monthly uniques in the U.S. have really inflected kind of in the past 3 or so quarters versus '24 that was maybe a little bit more flattish, plus or minus. Can you share just kind of what's been driving that higher growth here in the monthly uniques in the U.S.

Sure, Brad. It's Jason. I'll take the content to the LLM. So LLMs are a new form of search, obviously. But at the end of the day, they are a form of search, which is not sort of -- behaviorally, it's very different from traditional search. But structurally, it's not very different from traditional search. And so consumers are looking for guidance. They're looking for direction. They're looking for answers.

And it's a question of can the search engine provide a sufficient enough answer for them to take the next action they need to? Or do they need to go to a specialist. And ultimately, we are seeing and we feel and see from our research, what we're seeing in the ultimate data as well that a small but growing percent of consumers are starting at the LLMs, and they are using that as a starting point, and then they are coming to us to gain the confidence and the trust and the validation and the context and the workflow that they need in order to go through the shopping journey. And so we are taking an approach, which is we want to make sure we're positioned well there because it's clearly a growing channel. And so we are sharing our data with LLMs in a controlled way. We're giving them access to certain data, not all data.

We are helping them to create a good experience for the consumer because we have the most data in the industry, and we're able to give them the sort of keys to that content or which content we decide to give to them. But then the experience that they get there is a fraction of what they get on our site.

And so we're very conscious about the totality of the consumer experience, which is not a one and done. Typically, this is a multi-month experience for consumers. And so it's about the data, but it's also about the experience and the context.

And just as we're a company that's now about 20 years old, and we've adapted through many iterations in technologies, and each one is unique and AI is certainly unique. But as it's gone from SEO to SEM and social has risen and video has risen and app has grown in phone and e-mail and text and chat and mobile versus web. I mean, we've adapted through all of that and continue to deliver a really good service. So in terms of traffic trends, Sam, do you want to talk about it, and I can add if helpful.

Brad, thanks for the question. We're really proud of 2 aspects that are driving that kind of traffic growth for our business. The first is the consumer experience and continuing to invest in that, particularly with AI at the center of that.

Number two is broadening our reach to consumers in different elements of their purchase cycle. So moving back to research and decision-making as well as the purchase process, as you've seen with products we've launched and our brand campaign is taking us there. I'm really proud of the Big Deal campaign.

We just announced this week that we launched our newest Big Deal campaign for TV and online and social advertising. You're going to see more from us on that. And that highlights Discover, as Jason talked about, our AI-driven, very differentiated conversational search tool as well as Dealership Mode, which we talked about as well.

So I think the 2 things you're seeing from that are: one, app downloads and monthly active users and traffic and leads coming from that for us as a very important own channel for us, really important that, that comes directly to us. And number two is our direct traffic up 16% because those brand campaigns have reached a broader audience in the marketplace and then taking them down through Discover or Dealership Mode on their app and making them real live purchasing customers for our dealers to see with full attribution. We're really proud of that growth.

Our next question comes from the line of Andrew Boone with Citizens Bank.

I'd like to follow-up on the last question. And what I'm trying to get at is the proprietary information that's available to you guys. As I think about certainly franchise dealers, but I really want you to focus on kind of the independents. Can you talk about the long tail of what may be available to CarGurus that won't be able to be scraped or picked up by an agent in the future that really does make your information proprietary rather than what could be built by an LLM in the future?

And then as I think about your own AI products and AI consumer mode with Discover, can you help us understand whether that's improving conversion? Or how does it actually change the consumer action on the site as it relates to improving financials or leads or whatever you want to highlight?

Sure. And the second part of your question was specific to Discover. Is that right, Andrew?

Yes, sir.

Yes. Okay. So on the first part, -- and I mean, we can talk in terms of indie dealers, but I think it's a pretty similar answer to both. I mean I would first start with we have really deep relationships with over half the market of dealers, and we have relationships with tens of millions and are visited by tens of millions of consumers every month, and we have a relationship with them.

And both sides of that marketplace, if you will, have trust in us and have history with us. And so yes, the data may be the keys to the kingdom, but there's also all of the infrastructure that goes around it in the form of relationships.

In terms of getting more tactical, I guess, we have a -- that has been crafted over many, many years, an understanding of what the consumer wants to know and needs to know in shopping for a car. And we have brought that to an experience that includes things like IMV and a deal rating and a dealer rating and understanding collaborative filtering and what other shoppers have shopped for, and we have personalization that creates a good sort order, the most relevant sort order for a consumer. We have data that we're catering to dealers who are using that data to perform their businesses better. And so from a data and actual data perspective, I would say it's like a system of competitive advantages that it's when you bring it all together that it really starts to sing.

And then we could also choose and a number of marketplaces are choosing this too, we could choose to cut off access to certain parts of our data. And we may do that. I mean this is obviously an extremely dynamic environment right now. And so we're watching things every day as to how to approach it.

But I think of it as there's both sides of relationship, there's trust and there's the totality of the workflow in the context. And then maybe it might be these custom indices that we've done like IMV and Deal rating, but it's the belief in those. And you're increasingly hearing that somebody can build an agent or there may be an agent that goes and performs a task or get some information. But if there isn't trust in that and an ability to verify and validate, then it's not that useful. And for a purchase is considered and big as this, a consumer absolutely needs that trust and confidence.

For Discover, and this is, I think, another worthwhile angle to mention, which is like we're also not sort of sitting on our hands here ignoring AI in our own products while the world develops quickly on AI. So Discover is a good example of that Dealership Mode is a good example of that.

I would say the key things that Discover is doing is, number one, it's getting us access to a higher funnel customer. Historically, if you didn't know what type of car you wanted, we really weren't that useful to you. We would send you to some filters and would -- it would be a clunky search experience.

Now we're able to be extremely useful to people that don't know what type of car they want that may not even know the make model trim ontology at all, and we can help them find that. So that's number one.

Number two is we are helping them get to a better answer for what they're looking for. And that's because it's conversational and it's colloquial and it has memory and is personalized and is not confined to the filters and the ontological structure. And so it allows people to ask for things like larger wheels or shinier hub caps that they -- the filter didn't exist for that they can now get to. So number two is it helps them get to a better answer.

Number three is it engages them more. And so we're seeing multiple interactions. And I think that ultimately leads to them having more confidence when they then connect with a dealer. And so adoption is growing. Traffic grew 3.5x.

Leads were up 10x quarter-over-quarter. And so clearly, we're finding an audience for it and they're finding engagement. And then we're closely watching how those leads convert once they get to the dealer. But our -- like a lot of things in our business, we have found that the more information you give a consumer, the more qualified they become as a lead.

Our next question comes from the line of Rajat Gupta with J.P. Morgan.

I just had a couple of quick ones. You mentioned earlier, Jason, that the industry is consolidating towards fewer providers over time. I'm curious if you can give us any update on that, how that trend has progressed?

Has that accelerated in any meaningful way recently? What are you baking in, in your outlook in terms of that shift in the industry? I'd love to hear your thoughts. And I have a quick follow-up.

And Rajat, consolidating.

I mean the dealers are consolidating towards like fewer marketplace options.

Yes. So I mean, the stat that we've given, which may be a couple of quarters dated now is that it used to be, I think, dealers on average would subscribe to 3-plus marketplaces, and that now is down to sub 2. I think it was 1.8 or 1.7. And so that's like a self-reported dealers are sharing that.

You'll also hear dealers when you talk to them, they will talk about the need to streamline their partners and their vendors. And that's across many areas of their business, not just consumer marketplaces. And so that's sort of another way to triangulate into the theme or the trend anyway.

And then I just -- I mean, you just look at the spend and the wallet share and where the wallet share is going. And we have continued to outpace market growth, which means we're gaining share. I do think there's growth in the market, penetration of dealers of digital marketing among dealers relative to offline or traditional marketing continues to grow. So I think we're in a growing market for sure, but we're certainly -- we're not in a 14%, 15% growth market. And so we're definitely taking share there. And I also think our -- especially over the last year to 2 years, our pace of innovation has accelerated, and that's creating a better and better consumer experience as evidenced by Discover, Dealership Mode, a number of things, and that continues to grow our leadership position.

Understood. And a quick follow-up. We were at [ NADA ] we have these -- we have a lot of like new and upcoming DMS providers, cloud-based and also some of the legacy ones moving from single-tenant to multi-tenant. I'm curious, like is any of this helping with this consolidation, making it easier for you to penetrate just given you have a little more advanced tools and tech internally. I'm curious if any of that is driving this market share shift at all?

I don't think a proliferation of start-ups is driving -- is driving acceleration consolidation. I would say that we are -- have -- among the many things we've accelerated recently, one area is integration with other technology providers. And that's both been a push and a pull. We've been pulled to them by them seeking our data, knowing that we're the largest marketplace.

And we've been pushed to them by dealers who say, I love CarGurus. I use your data all the time. I love XYZ the other provider. I really wish you 2 would work together. And so through partnerships and integration, we are becoming more intertwined with the leading providers or the providers of the future, which is a service to our customers.

Our next questions come from the line of Ron Josey with Citigroup.

This is Jamesmichael Sherman-Lewis on for Ron. Two, if I may. First, great to see the traction with DDI and PriceVantage in the 8-figure revenue level target for monetized dealer products. But at a higher level, how are you balancing monetizing these newer analytics via subscriptions versus passing on greater value to dealers, particularly as we think about newer offerings like the instant market value model coming online soon? And then I have a follow-up.

Yes. It's a -- that is a very -- we could talk for hours about that. And so there's not a simple answer. What I would say is that as our data grows, I guess. But as the uses of our data grows in the form of DDI, which would be more insights and intelligence in the form of products and informing products like PriceVantage, we are having to get certainly more sophisticated and think more broadly about how to price, package, bundle, deliver, take to market, onboard, et cetera, our customers, which is why we've invested in account management, by the way. And so there's not a silver bullet. There's not even really a single formula for when to bundle, when to offer freemium, when to give away, when to charge a la carte.

But what we're trying to do is really, we're just trying to grow engagement because what we've shown is as dealers engage more with our products, they perform better on our platform. And so it's us doing some art and science every day, every month as we're rolling out products to look at what are the behaviors and what do we think is going to drive long-term engagement.

Great. I appreciate the color. My follow-up, I know we've hit on the margin trajectory a few times on this call. But could you add more color on relative investment intensity across product offerings, brand investments, international expansion or other areas? Any way to rank order the magnitude here as where you might lean in more this year?

It's what I said before that number one is in product and tech; number two, international; number three, account management.

Our next question comes from the line of Tom White with D.A. Davidson.

This is Wyatt on for Tom. I'd like to hear more about international ambitions over the course of 2026, given that you're the #2, 3, 4 player in your international markets. What are you doing this year to continue grabbing share from your competitors internationally?

Thanks, Wyatt. It's Sam. I'll take the call. Not 3 or 4, we're #2 and fast gaining on #1. So sorry to give that specificity. These were businesses that if you haven't heard of it before, they were small, slow growing and unprofitable, and now they're sizable, fast-growing and profitable. That's happening because we are driving tremendous lead growth and lead quality.

We're ranked #1 by our dealers in ROI in those markets against all of the other players that have existed. We're actually coming up on our 10th year anniversary in both markets, Canada and the U.K., very proud of what we've accomplished over those years. And I think what you've seen in the -- both the dealer adds in international and the QARSD growth is a combination multiplier effect that says there is a market out there for many of these dealers, if not all of them, to say, and these are the calls I'm fielding day-to-day.

We want to diversify our spend from the partner who we worked with in the early years, who's the largest player in the market, and it's not being effective for us anymore. And we don't like some of the things that have happened in those markets. You may have read about the U.K.'s fourth quarter initiatives that one of our competitors made to try to force products to our dealers and price increases that led them to say, we're going to move over to other players in the marketplace. We'd like to look at the player that gives us more ROI.

So it's opened up many more of those conversations for net dealer adds and the growth in our QARSD. And I think you don't get 27% growth in an international market in the fourth quarter without that kind of momentum. So I -- we feel very strongly about it. Jason just mentioned that -- it's the second biggest area of investment for our business. We're differentiating. So think about the products we're bringing to market here in the U.S. with Discover and Dealership Mode.

Those are differentiators for consumers and dealers that our competitors have not built. So our thought is we will hope to build a product road map that's similar to the one we built here and then bring it to international markets. That's a road map for us. So thank you for asking. We're really excited about the growth and the potential.

Got it. Yes, that's fair. And that's really helpful. And on the topic of international, could you maybe quantify the EBITDA drag related to international expansion that's reflected in the margin outlook for the full year?

Yes. We don't break it out. So as Sam mentioned, there -- we're one segment now. And as Sam mentioned, there high-growth profitable businesses. And the stats you heard in the script on Q4 give a good sense of the growth. So that's the extent of the detail we can give.

Our next question comes from the line of Joseph Spak with UBS.

It's Alejandro on for Joe. Maybe could you help us sort of disaggregate the QARSD growth for the year for pricing versus the addition of new products and how that's compared to sort of year prior?

'25 versus '24?

It actually does.

Pricing is towards the [ bot ] -- unit pricing is one of the weakest levers -- or one of the least strong levers because we don't pull it very hard, frankly. So up-leveling is number one, upgrades rather, adoption of add-on products is number two. And then the bottom half are price increases and lead quantity and lead quality. And then there's a number of things that we continue to add to the products that are free, which support these things like listing upgrades and even price increases, I suppose. And so it's not like we're trying to squeeze water from a stone. We're actually putting more value in the product to drive those.

Understood. And then maybe just as a follow-up, like given the high level of investments that you're expecting, like how should we think about the capital allocation repurchases going forward?

How should you think about the stock repurchases you said?

Correct.

I mean -- well, the high investments going forward, we gave a sense of what our margins would be. And I think if you looked at our margins -- EBITDA margins are still extremely strong compared to benchmarks. And if you look at our investment in things like technology and R&D, we're also right in line. So I guess, high is a relative comment or a trend-based comment, but not necessarily a relative comparative one.

In terms of share repurchases, we announced a $250 million new program. And so I would think about that being the potential for it. And the way we think about it is we think about investment in the business, which goes into our EBITDA calculations and output.

We think about M&A, which we're always evaluating. And then we think about returning capital to shareholders, and that's the share repurchase.

Our next question comes from the line of Marvin Fong with BTIG.

Just a housekeeping question. I just wanted to double-click on the statement you made about new products reaching 8 figures for the 8-figure revenue level, you said. So first part, is that for the total of 2026? Or do you think you can actually hit 8 figures in any given quarter? And then kind of relatedly, just how should we think about the trajectory of revenue growth as I sort of think about these new products? Do you -- is the upper end of your guidance -- revenue guidance imply that you could exit 2026 still at double-digit growth?

It's a full year. It's in reference to full year. the 8-figure comment. And we haven't commented on exit rates. I think if you -- given the Q1 guide and the full year guide and looking at historical, you can get a sense for how you think they might play out quarterly, but we haven't spoken to quarters to give you exit rates.

Go ahead. You're fine.

Sorry. Okay. Yes. So we knew that was the end. So thanks, everyone, for your time this evening. Thanks so much for the questions. We're extremely proud of a really strong 2025. So huge thanks to our team and our customers, and we're extremely excited for 2026 and the momentum behind us. Thanks, everyone. Have a good night.

Thank you. And this concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.

Good day, and welcome to the CarGurus Earnings Call. Please note that this event is being recorded. I would now like to turn the conference over to Kirndeep Singh, Vice President and Head of Investor Relations. Please go ahead.

Thank you, operator. Good afternoon. I'm delighted to welcome you to CarGurus' Third Quarter 2025 Earnings Call. With me on the call today are Jason Trevisan, Chief Executive Officer; and Sam Zales, President and Chief Operating Officer.

During the call, we will be making forward-looking statements, which are based on our current expectations and beliefs. These statements are subject to risks and uncertainties, which could cause our actual results to differ materially from those reflected in such statements. Information concerning those risks and uncertainties is discussed in our SEC filings, which can be found on the SEC's website and in the Investor Relations section of our website. We undertake no obligation to update or revise forward-looking statements, except as required by law. Further, during the course of our call today, we will refer to certain non-GAAP financial measures.

A reconciliation of GAAP to comparable non-GAAP measures is included in our press release issued today as well as in our updated investor presentation, which can be found on the Investor Relations section of our website. We believe that these non-GAAP financial measures and other business metrics provide useful information about our operating results, enhance the overall understanding of past financial performance and future prospects and allow for greater transparency as it relates to metrics used by our management in its financial and operational decision-making.

With that, I'll now turn the call over to Jason.

Thank you, Kirndeep, and thanks to everyone joining us today.

In the third quarter, we delivered double-digit year-over-year marketplace revenue growth while also expanding profitability across our U.S. and international businesses. Marketplace revenue and Marketplace EBITDA both finished above the midpoint of our guidance range, reflecting focused investment to drive sustainable top line growth and disciplined execution of our strategic priorities.

Marketplace revenue grew approximately 14% year-over-year or $28 million, and Marketplace adjusted EBITDA was up 18% during the same period. Growth was driven by continued expansion in QARSD, led by dealer upgrades to higher tiers, broader adoption of our add-on products, like-for-like price increases and higher lead quantity and quality. We also added 1,989 net new dealers globally year-over-year, supported by stronger retention.

Our international operations contributed meaningfully with revenue up 27% year-over-year, driven by momentum in both Canada and the U.K. QARSD grew 15%, and we added 807 net new dealers year-over-year. At the foundation of these results is the strength of our market-leading 2-sided marketplace.

Built on trust and transparency, CarGurus connects the largest audience of car shoppers with the broadest network of dealers, giving consumers confidence and dealers high-quality demand and intelligence, both of which bolster marketplace liquidity through rising engagement and adoption. As our marketplace continues to scale, it generates vast proprietary data and machine learning signals that fuel a uniquely advantaged analytics and intelligence platform for dealers.

With this expanding data set and our accelerating AI capabilities, we turn data into intelligence, delivering predictive tools and insights that help dealers make faster, smarter decisions and achieve stronger outcomes. These dynamics reinforce 2 durable advantages, scale and data intelligence. Scale delivers reach and liquidity. With the broadest dealer network and deepest inventory, our marketplace offers car shoppers unmatched selection and transparency, attracting the largest consumer audience and in turn, more dealers.

That flywheel has supported faster growth and share gains from our primary competitors. Data intelligence transforms that scale into smarter products. We believe our growing size generates the most comprehensive retail demand and pricing signals in the markets where we operate, which we productize into solutions that improve dealer profitability. For example, our retail demand analysis recommends vehicles aligned with local shopper interest.

And when dealers follow those recommendations, we've proven their inventory turns faster. Our pricing models enable dealers to price with precision, improve margins and outperform competitors, while behavioral and intent data enriches leads to improve conversion and ROI. This creates a virtuous cycle in which scale drives richer data and intelligence derived from that data improves dealer performance and the consumer experience, which in turn, we believe drives ever-increasing adoption and engagement.

Building on our position as the #1 most visited automotive marketplace, we've continued to expand our platform with software and data products that help dealers make more intelligent decisions across 4 key workflows: inventory, marketing, conversion and data. We've already introduced a variety of offerings in each of these 4 pillars.

In inventory, products like Sell My Car, Acquisition Insights and next Best Deal Rating help dealers source the right vehicles, merchandise and price each inventory unit with precision. In marketing, solutions such as our core listings packages, highlight, RPM and new car exposure connect dealers with high-quality, ready-to-purchase shoppers efficiently and generate significant dealer awareness and walk-in traffic.

In conversion, offerings like Lead AI, our in-person engagement team and Digital Deal help dealers convert leads into sales, driving better attribution and higher close rates. And in data, our Dealer data insights suite delivers local market intelligence that powers smarter, more profitable decisions.

Over the past few years, we have built a strong foundation and garnered dealer engagement across these pillars and are now advancing from add-in features in these areas to differentiated software and data products, each with a clear value proposition and measurable ROI. We believe these products will expand our addressable market from the current $3.5 billion spent by U.S. dealers on marketplaces by roughly an additional $4 billion U.S. dealers spend on software and data products in these segments.

We believe that our growing product suite positions CarGurus as an intelligence-driven partner that helps dealers optimize every stage of their workflow beyond simply marketplaces. We plan to deepen monetization across these pillars through scalable software and data solutions, and we're excited to share that we've begun that this quarter with our newly launched Price Vantage, which I will cover shortly.

Much like we've done for our dealer partners, we are expanding our offerings along the consumer journey, continuing to lead the market in trust and transparency while broadening our role more upstream with research and downstream to purchase. With the largest selection and a seamless online to off-line experience, shoppers can research with confidence, connect with dealers and complete the transaction on our platform or at the dealership in the way that works best for them.

We believe this expansion of our product suite on top of our market-leading marketplace will continue to reinforce our scale and data intelligence flywheels and result in us capturing more dealer wallet share and deepening consumer engagement to support long-term growth. With that context,

I'll now walk through our progress across our 3 drivers of value creation. Driver number one, expanding our suite of data-driven solutions across dealers' workflows to help them drive more profitable businesses. Core to our mission of helping dealers make more profitable decisions, we recently launched PriceVantage, a major machine learning-based evolution of our pricing tool.

It is the only used vehicle pricing solution powered by real-time consumer demand from the #1 most visited car shopping marketplace, giving dealers an edge to predict the market rather than just react to it, enabling smarter pricing, faster turns and improved profitability. Built on the industry's largest data set of shopper behavior and market supply, PriceVantage leverages AI to deliver VIN level activity, turn time predictions, lead potential, market day supply and visibility into comparable listings, all within a single unified workflow that directly syndicates into dealers' inventory management systems.

It translates live market dynamics into data-driven pricing recommendations aligned with each dealer's goals, giving dealers greater speed, control and confidence in every pricing decision. Early beta results demonstrate the power of the software. The most engaged dealers using PriceVantage saw a 5x improvement in turn time compared to their top 5 competitors on CarGurus. Taking price drop recommendations drove a 68% median increase in daily VDP views and 77% of recommendations met or exceeded predicted sales velocity outcomes.

We launched a Chrome-based browser extension that embeds these insights into the platforms where dealers already operate, such as their IMS or auction sites. Dealers can access real-time price recommendations without leaving their workflow with future releases planned to extend into sourcing and merchandising. PriceVantage is the latest and most substantial addition to CarGurus's expanding suite of dealer intelligence software solutions.

Other offerings continue to grow, especially our Dealer Data Insights suite, which strengthens dealers' predictive capabilities, delivering greater efficiency and faster sales. Next Best Deal Rating is now used by nearly 20,000 dealers, growing over 70% year-over-year. Merchandising Insights adoption grew to 9,791 dealers, while Max Margin Insights adoption rose to 5,032 dealers.

In the third quarter alone, dealers made over 700,000 price changes through Next Best Deal Rating. We've seen a median 48% increase in VDP views and faster turn times for vehicles using our recommendations. Engagement remains high, with next best deal rating driving nearly 50 price changes per dealer in Q3 and dealer data insights reports overall driving 75 price and inventory changes per dealer. Over 2/3 of recommendations we send to dealers are being opened and red, indicating the value of these insights.

Last quarter, we also introduced new car exposure to give dealers more sophisticated control of their new vehicle marketing. New car exposure continues its rollout across markets, now reaching 94 DMAs and brand combinations. To date, it has driven 31% of new car VBP views and 13% of new car leads, with participating dealers capturing a greater share of new car leads than those relying solely on organic placements.

Innovations like this are deepening dealer engagement by enabling smarter decisions across inventory, marketing, conversion and data. Dealers are upgrading into premium tiers more frequently. They're adopting our products and solutions at higher rates, and they're signing long-term contracts. Together, we believe these factors support our ability to grow QARSD. QARSD growth has been manifesting in 3 trends: first, customers who remain on our platform consistently increase their spend over time; second, new customers are joining at higher average order sizes than in prior years; third, newer customers are ramping their spend faster than prior new customers did.

On all these observable dimensions, we're seeing clear evidence that the growing quality and breadth of our products have been driving measurable QARSD growth. Driver number two, meeting the evolving needs of car shoppers by powering a more intelligent and seamless journey. As I said earlier, we're expanding the CarGurus experience across the full car buying journey from research through consideration and purchase. This quarter, we advanced 2 key innovations that bring that vision to life. First, consideration.

We expanded CG Discover, our Gen AI-powered shopping assistant. Unlike others that use GenAI to repackage traditional filters, discover uses conversational understanding and real-time reasoning to interpret a shopper's intent and curate the best fit vehicles for millions of listings. It helps consumers refine choices and explore inventory with greater speed and clarity while giving CarGurus richer demand signals and pricing insights to strengthen the data intelligence flywheel.

Early engagement has been strong, and we have since expanded Discover to our homepage and app, creating more prominent entry points that have driven higher traffic into the experience. Research shows 80% of consumers are open to using AI in their car buying journey, underscoring the scale of this opportunity. Traffic to CG Discover has nearly tripled quarter-over-quarter, and leads have grown 3.3x.

Discover VDP to lead conversion is 6,000 basis points higher than standard VDP to lead conversion. As Discover scales, every interaction generates signals and insights, making the platform smarter and strengthening both dealer and consumer experiences. Next, purchase. Car shoppers want confidence at every step from discovery to purchase. Research shows consumers feel the hardest part of car buying happens in the dealership when shoppers feel anxious about pricing, alternatives and making a rush decision.

Our goal is to reduce that anxiety with transparent dealership ratings and reviews and by extending the CarGurus experience into the dealership where support matters most. We're excited to introduce dealership mode, a major innovation in the purchase step designed to deliver real-time support at the exact moment shoppers need it. When a CarGurus user visits a participating dealer lot, the app activates through geofencing and push notifications to provide VIN level pricing and ratings, reduce payment anxiety with a financing calculator, compare cars on the lot or highlight alternatives at the dealership and surface reviews to validate quality.

Dealership mode gives consumers clarity and confidence at the most stressful point in the process. For dealers, dealership mode strengthens attribution and ROI. While we already maintain significant attribution on closed sales data through DMS integrations and third-party data providers, dealership mode now enables us to close the purchase loop more fully, connecting online engagement to in-store activity, which we believe demonstrates clear ROI and higher quality leads.

With millions of monthly app users making hundreds of thousands of lot visits, we believe the opportunity is significant. Based on an early analysis, 56% of consumers who see dealership mode in the app navigation have clicked into the experience and over half of our users have opted in for push notifications. Over time, we expect dealership mode will drive even greater app adoption, build consumer trust and help dealers convert more sales. By improving the consumer experience and extending our brand awareness, we are giving shoppers more reasons to start and end with CarGurus.

This deeper engagement is translating into higher intent activity with CarGurus-led sales growing year-over-year in the past 2 years. Separately, as we implement changes to comply with cookie consent regulations across markets, reported uniques and sessions are expected to decline as some users do not opt into tracking. This represents a change in how traffic is measured rather than an indication of an underlying change in site traffic or in the leads and connections we believe we're delivering to our dealer partners.

Driver number three, enabling dealers and consumers to complete more of the transaction online, streamlining the final steps of the deal. In the third quarter, we advanced our transaction capabilities through continued progress across Digital Deal and Sell My Car. These offerings are delivering a seamless online to off-line journey for shoppers. Digital Deal adoption surpassed 12,500 dealers this quarter with over 1 million listings digitally enabled.

With more digital deal listings and improved user experience, we have driven 45% year-over-year growth in high-value actions such as financing applications, appointment scheduling and deposits. Users who complete these high-value actions closed at up to a 3x higher rate than standard e-mail leads. In fact, our strongest close rate comes from reservations. Reservations closed at nearly 16x the rate of standard leads for out-of-market shoppers and 9x for in-market shoppers.

Appointments are up approximately 20% year-over-year. Financing adoption is also strengthening, supported by direct credit applications, prequalification and SRP filters that surface vehicles consumers are already approved to finance. Digital deal leads with a financing element have grown 77% year-over-year. We also embedded high-value actions into the core site experience.

This quarter, we introduced a post-lead high-value action menu that surfaces additional steps such as scheduling an appointment or submitting a deposit immediately after a shopper submits a core lead. This creates a natural ramp for consumers and provides dealers with even stronger intent signals. Alongside a broader redesign of the digital deal experience, initial testing shows several hundred thousand incremental leads from the new experience.

We now expect that by year-end, nearly 30% of the digital deal-enabled dealers' e-mail leads will come through Digital Deal. These leads include verified contact information, full name, e-mail and phone number and around 45% of them historically carry at least one high-value action. Beyond enabling more of the transaction online, we're helping dealers source inventory with greater efficiency.

Sell My Car adoption has continued to grow and is now live in 115 markets, reaching roughly 75% of our eligible traffic. Lead quality and conversion have continued to strengthen. A growing share of Sell My Car acquired vehicles are listed on our marketplace soon after purchase, demonstrating that these are high-quality retail-ready leads. Collectively, these advancements are streamlining the transaction for both dealers and consumers, improving lead quality, accelerating conversions and reinforcing our ability to meet customers wherever they are in their journey.

Across all of our value creation levers, I'd like to discuss the accelerating use of agentic AI. AI has been foundational to CarGurus since our inception and continues to power innovation across the platform. We're embedding Agentic AI in numerous places throughout our products and systems to create smarter, faster and more intuitive experiences for both consumers and dealers.

CarGurus Discover, our conversational Gen AI-powered shopping assistant, uses large language models to help consumers refine choices and explore inventory with greater speed and clarity. In our mobile app, dealership mode activates when a shopper visits the participating dealership lot, providing AI-generated comparisons and summaries of vehicles.

In our dealer dashboard, PriceVantage extends these capabilities to dealers by using predictive AI and real-time demand data to deliver bin-level pricing insights, turn time forecasts and competitive benchmarking directly into their workflows. We also continue to scale AI-driven content and quality improvements across the platform to drive consumer traffic and reduce operational overhead across internal teams.

SEO content generation powered by generative AI and guided by our editorial expertise has expanded high-quality content roughly tenfold across CarGurus in our core channels, driving a 60% increase in top and mid-funnel sessions year-to-date. Pricing and compliance monitoring now also uses AI to identify inconsistencies and ensure data integrity across millions of listings.

Internally, AI is transforming how teams work. Over the past year, we've deployed numerous solutions that have improved speed, precision and efficiency across nearly every function. Our Gen AI sales tools have provided account summaries, tailored recommendations and predictive insights that have helped teams identify opportunities to strengthen retention and deepen dealer relationships.

Nearly 80% of managed leads in October, chat and text were handled and closed by AI. This automation has enabled us to reduce the outsourced team by over 40%, driving meaningful efficiency gains and cost reduction. Engineering productivity has risen by nearly 25% in the past year through the use of AI coding tools and code review agents.

Our LLM gateway democratizes LLM integration, allowing teams to embed new use cases directly into workflows and bring ideas to market faster, while our enterprise LLM-based search platform enhances knowledge retrieval and workflow automation. AI also strengthens fraud detection and prevention, enhancing data integrity and platform trust. Adoption is broad and disciplined. 91% of employees report using AI weekly, driving faster execution, sharper insights and greater collaboration across the company.

Looking ahead, we believe that the combination of proprietary data, machine learning, predictive analytics and agentic AI positions CarGurus to deliver new levels of intelligence, automation and efficiency to both dealers and consumers. AI remains central to how we innovate, operate and lead in automotive technology. In Q3, we delivered strong revenue growth, healthy margins and disciplined execution.

We advanced products that give dealers greater control, efficiency and intelligence while creating more confidence and clarity for consumers. These innovations are expanding our reach beyond the $3.5 billion U.S. marketplace segment into an additional $4 billion dealer software and data products TAM, which we believe broadens our long-term growth opportunity. Innovation remains at the center of this progress.

We're extending our platform across each of our 4 pillars: inventory, marketing, conversion and data with scalable software and intelligence solutions that address more of the dealer workflow and consumer journey. These advancements reinforce our leadership as a data and technology-driven company, which we believe unlocks new sources of growth and value creation. Across every initiative, our focus remains on measurable value, capturing more dealer wallet share, deepening consumer engagement and strengthening our platform's foundation.

With that momentum, we believe that we're scaling solutions that reinforce our leadership, support durable growth and create long-term value for our customers and stockholders. Now let me walk through our third quarter financial results, followed by our guidance for the fourth quarter and full year 2025. Third quarter consolidated revenue was $239 million, up 3% year-over-year. Marketplace revenue was $232 million for the third quarter, up 14% year-over-year toward the high end of our guidance range.

Marketplace revenue growth was driven by strength in our subscription-based listings revenue. In the third quarter, U.S. QARSD grew 8% year-over-year, and we added 1,182 paying U.S. dealers year-over-year, marking our seventh consecutive quarter with positive net dealer adds and our fourth straight quarter of accelerating year-over-year dealer count growth.

We continue to expand our footprint while taking greater wallet share in our growing base, driven by upgrades, broader adoption of add-on products, like-for-like price increases and higher lead quantity and quality. Our international business had yet another strong quarter with revenue up 27% year-over-year and international QARSD up 15% year-over-year, the ninth consecutive quarter of double-digit year-over-year international QARSD growth.

Wholesale revenue was approximately $2 million for the third quarter and product revenue was roughly $5 million for the third quarter as we ceased facilitating transactions in the quarter as a result of our decision in August to wind down the CarOffer transactions business. As a reminder, we expect to account for the wind down of CarOffer as a discontinued operation in the fourth quarter.

As such, we do not expect there to be revenue associated with digital wholesale going forward. I'll now discuss our profitability and expenses on a non-GAAP basis. Third quarter non-GAAP gross profit was $214 million, up 11% year-over-year. Non-GAAP gross margin was 90%, up about 650 basis points year-over-year. Marketplace non-GAAP gross profit was up 13% year-over-year and non-GAAP gross margin was stable at 93%. On a consolidated basis, adjusted EBITDA was approximately $79 million, up 21% year-over-year.

Adjusted EBITDA margin was 33%, up about 490 basis points year-over-year. Marketplace adjusted EBITDA grew 18% year-over-year to approximately $82 million, above the midpoint of our guidance range. As a reminder, we guided to Marketplace EBITDA only this quarter as we sunset the CarOffer transactions business. Margin rose about 120 basis points year-over-year to 36%, but declined slightly quarter-over-quarter due to investments in new product innovation and sequentially higher sales and marketing expense.

Digital Wholesale adjusted EBITDA loss of approximately $4 million was modestly higher quarter-over-quarter as expected. The sequentially larger loss was driven by lower volumes due to the cessation of transactions in the third quarter as a result of our decision to wind down the CarOffer transactions business. Moving to OpEx.

Our third quarter consolidated non-GAAP operating expenses totaled $142 million, up 7% year-over-year and 4% quarter-over-quarter, reflecting sequentially higher sales and marketing expense and investment in new product innovation, as I mentioned earlier.

During the third quarter, we incurred $3.8 million in onetime cash restructuring charges, and we expect remaining cash restructuring charges of $2 million in the fourth quarter. Accordingly, we have narrowed our previously estimated range from $5 million to $7 million to $5 million to $6 million.

We still expect to substantially complete the CarOffer wind down by year-end, with total wind-down related charges expected to be in the range of $13 million to $15 million, which is lower than the original range of $14 million to $19 -- non-GAAP diluted earnings per share attributable to common stockholders was $0.57 for the third quarter, up $0.13 or 30% year-over-year, reflecting primarily the increase in adjusted EBITDA and lower diluted share count.

We continue to generate strong free cash flow, and we ended the quarter with $179 million in cash and cash equivalents, a decrease of $52 million from the end of the second quarter, primarily driven by $111 million in share repurchases in the quarter, partly offset by higher adjusted EBITDA.

As of September 30, we have approximately $55 million remaining on our share repurchase authorization. I will now close my prepared remarks with our guidance and outlook for the fourth quarter and full year 2025. As a reminder, due to the wind down of CarOffer, last quarter, we stopped guiding to consolidated revenue and consolidated adjusted EBITDA and instead, we'll guide to Marketplace revenue and Marketplace adjusted EBITDA as that is representative of our go-forward operations.

We expect our fourth quarter marketplace revenue to be in the range of $236 million to $241 million, up between 12% and 15% year-over-year, respectively. And we expect full year marketplace revenue to be in the range of $902 million to $907 million, up between 13% and 14% year-over-year, respectively. For the fourth quarter, we expect our non-GAAP Marketplace adjusted EBITDA to be in the range of $83 million to $91 million, up between 5% and 15% year-over-year, respectively, and we expect full year Marketplace adjusted EBITDA to be in the range of $313 million to $321 million, up between 18% and 21% year-over-year, respectively.

We expect to meet the discontinued operations criteria in the fourth quarter. As a result, we expect our full year guidance, similar to the third and fourth quarters to reflect Marketplace absorbing approximately $1 million in ongoing quarterly CarOffer expenses as a result of the wind down. Accordingly, we've included about $2 million of first half costs that we expect to be recast to continuing operations once the criteria are met.

These estimates are preliminary and subject to change. The midpoint of our Q4 guidance implies a full year marketplace EBITDA margin of approximately 35%. We're pleased with the strength and growth of our marketplace and excited by the early results of our various new product investments. That innovation has delivered growing adoption across more dealer pillars and deeper consumer engagement across their shopping journey.

That success reinforces our confidence to continue growing our investments in new, primarily AI-centric innovation across our dealer and consumer product suites that we believe will drive long-term growth. We expect fourth quarter non-GAAP consolidated earnings per share to be in the range of $0.61 to $0.67, up between 13% and 24% year-over-year, respectively, and full year consolidated earnings per share to be in the range of $2.19 to $2.25, up between 29% and 32% year-over-year, respectively. And we expect fourth quarter and full year diluted weighted average common shares outstanding to be approximately 97 million and 101 million, respectively.

With that, let's open the call for Q&A.

Our first question comes from the line of Chris Pierce with Needham & Company.

If I'm looking at the deck on Slide 5, I think you have a stat that you shared for the first time that may or may not be right, but it says 25% of CarGurus dealers only pay for CarGurus. Is there a way to think about where that stat was a year ago, 2 years ago and some sort of upper bound as maybe you guys have separation versus peers?

Chris, it's Jason. Thanks for the question. I don't think we've given a trend on that stat. But what we have seen is that in surveying dealers that dealers use fewer and fewer marketplaces -- marketplace partners. In fact, over the last few years, I don't remember the exact years, but it's gone from about an average of using 3 to using under 2, around 1.8.

So there's consolidation and concentration with those that typically offer the best ROI. So that's the sort of macro trend on that dynamic, but we haven't given a trend number on the 25%.

Okay. And then on the ROI that you're talking about, specifically on digital deals, are you seeing dealers more willing to engage here given there seems to be an acceptance that fully digital transactions are growing within the industry?

And like I guess what will be the right time to flex pricing power here given the conversion metrics you cited and sort of the dealers need to do something specifically on their end to accept these leads or is it sort of just kind of easy housekeeping on their end and a customer can walk in, have their loan in place, take their test drive and leave the dealership within, call it, an hour, something like that?

Chris, it's Sam here. Thanks so much. We have constantly spoken about the research we've done showing 80% of consumers want to do more online, but still want to touch and feel the car and come in for a test drive. So we think we've got a perfect product in that regard. You've seen that we've got 12,500 customers now on the program. It is packaged into one of our premium tiers.

So the dealers who get access to digital deal have to pay more. I see your point that as that trend continues to move, that's an opportunity for us. But I think the thing we're most excited about is more and more of our consumers doing highly -- what we call high-value actions. So taking a process to either put -- set an appointment to put a deposit down to look at financing and try to provide some information on their credit ability that we think is driving a higher quality lead, a further down funnel lead, and we believe that's driving further and further ROI for our dealers.

So long term, it gives us the opportunity, as you said, to say, how much more will that continue to grow? That gives you an opportunity for pricing power, and we'll consider that as we go forward.

Our next question comes from Marvin Fong with BTIG.

A question just on the international CARSD and international in general is doing so well, very good growth across the board. I just wanted your thoughts on how much faster and higher you think CARSD grow? Obviously, we can look at in the U.K., the dominant player there and generating revenue per dealer is much higher than [Technical Difficulty].

Sorry to interrupt you. Mavin, we are unable to hear you clearly. Could you please use your handset?

Sorry about that. Yes, I just wanted to ask about CARSD, particularly in the international segment. I believe the incumbents in the U.K. in Canada charge a lot more than you are right now, and we're seeing very nice growth in international. So just wanted your thoughts there and how quickly you can pull that lever and close that gap?

Thanks, Marvin. It's Sam Zales. We're really, really proud of the international markets and what we're doing there. You'll recall that we're competing against 2 big players who had monopoly power in those markets. But I think what we're showing is 2 things.

When you drive lead quality and lead quantity in an aggressive amount, it makes dealers stand up to say and you price at a lower price point, it makes dealers stand up and say the ROI is better, and we've shown you the research in the markets to show that our ROI is advantaged versus our competitors. I think though, we're still in a market zone of adoption right now.

We're keeping our prices at a lower level because we are winning more and more customers. And you saw we added 800-plus customers in the international market. So our goal there is to say, let's be smart about pricing. Let's price to the value that we're offering to our dealers. And we know we can always grow that over time, but we're looking to build more market share. So you may have read in Canada that one of the largest dealers in a press release that was out Auto Canada converted by saying, I'm no longer going to be on the Auto Trader program, and I want CarGurus as my preferred partner in that regard.

Those are the kinds of things that will give us that opportunity to continue growing not only dealers, and that leads to other dealers picking up their heads and saying, I might do the same thing. It allows us to keep growing our customer base, but also growing QARSD. The 15% growth, we're really proud of. We'll continue to push in that direction, but we don't want to get too aggressive on that front at a time we're still signing more dealers in both Canada and the U.K.

And that will replicate if we can, the market experience we have here in the U.S. We started with lower pricing. We got the largest base of dealers to our franchise and joined us, and then we raised prices over time, and we think that's a good model to try to take on in that arena. So thank you for recognizing 27% growth in international. We really proud of it and excited to try to push forward.

And if I may, I'd just add to that and echo something that was said in the call. So international QARSD is about 1/3 of the U.S. and the levers that drive QARSD in the U.S., upsell, cross-sell, lead growth, lead quality, pricing, those are all available to us in international, and they're all much earlier and less mature. And so they all have more runway in international.

But the other thing I'd point to from the script is to just call out some of the trends we're seeing in U.S. QARSD, which I think we have every reason to believe will exist in international. And that's among our -- it was 3 themes from the script. Among our paying dealers, they increase their spend the longer they stay with us. The second trend is new dealers are joining at higher AOS than old dealers.

And the third is that despite joining at a higher AOS, they're actually ramping their spend faster than prior cohorts ramp their own spend. And so we're seeing that in the U.S., which is a much more mature market, and we're incredibly proud of that. And international has all of those available and earlier stage.

Those are terrific insights. And maybe a follow-up question, just maybe Sam, this is up here But Jason referenced that you're really attacking, I believe you said $4 billion solutions market. Is that how you're presenting it to dealers? I know that dealers like to think about things in a cost per sale.

But are you actually kind of talking to them about these new analytics in the sense now you don't have to pay for vendor A or vendor Z. Is that how dealers are thinking about it? And is it kind of clear to them that you're presenting both a listing service as well as a solution -- software solution?

Marvin, I'll jump in and then let Jason add color. I think what we're doing every day is talking to customers about driving profit maximization. And that can come from our marketplace business as we spoke about in the call, you start to build solutions like DDI that we've talked about previously, which helps dealers convert more of the leads that they're getting today, helps them increase their profitability.

And then you -- from product-led growth, you're seeing customers adopt those products. So our pricing tool led us to build this software product called PriceVantage. So what we're doing for dealers with that product is simply helping them grow their profitability by reducing their turn times and allowing them to price as most effectively to manage their inventory.

So it all comes out of the marketplace business that then leads to other products, as we said, inventory, marketing, conversion and data. They all work together with the value proposition that says, we're going to help you, Mr. or Mrs. Dealer to grow your profitability by utilizing our marketing tools, our data and now software that lets you run your business more efficiently, that leads to QARSD growth that leads to retention of our customers long term. Jason, anything you'd add?

Just that it is -- they are all connected. The dealer historically has thought of them as steps in the workflow and as such, has allocated different wallets to those different steps. And these products are allowing us to start to tap into those new wallets. But what makes them particularly compelling is we're not selling a stand-alone product here and a stand-alone product there.

When Sam talks about them being tied together for something like PriceVantage, it's saying, if you do this to a price, this is exactly -- or this is what will happen from a leads perspective, from a turn time perspective. So it's giving them recommendations and the ability to act on those with a strong forecast of the results because the results are what occur on our marketplace.

Our next question comes from John Colantuoni with Jefferies.

This is Vincent on for John. Just one with a few parts for me. So it looks like some of the investments you've talked about in recent quarters is really paying off, given both U.S. and international dealer rooftops saw accelerated growth during the quarter. At the same time, QARSD growth slowed a little bit across both geographic segments despite the traction you called out for the product suite.

Maybe talk a bit about what the growth algorithm between rooftops and QARSD ought to look like going forward, touching a bit on the drivers of slightly slowed QARSD growth as well as the relative contributions of improved dealer retention versus net new adds to rooftop growth?

Sure. Vincent, it's Jason. So the relationship between -- so thank you for acknowledging the investments paying off. We are incredibly excited about a bunch of the things that we shared with you all tonight in terms of new launches. The relationship between rooftops and QARSD is math in so much as QARSD is revenue divided by the average active rooftops.

And so what happens is when we grow rooftops much faster, that's a natural math-based headwind or depressant to QARSD. And so this past quarter, QARSID was up about 8% year-over-year. rooftops were up about 5% year-over-year. And if you add those 2 together, you actually get something close to our total marketplace revenue growth for year-over-year, around 13%.

And so if you look at the last several quarters, you'll see that type of relationship. It's not perfect, but I think it illustrates the math pretty well and how the math works. In terms of retention versus adding, we've talked about our retention has been improving nicely over the last set of quarters, even a couple of years. And that's a function of a number of things. We've invested in account management, as you've heard, but I would say a lot of it is through the investment in product.

And a lot of that product is in dealer data insights and things that we're adding to our core marketplace and thus far haven't really been charging for a good portion of them. And so between better account management, between more feature functionality, between more insights that help them perform better on our marketplace, our in-dealer partnership team that helps them perform better. So much of what we build here is to just help them perform better on CarGurus. And when they do that, they tend to stay.

So -- and some of the cohort information I just gave shows that they're, in fact, ramping even faster. And then as you heard about some of the adoption numbers from the script, we're getting really broad adoption of a lot of these things.

Our next question comes from Ron Josey with Citigroup.

This is Jamesmichael Sherman-Lewis on for Ron. First here, on 3G Discover now more deeply embedded, can you help us understand how this new car buying journey and purchase funnel differs from traditional car buying? Clearly, we're seeing traffic and conversion ramp, but curious how you see user engagement and this channels contribution evolve longer term?

Happy to. This is Jason, again. So Discover is definitely a new experience and one that is getting great traction as we talked about, sort of explosive growth, granted it's early, but explosive growth. So I mean, the key thing to recognize is that it's outside of the structure of the SRP or search results page.

And so think of it more as a conversation versus a filter-driven onetime query. And so you -- and I'm sure I encourage you to use it and try it if you haven't. But you can ask questions naturally. You'll get explanations and follow-ups, and you can then continue those follow-ups and ask questions that build on prior questions. And so the discovery goes beyond listings. It actually reasons with the shopper.

It explains why cars are ranked the way they are. It offers side-by-side comparisons. It offers contextual intelligence, market value ownership costs, confidence scores, YouTube videos, side-by-side comparisons of different makes and models that we offer. And so it also offers things that would be outside of a search. So whereas a typical search might offer just sedans, this may offer based on your inputs, some small or midsized SUVs that would solve some of the things you're looking to solve that aren't a sedan.

And so it's actually making recommendations outside of what you're specifically prompting. It creates a ton of opportunities for us on our platform. It also offers opportunities though for dealers because they're going to learn a lot more about the consumer and what they're looking for through the information that we share on the dialogue. And so it really is a 2-way conversation.

What it's leading to is shoppers who use it do almost 3 follow-up prompts. And so it is a conversation. It's converting from a vehicle detail page to a lead at much, much higher rates. And then those leads are much richer to the dealer because we're passing along a lot of that information.

So it really does -- it helps the consumer, it helps the dealer and it helps us. and it's built for agentic expansion. And so the architecture of it allows very easily things like personalized deal alerts, watch lists, comparison across trims and markets as new cars come out. And so it's beginning to and will easily act on behalf of the consumer for future opportunities created by what the consumer has given to the agent.

So we're really excited about it. It is not, by any means, a glorified filter, which some other folks are doing. And so we think that it's going to be a really big opportunity for us that can scale nicely.

That's helpful color. Follow-up, if I may. As we look out to 2026 can you unpack the key investment areas across product, international, brand or other areas? And any changes to relative investment intensity versus 2025 investment year?

Thanks. I wouldn't say there's change to relative intensity. I think what we're really excited about, we had said a couple of quarters ago that we were going to increase investment. And I think this quarter, in particular, is showing a lot of the benefits of that.

We have shown a really quick speed to market with a lot of our introductions. We're showing much deeper engagement, price advantage, new car exposure, dealership mode, Discover. And so we're going to continue to invest in as you said, product, go-to-market, international and focus on getting adoption of those across the 4 dealer pillars and across deeper consumer engagement.

And so I would say, if anything, this sort of reinforces our confidence to continue growing our investments in mostly AI-centric innovation across both dealer and consumer, but we're going to be smart about it. I mean I think we've proven the ability to be really disciplined and to prove execution has to follow innovation and that we're -- we pride ourselves on being a company that balances long-term sustainable growth, high-quality revenue with margin.

Our next question comes from Ryan Powell with B. Riley Securities.

It's Ryan on for Naved. I wanted to ask on dealership mode. Obviously, you mentioned some good metrics on initial adoption. What kind of consumer insights are you able to generate from users engaging with dealership mode? Does this have anything to do with improving recommendations for users? And then I have a follow-up.

Sure. So, number one, to maybe state the obvious, it's giving us a lot of information about who actually goes to the dealership, which may seem like a basic thing. But prior to this, that was oftentimes something that we had to triangulate into. And so this gives us a lot more fidelity on that.

Number two is it helps us and it helps the dealer, frankly, probably more than us, understand what other cars a consumer is interested in to compare, and it helps the dealer cross-sell. I mean a good percentage, I think a lot of times, a surprisingly high percentage of consumers who buy a car through our platform at a dealer end up buying a car that is different from the one that they submitted a lead on.

And so this helps the dealer in that regard. It helps us and again, the dealer understand financing needs, having a calculator there looking at financing options is really valuable because the dealer wants to get them in the right loan. And it just allows them to -- we have -- that's primarily built tool and the consumer can engage with that. And all of the things that I just talked about with Discover are happening in dealership mode.

And so again, it's -- we call it lead enrichment here, but it keeps enriching and enhancing our leads. And so we just capture more and more data on the consumer. So consumers often cite the in-dealership experience as a time when they're trying to comprehend a lot of data, understand a lot of different things thrown at them, and this helps them do that, but it also helps the dealers, and you need to be a paying dealer to be part of this. It helps the dealers understand their customers much, much better.

Great. That's very helpful. And then secondly, on CG Discover. So it was also live in the second quarter. What do you think led to the pretty significant increase in adoption amongst users?

I mean the biggest thing is we made it more available. It was in testing mode, some -- an earlier form of testing mode as we gain more confidence and saw the increased engagement of consumers saw all the stats that we shared on improved conversion rate, all the rich data that we were getting, we made it more available and realized pretty quickly that it was helping both consumers and dealers.

And so I would say that's the primary one. It's definitely improved. We continue to work on it. It's gotten better. But I would say the biggest thing is just exposure to our audience. Like this quarter, we released it in the app. It had not been in the app before. And app is our fastest-growing channel. And so putting it on that really accelerated things.

Our next question comes from Rajat Gupta with JPMorgan.

I wanted to ask a little bit zooming out on the industry backdrop. Clearly, there have been some signs of stress on profitability at some large used car dealers, some stress at like smaller independents as well. And we're also seeing some of the -- at least the publicly listed franchise dealers seeing some profit pressure in the near term.

But cyclically, it looks like inventory is going up, which should be supportive for your business. I'm just curious what are you hearing from customers in terms of budgets? In response to an earlier question from Chris, you mentioned maybe consolidating their vendors. I'm just curious like what's the latest that you're hearing on their planning as we head into '26? And I have a quick follow-up.

Yes, I can start Rajat and Sam may add to it. So we oftentimes will try to distill down macro factors into just a few key points. And we've also said that our business as a subscription business and dealers need to sell cars in good times and bad is pretty resilient to a lot of the cyclical trends that exist, even seasonal trends.

And furthermore, as the largest marketplace with dealers consolidating spend, we're, I think, even more immune and sound. And then lastly, I would say used cars tend to fluctuate far, far less than new cars. And so we're in a bit of a sheltered harbor in that respect, too. So we do, though, try to acknowledge macro factors.

So number one, retail sales for used cars are up mildly. Number two, days on market -- and days on market are down a little bit, but frankly, call them flat, and they're actually rising sequentially right now, but they're pretty steady year-over-year, rising a little bit right now. And pricing is up a little bit, not very much. Inventory, as you just said, is up significantly. It's up double digits.

Year-over-year, it's up 10%. I think though, the biggest thing in all of that is the consumer sentiment is down. And interest rates, I mean, granted, they dropped recently, but they still remain pretty high on a relative basis. And so consumer sentiment down, interest rates still elevated, pricing not having come down and inventory up. And you've got dealers that need to move cars and need to sell cars.

And oftentimes, it's better for them to market smarter than it is for them to drop prices. And we are the largest scale and typically surveyed or frequently surveyed as the best ROI. So they may have some profit pressure. Their advertising spend has steadily climbed year-over-year based on the publics anyway. And we tend to be gaining -- we are gaining share every quarter. So we don't face a lot of pressure despite what dealers may be facing as margin pressure.

Rajat, sorry, I'll just add that the other immunity to short-term pressures that Jason mentioned is the breadth of our dealer base. We appeal to the small independent to the multi-sized independent and franchise dealer and those national accounts you speak to. Our consumer base will buy from all of those types of dealers. And so our breadth, and we're not tied to one particular segment.

We have the largest base of dealers that continues to grow. And I'd just add again, the new car exposure product that we launched just in the last quarter was a relevance to dealers saying, hey, there's a high price point for new cars. Can you help us be more profitable selling those new cars? So giving them an opportunity to win their make in a local market, convert consumers who are coming in saying, I might want a used car, I might want a new car. Oh, my payments might be similar on both. I'm going to buy that new car.

We're helping them create the profitability in a market trend that we saw coming very quickly and built a product to get there and make them more profitable. So I think it's that breadth of dealer base that also adds to the immunity of short-term impacts and our constant growth through those cycles in the macro environment.

Understood. That makes a lot of sense. And just one quick follow-up. I hear a reiteration of the double-digit revenue growth exit rate unless it was meant to be just a fourth quarter number when you mentioned that last call. Could you just give us an update on that? And then how should we think about as we head into '26, the trade-off between growth and margins like you had in the last 2, 3 quarters?

Rajat, can you -- I got the second part of the question, relationship growth and margins in '26. Can you repeat the first part of the question about Q4?

It was not a Q4 question. I think you mentioned on the last 2 earnings calls that you expect to exit the year at double-digit revenue growth for marketplace. I wasn't sure if that was an implied fourth quarter number or you meant exiting like the year into '26 with double-digit revenue. I did not hear an update on that today. So I'm just curious if that is still on track?

Yes. I think that my hunch is that the comment made was in reference to Q4 being a Q4 year-over-year revenue growth rate. And -- but then if you look at what that implies for a full year, you would -- the math would illustrate that, I think, is also a year-over-year -- or a full year year-over-year double-digit growth rate.

So I think the Q4 guide sort of answers both of those questions. And we obviously haven't commented on '26. And so from a growth and margin standpoint, I would probably cite back to comments we've made in the past couple of quarters and this quarter around our enthusiasm around the investments, the growth they're driving, the CASA trends we talked about and the speed with which we're introducing new products.

We move to our next question from Andrew Boone with Citizens.

This is [ Briana ] on for Andrew Boone. You mentioned that 80% of managed leads in October chat and text were handled by AI and that 91% of employees are using AI internally, which has reduced reliance on outsourced teams. Where do you see still the biggest friction points either internally or across dealer workflows where AI can further improve efficiency within the business? And how should we see that coming through on the margin?

2

And is your question related to friction in our business or at dealers' businesses?

Within dealer business.

Within dealer businesses. Well, I think one of the biggest areas of opportunity in the dealers business, 2 dimensions. One is how all of their different steps of their workflow tie together. So -- and you've heard us talk about that, and that's what we're focused on is how can they source smarter, price smarter based on retail signals that they're seeing.

Dealers have, for a long time, been making purchase and appraisal decisions on wholesale data, and that's just not as useful to them. They're more interested in retail data, what can they sell the car for, how much demand does that car have today. Same with conversion. And so how all of the steps of their workflow tie together is one area. The second area is around predictability. It's -- I can use the same example, which is not only were they using wholesale data, they were using wholesale data for appraisal that was 30 days or 60 days old.

And so using AI, a lot of our insights and our price advantage tool and other things that we're providing them now are about predicting what the environment will be, what the implications will be 30 days from now once they have the car and once they price the car and merchandise it, et cetera. So those would be the 2 dimensions. Internally, it's about speed of development and execution and quality of product.

And so I think that shows up in a lot of different ways in product and engineering, but also in other parts of our company. It shows up in how well we serve our customers with our sales team and account management teams, knowing exactly what they should be talking about with our customers. And so I don't think it's friction internally. I think it's just how quickly we can build the internal agents and the internal products to be faster.

At the dealer, I think it's those 2 vectors and how quickly they can change behavior to capitalize on those vectors. And so that's what we're trying to help them do with account management. we are -- and how that translates -- you asked how that translates into the results. I mean, I think that's about growth and speed of growth for us, and that's how we're thinking about it, more so than a margin enhancer in the near term.

Thank you. Ladies and gentlemen, that concludes our question-and-answer session for today. I would now like to hand the conference over to Jason Trevisan for closing comments.

Thanks. I would just like to thank all of our colleagues certainly here at the company, all of our customers and everyone who joined us on this call tonight. Have a great evening.

Thank you. This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.

All right. Well, it's my pleasure to be joined on stage by Jason Trevisan, CEO of CarGurus. Jason, thanks for coming to the conference and helping kick off the company session.

Of course. Thanks for having us.

Sure. So maybe just to level set, for those less familiar with the story, maybe just talk about the evolution of CarGurus and what that journey has been on and some of the key messages coming out of the quarter.

We began as a 2-sided marketplace, helping consumers shop for and buy cars and helping dealers get access to the largest audience of car shoppers. And so it's a 2-sided marketplace. And we entered a market in the U.S. that was -- that already had existing competitors, but we came with a different model. We came with a freemium model where we invited all inventory and all dealers onto our platform, which allowed us to give consumers more access to more inventory. We gave them more information on the inventory and on the cars and then we sorted it in a way that made sense to them largely through our deal rating system. And that allowed us to build the largest consumer audience and a 2-sided marketplace as you get the network effects, we also then built the largest dealer base as well.

So today, we are by every measure, the largest marketplace. Since then, we have expanded with both dealers and consumers. So among dealers, we've helped them with not just market their cars, but with more aspects of running their dealerships. So we now have data and technology that helps them source better merchandise, price and sell cars. And then we help consumers a little more upstream with determining what types of cars to buy. So what we call the consideration step. We've always helped them with the decision step of which car to buy. And then we're also helping them purchase the car by allowing them to do more steps of the purchase on our site.

And maybe before we get into the more product and platform stuff, I'd love just to get your perspective on what you're seeing from a broader macro perspective in the auto market, any impact from tariffs that you're seeing and your base case around broader macro and impact on the platform?

So we -- as the largest marketplace, we are not terribly affected positively or negatively by a lot of macro effects in auto. As dealers are trying to sell cars and consumers are searching for cars, we play a very valuable role. Right now, we're seeing pretty strong demand for cars. Tariffs have not translated into -- the auto manufacturers of new cars have not passed a lot of those tariffs on to the consumers yet. They've reduced some rebates. But it has had a marginal effect on raising new car prices, and it's had a marginal effect on raising used car prices, but demand has remained strong.

Consumer confidence, though, is quite low and cost of borrowing are still quite high. And so consumers are shopping much more price sensitive now than they were a year or 2 ago. Inventory levels are back up to pre-COVID levels. And so you've got pretty strong inventory. You've got pretty strong demand and -- but you've got a consumer that's certainly feeling pinched and so therefore, searching for lower-priced cars.

Okay. You recently wound down or made the decision to wind down the CarOffer transaction business following the strategic reassessment. Maybe just talk about that decision, what went into that as you were assessing that business? Any impact to financials that you're flagging to folks for the back half of this year? And then how you anticipate the underlying technology of that business still being able to be utilized inside of or enhancing the broader platform?

Sure. Our belief has always been that in order for dealers to sell cars well, they need to use data and intelligence across the entire spectrum of their workflow, and that starts with sourcing the right type of car. So we have long believed that in order to sell cars effectively, they need to buy the right types of cars and the right cars at the right prices. And so sourcing has always been part of our effort to expand solving the pain points of dealers.

So we -- so CarOffer is a digital wholesale transaction platform. Unlike that segment of the market, which has historically run with an auction model, CarOffer ran with what's called an instant trade model, which is different and unique from an auction. And during COVID, it skyrocketed. It's very easy to buy and sell on the platform when dealers have high confidence in the prices of cars in wholesale. Since then, since post-COVID, it has struggled to grow volume.

And so there are a couple of different parts of the business, if you will. One is the actual running of the transactions themselves. The other is the technology that we built around CarOffer. And that technology is stocking recommendations, appraisal and price assessments, acquisition, inventory acquisition intelligence, so telling dealers which cars to buy, how much to pay for them and being able to predict, and a lot of this is using AI, being able to predict how quickly they will sell those cars at what prices and how much gross margin they can make.

So we made the decision and announced at our last earnings to wind down the transaction piece of that business. But we are taking the technology and the analytics, and we are bringing that and making that available to all of our CarGurus customers with a software-like and data approach rather than a logistics-intensive transaction approach of CarOffer.

So on August 7, we announced at earnings, we announced the wind down. We said -- from your second part of your question, the financial implications, we said there will be between $14 million and $19 million of total expenditure, the bulk of which will happen in the second half of the year. At the time, we gave a marketplace guide on revenue and EBITDA, which is different than historically when we've given a consolidated. And we did that because we view that as much more relevant to the go-forward nature of our business. So we gave a marketplace revenue and EBITDA guide. And just to note for all of you that some of the aggregators are showing that marketplace guide as the consolidated guide. And that does not take into account CarOffer losses. In Q3, because we're winding down transactions, there will be lower fixed cost coverage. And so CarOffer burn in Q3 is likely to be higher than in the last couple of quarters because of that wind down of transactions. And so -- and then in Q4, it will be discontinued operations.

Okay. Well, that's a good segue into the marketplace business. Maybe just help us understand, as you think about the growth algorithm for the core of the business, how do you think about the number of dealers being a driver versus wallet share. I think you guys disclosed a metric called QARSD. And just within that QARSD metric, how you think about the drivers of wallet share and the rank order of the various factors that help drive that growth?

So we have U.S. -- the bulk of our business is the U.S., but we also have marketplaces in the U.K. and Canada that are growing much faster than our U.S. business and have more runway. And so if you look at the total spend in those 3 countries of dealers in marketplaces, it's about $4.5 billion. And today, we're less than 30% market share, but yet we're the largest audience. So we see a lot of upside there.

So the 2 metrics we look at, one is quarterly average revenue spend per dealer or QARSD and the other is number of dealers on our platform, paying dealers on our platform. If I focus on the U.S., we have about 26,000 paying dealers today, and there's about 42,000 to 44,000 total dealers in the country. We have -- we still have a freemium model. And so total dealers on our platform is over 30,000. We've been growing our dealer rooftop count. But that, I would say, honestly, is not a key focus of ours. A key focus of ours is just gaining more wallet share in total. And so for us, the more important factor over the last few years has been QARSD and growing QARSD.

So the way we grow QARSD is there's a handful of key levers. Number one is upselling to higher tiers of our product. We do that. They get dealers get more benefit to higher tiers, but we've also been innovating a lot of data and intelligence and including those, bundling those into the higher tiers. And so upsell has been our #1 driver over the last couple of years. And yet we still have less than half of our dealers who are beyond the basic package. So we have a lot of runway there.

Second is add-on products. So these are not bundled, but they are other marketing-related products that dealers can buy from us. We have a handful of those, and so adding products is the second.

The third is, as we grow lead volume and grow lead quality, that grows the value to dealers because they will sell more cars as a result of working with us. And so that is just -- as long as we keep growing leads, and we've said we're growing leads. And as long as we keep growing quality, which we've indicated we have, then that's value that just grows by itself.

And then the last is unit pricing. So we started in order to earn our way to the leadership position, we priced below our competitors. We believe we are still priced on a unit basis below our competitors in most cases. And so that's been a very -- a small lever for us, but it's been a consistent one that we think has a long runway as well. So that's all within marketing.

And then as I mentioned earlier, as we move out into other steps of the dealer workflow, we have every intention to bolster the technology and the analytics that we're delivering to them such that we can start to tap into their wallets that they're spending on data and their wallets that they're spending on software.

Okay. And that reported metric is obviously a blend of newer dealers coming into the platform and funnel and older kind of more mature dealers that have been on a journey already. Is there a way to illustrate what some of your most mature dealers look like and what that journey or adoption curve look like from either a product attach rate standpoint or what the journey looks like to move up the subscription tiers to just illustrate what -- how that cohort curve can look?

Sure. I mean I think -- so one helpful data point is that the -- if you look at the cohort of dealers who have been on our platform for 5 years and compare it to the cohort of dealers who have been on our platform for 1 year, the 5-year cohort is spending 80% more on average per rooftop than the 1-year cohort. And so that's growing with us through all of the QARSD levers that I described.

And also, it's just dealers that stay with us longer and are with us longer learn more about how to use all of the features and analytics that we provide. We're getting better at helping them do that. We're delivering it in different methods in the app, in the dashboard, in e-mail, through account management. So we're getting better at helping them, but those that really buy into the system are far more effective than those who just view us as a more simple marketplace.

Okay. Yes. And it seems like the strategy is to become a more integrated partner with your dealer partners over time. Maybe talk about what some of those key unlocks have been as the platform has evolved. Talk about some of the specific products that you've rolled out that have allowed that deeper, more tight relationship with your dealer partners.

Sure. So part of it is establishing a relationship with more people at the dealership. And so I mentioned an inventory acquisition report. That is typically focused on the person at the dealership who is in charge of sourcing cars, which is different than the person, say, the Internet Marketing Manager who's focused on marketing the cars. So we're building relationships with more roles at the dealership by providing some of this.

Pricing is a big area where we've had a lot of progress, and that's changed behaviors at dealers. And so we have a couple of products. One is called Next Best Deal Rating, and one is called Max Margin. And that helps dealers understand how they can make oftentimes small changes to the pricing of their cars and have an outsized positive impact on either the leads they're generating from our site or the margin they're making on the car. But what's -- and so now we have over half of our dealers getting that information in those reports.

What's most exciting to me is that they're getting those daily and weekly. And those are now going to multiple people at the dealership. It's going to the GM, it's going to the Internet Marketing Manager. And so their engagement in our platform and their habitual use of our analytics has gone from occasionally going into our dashboard to something that they're using every day.

Okay. Let's maybe move over to the consumer shopper side of the platform and talk about how the company has innovated on the consumer side and the car buying journey to drive deeper engagement with you, better conversion for your dealer partners, reduce friction overall. Maybe just talk about some of the product and feature evolution that you've gone through there?

Sure. I'll take it in a -- on a few different dimensions. One is, I mentioned earlier that we're helping with more steps of the car buying journey. So there is an initial period where someone is trying to figure out what types of cars exist and what they might want to think about getting. There's then sort of a deeper consideration of what type of car to get. There's then the decision of which car to go pursue and then there's the purchase of it.

So on the consideration piece, we have an AI-based virtual assistant that helps consumers figure out what type of car they want based on colloquial inputs that they have, how many kids they have, what they need the car to do, what aspects of a car they like and don't like and helps get them to make model trim and then actual cars that are available. Not only is that growing the conversion rate of our audience and really deepening the engagement that they have with us, but it's giving us really valuable data that we're then passing on to the dealer so that when the dealer gets this lead and knows all of this about Jason, they're able to convert that lead better. So that's in the consideration piece.

On the decision piece, we, again, mostly AI-driven. We're getting much better at personalization. We are getting much better at sort order and recommendations. And so again, that's making for more engagement. We have -- our audience is much more engaged in our platform than any others. We have 75%, I believe, percent more session time than our next closest competitor. So that's in the decision piece.

And then in the purchase piece, we're helping consumers do a lot more on our site so that they can make their time in the dealership more efficient. So we now help consumers get a trade-in value. They can get prequalified or fully qualified for financing. They can set up an appointment. They can put down a deposit, all on our platform. And so when they get to the dealership, instead of spending 5 hours, they come in, dealership knows them much better, they're spending 1 to 2 hours.

And is there a flywheel or feedback loop where the better the experience, the reduction in friction on the consumer or shopper side leads or gives -- empowers you to go to dealers that drives more of that higher product attach rate or moving up subscription tiers. Does it kind of helps the dealer side as well in terms of that wallet share story that you're talking about?

It definitely does, but it is also hard to measure. So we focus a lot on attribution and helping dealers understand not only did we send you 150 leads this month and not only did we give you all these other capabilities and technology and analytics. But importantly, that translated into x number of sales for you a month. And we focus a lot on quality. So we know, for instance, the type of lead that has a higher propensity to close, and we incorporate that into our performance marketing way upstream. And so we track a metric of estimated sold cars, and that's how we measure our success, and that's growing. And the more we're able to help dealers understand that, then yes, the more they're buying into the system.

Okay. Another part of the key strategic initiative story that you've talked about a lot on prior calls is moving more transactions towards the bottom of funnel. Maybe just talk about that initiative, talk about digital deal, what that is, how that's progressed? And then additionally, how like financing plays into driving more of the transactions-oriented stuff at the bottom of the funnel?

Yes, dealers -- When COVID happened, dealers materially reduced their sales team's size and have not really built them back up. And so you'll hear a lot of dealers now say that they have trouble getting the sales staff that they do have to actually effectively respond to all the leads. And so dealers are always looking for signals as to which is a higher quality lead so they can prioritize and be most efficient with their sales team's time.

And so we have a number of ways to further qualify consumers on our site, which also creates huge consumer satisfaction. And the examples I just mentioned are consumers can do elements of the transaction on our site. So appointment, down payment, trade-in value. They can buy other products from the dealer. Dealers have -- as any of you have bought a car know, they are going to try to sell you 5 to 10 other things that they have. They can market those and sell those on our site to the consumer as well. And so it makes the consumer happier. The vast majority of consumers say they want to do more online. While full online transactions are growing quickly, it's still a very small percent. It's low single-digit percentages of consumers that want to do it all online. So most consumers want to do most of it online, but then still go into the dealership for the test drive and to complete the transaction.

And so we are trying to find a win-win always between consumers and dealers to make that more efficient. And it then makes the conversion of those leads much more efficient for the dealer, which is much more profitable for them.

Okay. We talked about it a little bit earlier. You touched on it is international. Maybe talk about just the international efforts in Canada and the U.K., how that has progressed relative to the progression in the U.S. over time? And any plans to extend that to other countries beyond the U.K. and Canada anytime soon?

Canada and the U.K. are a similar playbook to what we've executed in the U.S., but a few years behind. So in -- I'll talk -- there are a lot of parallels or similarities between the 2 countries, but I'll talk about them independently.

So in Canada, which was the first country we entered, AutoTrader Canada is the current leader and we are the #2 and growing very quickly. So our international business in Canada has been growing 20% to 30% for the last few years. The market leader has been growing much less. It's private, but from our data points. And so we're gaining market share. And importantly there, our lead volume is starting to get close to AutoTrader Canada's lead volume, which is allowing dealers to finally -- and we charge materially less. So it's a similar playbook to the U.S. of we're going to grow our consumer audience. We're going to focus on lead quality and quantity and we're going to charge less. So the ROI story is bulletproof. And so with getting to the lead volumes we're at, we're now having dealers who are finally saying, finally, I'm able to work exclusively with CarGurus.

So AutoCanada, one of the largest groups in the country, just released a press announcement that they're making us their exclusive marketplace provider. So that market is doing very well for us. App is growing exceedingly fast as well. And as we introduce -- build and introduce new products in the U.S., we're then bringing the majority of those to Canada where -- which is helping fuel the growth.

U.K., similar in so much as there's a market leader, Auto Trader U.K. and we are the #2, similar story that we're growing much faster than the market leader. So we're gaining share. There, we are a little bit further behind in lead quantity. So we're still growing our consumer audience there. But the ROI in a recent survey, we were just deemed by dealers the strongest ROI that they work with. And we're the #1 download app. So -- in the auto category.

So in both markets, you're seeing this slower growing very high margin incumbent that, frankly, dealers are eager to find competition for because they're displeased in a lot of cases. And we are the highly innovative, highest ROI, fastest-growing consumer audience and rapidly catching up to them.

I did want to shift here and talk about AI. Obviously, it's fast evolving and moving. What are you seeing in terms of any impact at all from AI search, zero-click search on your business in terms of customer acquisition and the broader funnel for search?

We are very focused on it. We're spending a lot of time making sure that we understand ways in which AI is a threat to our business and the ways in which AI is an opportunity in our business, and it is certainly both.

From an audience acquisition perspective, I would say there's probably 3 dimensions. Number one is we are aggressive users of the AI tools that the performance marketing platforms are building. And so it's helping us become more efficient in places like Google and Meta and so forth. So it's -- we're more efficient as a result of AI in the performance marketing channels, presumably because we're leaning harder into it than our competitors are.

Second way is making sure that we make our data and our content available for the generative AI search engines themselves, so the SEO for AI and that is more of a technology and content strategy, and it's still a very small percent of total searches, but we're excited by how we're showing up there. Again, it's all relative to the competition.

And I would say in that, the zero-click search that you mentioned, car shopping is rarely a zero-click or one-click activity. If you're searching for a data point or an answer to a question, it's fantastic. If you're searching for the right -- the perfect Subaru Outback that you want for your family, you're going to need to do some searching over 3 to 6 months, and that's going to be informed by generative AI responses, but it really is going to be ultimately decided by a platform like ours where the consumer goes deep over a period of time has multiple conversations, et cetera. So we have not seen a lot of our audience gets siphoned off to zero-click because they still need to do the searching process. That's the second.

And then the third I would say is we are diversifying our spend pretty aggressively. We've started this maybe 2 years ago to be less and less concentrated in performance marketing. And so we have spent a larger percentage of our marketing on brand and on awareness. We're tapping much more aggressively into online video, TV, social influencers so that we develop a more direct relationship with the consumer and a -- and it's working for us. So our app usage and the percent of our volume that comes from app is up significantly. Direct is one of our fastest-growing channels, and that's the type of relationship that circumvents any third parties because that's us directly with the consumer. And so app is a huge focus area for us.

Okay. And maybe it's too early and not entirely representative, but are you noticing any difference in the type of audience that's coming into the funnel or engaging from an AI search generative search versus traditional SEO, SEM in terms of conversion or demographics? Any noticeable difference yet?

We have not seen that yet. As I mentioned earlier, we are -- we optimize for the actual conversion to a sold car. And so that would get captured in our algorithm. What we are seeing, though, is that users -- our users who are using our virtual assistant and are using the AI to help them find the car are much more likely to convert to a lead and then much more likely to go from a lead to a sold car. And so as our use of AI helps create a smarter customer or consumer on our site, that's leading to a much more qualified consumer at the dealer.

Okay. And maybe last one on this topic. Just as you've obviously built out a muscle over time on SEO and SEM, how does that help you? Or how have you seen that help you show up more prominently in AI search? Or is that not something that necessarily translates in terms of optimizing for making sure that you're prominently showing up?

No, I think it does. I think it does translate. Number one, we have the most data. We have the most searches. We have, by far, the most intelligence and visibility on car shopping behaviors in the U.S. And so simply having that breadth and depth of content helps. We have the most inventory of cars. We have the most consumers, et cetera. So that puts us in a good starting position.

We have tried to be really smart about making that information accessible while still without giving it away too much so that it recommends us I think we've done a good job with that. And then the second -- the third factor I would say is that we have started to develop more content over the last couple of years, largely using AI. We've never had a large content team or large content investment. And we have said we're comfortable if we're not very high up in the funnel to a searcher who says, "I think I want to sit in, but I'm not sure which one." We've not focused as much there. But with AI, we're producing multiples, multiples more content than we have before. And it's still not for that trying to compete with car and driver necessarily. It's more in the, "Well, how should I think about Audi A4 versus BMW 323 or whatever. Yes. And so having that content helps significantly with the GEO.

Okay. I wanted to shift to the growth versus margin dynamic. And we talked a lot about priorities for growth around investing in product innovation, customer acquisition and international. What's the messaging around the priorities for investing for growth versus showing kind of improving margins, scaling free cash flow, et cetera?

When you focus on -- or when you isolate for our marketplace business, so excluding CarOffer, we have shown really nice margin expansion over the past few years, and it's now in the mid-30s. And we -- and that was our long-term margin target that we've indicated for years. And so we feel comfortable in that range.

And when you look at Rule of 40 and how we stack up, it's very, very strong. And so our focus going forward is much more on the growth levers that I talked about before and specifically, so continuing to grow our leadership position in the marketplace model, which -- we think it's a winner take most. We think there's a long runway there. So we're going to continue to do that, but also innovating to cement that, but also innovating to solve for more of the dealer workflow and by doing that, we think we're going to be able to build -- continue to build great features, but move those from features to actual data products and actual software products. And so as we do that, that will tap us into the $8 billion that dealers spend on software and get us out of simply the wallets that they think about from marketplace.

So long story -- or long answer to your question, we're much more focused on sustaining our growth for many, many years to come and less focused on continuing to expand the margin.

And are you seeing any efficiencies on the cost side from incorporating AI into your own kind of workflows and how meaningful or not has that been?

We certainly are. I would say it has been modest thus far, but we're seeing it in -- certainly in our engineering and code creation. We're seeing it in our product development cycles, we're seeing it in account management and sales. So our sales team is maintaining higher productivity and efficiency than they otherwise wouldn't be more effective with the customer. We're seeing it in smaller ways in finance and legal and HR. We have never had a big customer support piece of our business, and so that's not as relevant for us. So we have the company focused on how we can use AI to move faster and to build more for our customers and less around cost cutting, but we're seeing a little bit of both.

Okay. In the last couple of minutes here, I did want to end on capital allocation. You're starting to scale free cash flow at a pretty healthy rate. You just increased the share repurchase program. So maybe just talk about how investors should think about your priorities for capital allocation between investing in some of those product and feature initiatives returning capital, any kind of M&A that may or may not fit the platform? What's the messaging around that?

The messaging is similar to what it's been the last couple of years, and I think we've shown. We've executed on what we indicated. So 3 main options or opportunities. One is to invest in the business. And so I just spoke to that in the form of margin. The second is returning capital to shareholders. And so over the last few years, we've bought back over $0.5 billion of our shares. We do that assessment based on looking at our expected free cash flow returns to the business based on the share price. And as a result, we've been pretty aggressive. And as you said, we just expanded it another $150 million to last earnings.

And then the third is M&A. And we continue to look at a lot of M&A that is always a muscle that we're using. We've acquired 3 businesses in the past and we expect to acquire more. The area that we're most focused on is in that dealer software and data analytics so that we can continue to expand with dealers and get their get them using our platform more and more than they have in the past. And I think that's evidenced by dealers are at a much faster clip signing up with us for year-long contracts. And historically, it was a month-to-month. So that shows the commitment as well as the behavior.

Great. With that, I think we're out of time. We'll end it there. Please join me in thanking Jason and the team from CarGurus for being at the conference.

Thank you, Ben. Thanks.