backthought Operator: Greetings, and welcome to the CarGurus third quarter 2026 earnings conference call. At this time, all participants are in a listen-only mode.

A question-and-answer session will follow the formal presentation. If anyone should require operator assistance, please press star zero on your telephone keypad.

As a reminder, this conference is being recorded. It is now my pleasure to introduce Kirndeep Singh, Vice President 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; Matt Mandel, Chief Financial 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. Before we begin, I'd first like to welcome Matt Mandel to his first earnings call as our Chief Financial Officer.

Matt joined us in the second half of October and brings extensive financial, operational, and strategic leadership in rapidly growing companies, as well as a strong track record for developing cross-functional teams. His experience will be valuable as we continue to focus on expanding our platform across dealer workflows and the consumer shopping journey, all powered by our leading marketplace data.

Turning to the quarter, we delivered strong top- and bottom-line results in the third quarter. Revenue grew 12% year-over-year to $258 million, at the high end of our guidance range.

Adjusted EBITDA was $89.7 million, near the high end of the range, with an adjusted EBITDA margin of 35%, and we again generated robust free cash flow. I'll start by providing an update on the dealer environment, because it frames much of what I will cover.

As we noted last quarter, some of the global macro factors that weighed on dealers early in 2026 improved modestly over the course of the first half of the year. We see this trend slowly continue.

While we've been encouraged by the sequential improvement at the dealer level, we do not believe that dealer enthusiasm has returned to 2025 levels, driven primarily by dealer margin pressure. For the consumer, affordability remains stretched, and rising borrowing costs have added to that pressure.

These factors have continued to foster deliberate spending decisions among dealers. Despite these pressures, dealer activity in the quarter moved in the right direction.

Net US dealer adds reaccelerated, up 251 sequentially, our strongest quarter of net adds this year. Even with a tough backdrop for both consumers and dealers, CarGurus has gained share on both sides of the marketplace.

We continue to grow wallet share driven by our strong ROI and adoption of our dealer products beyond listings. US paying dealers that subscribe to at least one product outside our core listings tiers grew approximately 16% year-over-year.

Average sessions per paying dealer continued to grow year-over-year as engagement grew stronger. More engaged dealers have historically seen better results, which supports stronger retention, while broader product adoption diversifies our business.

Our listings, software, and data work together across inventory, marketing, and lead conversion, helping dealers get more value from the platform. On the consumer side, we outpace competitors in visitors and sessions, while also continuing to lead in AI referral traffic and AI visibility across categories we track, reinforcing our position as the number one most visited automotive marketplace in the US.

As we expand beyond listings, we're building products that support more of the decisions dealers and shoppers make every day. Using these offerings brings both sides of our marketplace into more frequent contact with our platform, deepening engagement and generating additional information about their needs and behavior.

Those interactions contribute to the nearly two decades worth of data, including approximately half a billion first-party signals we capture daily across demand, pricing intelligence, live inventory, and first-party consumer intent and behavior, many of which exist only on CarGurus. In addition, on the dealer side, we capture workflow data points from lead quality, conversion signals, response times, engagement, and dealer interactions.

We use these insights to improve our software and AI to help inform what we develop next. This is the virtuous cycle we have described, with broader use of our offerings informing development and helping us make our products more useful.

AI is also improving how we develop and deliver those products by increasing productivity across the company. This helps us accelerate innovation and bring products to market thoughtfully while managing costs.

We believe this approach supports our business across market environments and will continue to drive growth across our three value-creation drivers. I'll now walk through our third-quarter progress across each of those drivers.

Driver number one: expanding CarGurus' offerings into integral parts of the dealer workflow, connecting inventory, marketing, lead conversion, and data pillars through mutually reinforcing products. On the dealer side, product expansion has led to more dealer engagement, which drives performance across our platform.

Each product we have launched addresses a decision a dealer makes every day: which vehicles to acquire, how to price them, how to merchandise them, and which shoppers to prioritize. Early data indicate that dealers have seen increased performance gains when combining multiple products.

In our inventory pillar, we believe the opportunity is significant, and we are continuing to work toward an integrated suite of products that supports each step of a dealer's inventory workflow: sourcing, stocking, appraising, merchandising, and pricing. Our integrated products draw on our marketplace data, each decision a dealer makes informs the next, through predictive intelligence determining what will sell, how quickly, and for how much.

That starts with Sell My Car, which gives dealers efficient access to inventory sourced directly from consumers at a time when acquiring used vehicles remains a top challenge. Earlier this week, we launched StockVantage, which addresses the next decision in that workflow: choosing which vehicles to acquire.

It uses AI to combine retail demand signals from across the CarGurus marketplace with dealer-specific inventory fit, profit potential, and expected turn time, and shows the retail potential of a vehicle alongside wholesale values and vehicle history at a local level for dealers. That lets StockVantage sift through thousands of auction vehicles to surface those with the highest potential, buyers can focus on the units most relevant to their lot.

It applies the same intelligence to consumer trade-ins with a VIN scan, making it a strong complement to Sell My Car. Once that vehicle is on the lot, PriceVantage carries the decision through to retail pricing.

It has continued to demonstrate strong adoption, and dealers using it saw a median 14% lift in VDP views, while engaged franchise dealers saw 1.5 times the lead lift of dealers who haven't acted on a recommendation. More importantly, dealers reduced overall discounting by nearly 10% per unit while maintaining turn times, selling just as quickly while preserving more value on each sale.

Together, StockVantage and PriceVantage link what a dealer buys to how it is priced, helping them improve turn time and price realization, leveraging our leading data, signals, and intelligence. In our marketing pillar, VINMAX has had one of the fastest early adoption of any add-on product we've launched.

This extends CarGurus' consumer demand data and AI-powered predictive intelligence into marketing for the first time. VINMAX finds underperforming vehicles and leverages AI to boost them automatically across our merchandising products.

As a result, dealers are getting more leads and more views, with promotion decisions driven by CarGurus' real-time supply and demand data. Many of the vehicles VINMAX boosts are the ones a dealer would otherwise mark down to move.

Those boosted listings are 85% more likely to sell without any price reduction compared to similar non-boosted vehicles, helping dealers hold prices as they navigate current market conditions. VINMAX-boosted listings receive roughly 57% more leads per day and roughly 67% more VDP views per day versus comparable non-boosted listings.

In our conversion pillar, Shopper Signals is becoming a more regular part of the dealer sales process. Nearly 70% of eligible dealers used it during a recent 30-day period.

The profiles bring together buying readiness, preferences, and activity to help dealers prioritize follow-up. Hot shoppers show up to 4x the average close rate.

As we continue updating each profile, dealers can spot shoppers who may have gone quiet in their CRM, but are still active on CarGurus. Over time, that creates a dynamic, valuable prospect base for the dealer, one they continue to build while they remain on our platform.

In our data pillar, we leverage the same retail signals that underpin every product I just described and package them into our market and competitive intelligence. This enables dealers to track store-level KPIs, benchmark performance against competitors, and examine their positioning in specific markets through the dealer engagement platform.

These insights have been well received, with engagement up nearly 30% over the last 90 days and our email engagement achieving an open rate of greater than 70%. Taken together, these products put CarGurus inside the decisions dealers make every day, well beyond listings.

We believe that engagement leads to greater adoption, stronger retention, and more wallet share over time. Driver two: focusing 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.

AI is changing where research begins for a small but rapidly growing portion of shoppers, but we do not believe it changes what shoppers need: current inventory, transparent pricing, dealer reviews, dealer ratings, and guidance they can trust, which ultimately all drive confidence. We believe that is why AI platforms send more car shopping traffic to CarGurus than to any other automotive site.

We expect CarGurus to become more valuable in that world, not less, for five reasons. First, trust.

Trust is a foundational layer of the shopping experience. CarGurus has earned that trust over more than a decade, growing to 46 million monthly shoppers globally, establishing ourselves as the brand consumers turn to.

We were the first to show a listing's price change history and days on market, and are the only car shopping site to identify overpriced listings. We were also among the first platforms to show all-in pricing and to remove deal ratings from vehicles that do not disclose their fees, while continuing to find ways to make the car shopping journey seamless.

Second, scale. We are the number one visited automotive shopping site in the US, with the largest consumer audience and the most visited and most used car shopping app.

We have the largest base of paying dealers and the largest inventory selection. Shoppers who use our platform instantly benefit from the true two-sided marketplace we have built, with more selection and more information on every car, while dealers reach the largest audience of buyers.

This combined reach creates a compounding effect as our liquidity and marketplace grow. That is hard to replicate, and we believe it provides a durable advantage.

Third, proprietary data. Dealer actions and the shopper signals we capture every day make up a data set of over 10 billion intent signals from shoppers every month.

These include demand signals, consumer and dealer interactions, dealer demand trends, and local market supply. Together, these signals are unique to us and help us understand what shoppers are looking for, we can show them more relevant vehicles and provide guidance tailored to their needs.

Because we know each shopper as an individual, that personalization follows them through the journey, from early research to the lot. That same data supports the products we build for dealers, they can serve those shoppers better.

Fourth, embedded workflow. Car shopping has historically been a series of disconnected steps.

We connect those steps into one journey, from the first search to a scheduled visit to the moment a shopper stands on the lot with Dealership Mode. We are embedded inside the dealership itself, in how dealers source, price, and market their inventory, the dealer is connected at every point along the way.

Horizontal solutions are not integrated to that degree across the full transaction. We are closer to the transaction.

Our ability to carry the shopper through every stage of the purchase or operate inside the dealership is what sets us apart. Fifth, reach.

We meet shoppers wherever they begin, whether with Guru on CarGurus, our app inside ChatGPT, and now the Claude connector we launched this quarter. These products put CarGurus' inventory, deal ratings, and dealer connections inside the interfaces consumers are already using on their car shopping journey.

As shopper trends evolve, we will continue to evolve right alongside them. An example of this would be through the continuous development of Guru into a vertically integrated, full-service agent, which we believe is a stronger long-term solution compared to horizontal agents in the market today.

Together, we believe these capabilities give us the foundation to make CarGurus the most trusted, auto-specific expert guide across the full shopping experience. This is a journey, not a destination.

As discovery paths evolve, we want to bring CarGurus' data and trusted experience to shoppers wherever they begin their search. That includes the new generation of shopping agents.

We see AI agents and similar platforms as an expansion of our marketplace's reach and utility. We have a history of remaining nimble and taking advantage of new technologies, and agents are no different.

We believe that Guru is central to that strategy. It brings together our proprietary automotive data with an AI interface to support shoppers across research and consideration.

We've continued to add our own research and data so shoppers can access CarGurus' market knowledge directly within the experience. Then, as shoppers move toward purchase, Dealership Mode carries that guidance onto the lot, where they often need expert context most.

The value comes from embedding those experiences. Every action with a consumer or dealer creates another signal about intent, preferences, and outcomes.

Today, the intent a shopper shows on CarGurus, the vehicles they save, compare, and return to, flows through to the dealer they contact, the dealer understands what matters to that shopper before the first conversation and can close with more precision. Over time, that makes the experience more useful on both sides of the marketplace.

That strategy has begun to show in engagement. Guru usage reached new highs in the third quarter: unique usage up 24% sequentially and a growing share of shoppers returning to continue prior conversations.

In addition, our consumer app redesign improved in-app lead conversion, and we plan to continue launching additional capabilities around AI assistance, such as our text response assistant that helps dealers respond to and foster leads. At the point of purchase, Dealership Mode reached an all-time high in usage in the third quarter, up 105% from the second quarter.

We have connected verified online activity to lot visits for a growing number of dealers. This connection gives dealers more context on what happened before the shopper arrived and connects the online journey to the in-store experience.

Our objective is to make CarGurus the expert guide that has the shopper's back from research through consideration and purchase. With our trusted brand, scale, auto-specific data, embedded workflows, and reach across new interfaces, we want CarGurus to be the obvious choice for any car buyer wherever they start.

We believe that gives us a clear path to win as AI changes how consumers shop. Driver number three: deploying capital with discipline with the aim of growing long-term earnings power and stockholder value.

Our capital allocation priorities remain a relentless focus on investing in innovation across both the dealer and consumer side while operating with discipline, evaluating strategic opportunities, and opportunistically returning capital to stockholders. This is all supported by our strong non-GAAP free cash flow engine, which fuels our capital allocation strategy.

2026 is an investment year by design, and we have delivered on it. Since 2025, we have brought six new products to market across our dealer pillars, growing our TAM and wallet share potential while deepening engagement with dealers and consumers, and we are doing so more efficiently than ever before.

We believe that combination of innovation and disciplined capital stewardship is what creates durable value. We plan to continue investing internally where we see the strongest long-term returns, preserving flexibility for targeted M&A, and returning capital to stockholders.

We plan to continue repurchasing shares when we believe it is an attractive investment and consistent with our capital deployment priorities. As such, we are announcing a new repurchase authorization.

Looking ahead, while macro factors continue to pressure dealer spend, we're excited by what the third quarter showed across all three value-creation drivers. This year, we brought new products to market faster than at any time in our history, and we're driving value for both consumers and dealers across the entire journey.

On the dealer side, a growing share of our US dealers now use more than listings. VINMAX had one of the fastest early adoptions of any add-on product we've ever launched, and our inventory pillar with our launch of StockVantage is becoming an integrated suite that our dealers rely on.

Our listings business remains ranked number one by dealers in ROI, with our franchise dealers saying they're more satisfied with our lead quality and quantity than our competitors, and we continue to gain wallet share and are encouraged by our dealer adds in the quarter. On the consumer side, Guru set new highs in usage, and we're meeting shoppers inside the AI tools they already use, building toward being the expert guide for car shopping in an AI-led world.

What connects both is our underlying data layer: proprietary demand, pricing, inventory, and engagement signals that get richer with every interaction and harder for anyone else to replicate. We believe the result is a stickier business and better liquidity and transparency between consumers and dealers.

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

Thanks, Jason, and good afternoon, everyone. I'm pleased to be joining you for my first earnings call as Chief Financial Officer of CarGurus.

I'm excited to be part of the team, and I look forward to getting to know many of you in the months ahead. Now let me walk you through our financial results, followed by our guidance for the fourth quarter and full year 2026.

Third-quarter revenue grew 12% year-over-year to $258 million, at the high end of our guidance range, driven by expansion of existing dealers and growth in new dealer subscriptions. In the third quarter, US CarSight grew 7% year-over-year, and we added 659 paying US dealers year-over-year.

We continued to increase our dealer base while taking greater wallet share. In the third quarter, year-over-year CarSight growth was driven primarily by adoption of add-on products, followed by listing upgrades.

Our international business reported third-quarter revenue of 22% year-over-year, driven by account expansion with listings upgrades and adoption of add-on products and like-for-like price increases. I will now discuss our profitability and expenses on a non-GAAP basis.

Third-quarter non-GAAP gross profit grew 10% year-over-year to $237 million. Third-quarter non-GAAP gross margin was 92%, down about 90 basis points year-over-year.

Third-quarter non-GAAP adjusted EBITDA grew 9% year-over-year to $89.7 million, toward the high end of our guidance range, and adjusted EBITDA margin was roughly 35%, down about 90 basis points year-over-year. Adjusted EBITDA came in near the high end of our guidance range, supported by strong flow-through from our revenue performance and a favorable true-up related to the real estate tax reassessment of our corporate headquarters in Boston.

Third-quarter non-GAAP operating expenses totaled $156 million, up 13% year-over-year, reflecting higher sales and marketing expense from performance marketing and brand awareness campaigns and growth in our sales organization to support our expanding inventory products. We also continued to invest in product, technology, and development headcount.

Third-quarter non-GAAP net income per diluted share attributable to common stockholders was 70 cents, up 17% year-over-year. We generated strong cash flow, converting 82% of our adjusted EBITDA, or $74 million, to free cash flow in the quarter.

We ended the quarter with $165 million in cash and cash equivalents, an increase of $43 million from the end of the second quarter, as $21 million in share repurchases was more than offset by our cash generation. As of the end of Q3, we have $25 million remaining on our 2026 authorization.

Additionally, I'm pleased to share that our board has authorized a new $250 million share repurchase program, which will be effective November 4th, 2026 through December 31st, 2027. 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 fourth quarter and full year 2026. We expect our fourth-quarter revenue to be in the range of $258.5 million to $263.5 million, up between 7.2% and 9.3% year-over-year, reflecting continued deliberate dealer spending decisions due to margin pressure and lighter new bookings in the first half of this year.

For the fourth quarter, we expect our non-GAAP adjusted EBITDA to be in the range of $93 million to $101 million. We expect fourth-quarter non-GAAP earnings per share to be in the range of 73 cents to 79 cents, and diluted weighted average common shares outstanding to be approximately 90 million.

Turning to the full year, we now expect 2026 revenue in the range of $1 billion 11.5 million to $1 billion 16.5 million, representing growth of 11.5% to 12.1% year-over-year. We expect full-year non-GAAP adjusted EBITDA to be in the range of $347.5 million to $355.5 million, a margin change of -80 basis points to -20 basis points year-over-year.

We expect full-year non-GAAP earnings per share to be in the range of $2.66 to $2.72, and diluted weighted average common shares outstanding to be approximately 92 million. With that, I look forward to meeting many of our analysts and stockholders in the weeks and months ahead, on the road and at upcoming investor conferences.

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

All right, thanks so much for taking my questions. Jason, there's some concern out there in terms of the evolution of search with AI mode and AI overviews.

Can you talk about how that impacted you guys in the quarter? It looks like US UVs and sessions are a little bit, quote, "slowed" versus 2Q.

What are you guys seeing there? Then I'd love to talk about cross-selling.

Dealers with one additional product grew 16%. Can you unpack that stat for us and help us understand the drivers of better attach?

Thank you.

Sure. Thanks, Andrew, for the questions.

Maybe I'll take the second one first. As we've expanded our products, particularly outside of marketing and marketplace into software- and data-oriented products, and inventory and conversion and our data pillars across the four pillars that you've heard us talk about, we define multi-product attachment as products outside of our marketplace tiers.

That could include another marketing product, but in increasingly another à la carte marketing product, but increasingly that's referencing software and data products in the other pillars. We have, as you've probably heard in the last couple quarters, we've focused a lot more on engagement of our customers, and engagement can be driven by additional products and them working with additional products.

It can also be more people at the dealership using our products. It's a combination of those things.

We've we've demonstrated that as dealers are more engaged, they perform better on our platform. They perform better in terms of connections and leads per unit and turn times and so forth.

We're very focused on additional products that get attached because, one, it introduces us into new wallets at the customer by introducing us or expanding us into other pillars, and two, it had almost always leads to more engagement, which leads to performance, and performance is probably the single strongest correlated driver to retention and customer satisfaction. We're we're very focused on all of that, but new product attach is one particular data point that is trending well, and so we wanted to share that.

If you think about us over the last, I don't know, four to six quarters, we've had a lot of product introductions: PriceVantage, New Car Exposure, Dealership Mode, Performance Insights, Shopper Signals, VINMAX, StockVantage. Driving cross-product adoption is has been facilitated by the innovation.

On the your first question, I think you were you were asking us how are, and you were it was a little quiet so if you can maybe repeat it, but you were were you asking us how we're handling traffic that might be coming from bots? Was that the question?

Jason, there's a view out there that search is changing, and so if I look at the 3Q US engagement numbers of uniques and then sessions, it looks like it slowed down versus 2Q. Are you guys seeing any changes in terms of SEO or any changes in terms of search sending traffic to CarGurus in the quarter?

Got it. Thanks.

LLM search, which has been around for several quarters now, and we've given some data points on it, tends to be very top of funnel, especially in car shopping. While it may be some very early stages of search, we've shared that we've not seen it really cannibalize the traffic that we typically get.

We've shared in the past that it represented at the time I think 1% of traffic or less than 1% of traffic. It was a small amount, despite the fact that, as we've shared again this quarter, we are the ranked highest on visibility and highest on destination traffic from the LLMs.

Whenever there's been a technology evolution, we have tended to adapt very well and perform very well, and LLMs are no different for us in that. We're showing up well, we're performing well, it's still a small percentage, and we're still getting the lower shopper funnels who want the trust of the experience and the workflow of a multi-month shopping journey.

There has been some softer demand recently, I would say, unrelated to LLMs or search shifts, that is a function of higher gas prices, higher interest rates, and then some of the FTC-related pricing guidelines, which is more of a one-time event, we also think created a little bit of confusion and maybe some softness on the consumer demand side.

Thank you.

Great. Good evening.

Thanks for taking my questions. Welcome aboard, Matt.

Just on the FTC issue, I think last quarter we had said that maybe it disrupted some of your activity, particularly on the dealer addition side, and then this quarter we saw a nice sequential increase in dealers. Was that increase kind of due to that pent-up demand and you were able to kind of close those dealers now or kind of just help us understand kind of like that trend line and how you're thinking about it in terms of the fourth quarter?

Marvin, I'll take it, it's Sam Zales. Thanks for the question.

I think the macro it's really not related to the FTC. We've just had work to do to try to be out front and be the leader in the market on transparency, as you know we are, price transparency.

I think you saw in the first half of the year just the macro environment affecting dealers. We know there's profit pressure on both operating profits and gross profit per unit in market.

Dealers would say to us, as I said before, you're the number one player in the market, you will drive the biggest ROI, I will be there when I want to spend. In the first half of the year, it was a process of dealers saying, "I'm either not going to spend, I'm not going to upgrade, or I'm not going to join the program as I might have previously."

We said in the script that things have started to move a little bit better, I think from an environment perspective. Then you combine that with all of this new product development, really the investments we've made to fuel dealer product development has been a game-changer for dealers in the market.

When you can offer a dealer VINMAX, which is a product right now that is growing VDPs and leads for dealers, helping them sell without discounting and taking a vehicle that's underserved in the market and bringing it forward and boosting it through AI-powered predictive intelligence, you've just got something nobody else offers in the marketplace, and there's a reason that dealers would say, "I'm joining the program because of that." You've heard about our growth in the inventory sector, we've always had Sell My Car, PriceVantage has been a really nice new product to the business, but now you add StockVantage, and it's another reason dealers again are saying, "There's a new set of offerings here that, if combined together, drive my profitability and my reduction in turn times and my efficiency there, "I've got to join the program now.""

So slightly improvement on the macro side, and then the set of offerings that's just much broader in the market has led to that increase, and we're proud of where we are from that front.

Great. Thanks, Sam.

Probably another question for you also, just on international dealer count, a little bit lower than we expected. I think you've been doing some pricing actions there, if memory serves.

Just kind of update us on that, and should we kind of expect a few more quarters of churn there as you modulate your pricing there?

Thanks, Marvin. It's a good question.

The macro environment has probably hit the international markets harder than the US market. All the same pressures: profits down for those dealer retailers, but more importantly, consumer affordability, inflation in those markets, vehicle pricing, interest rates in those markets.

Then you add the Chinese EVs coming into the UK market. It leads to a pretty up and down and crazy markets.

In those cases, the tariffs have not helped at all. With those things happening, that's led to again the same kind of environment where dealers will say, "I'm standing on the sidelines for a while, I'm not going to join as much as I have in the past."

You said it well, in those markets, because we're smaller in our overall absolute size of dealer base, when we hold price and say to a dealer, "We're not going to cave, and we're going to build those annual performance reviews and hold tight on pricing," we're going to lose some of them, and they'll come back over time, but those impact us more than another situation where a small dealer doesn't make that much of an absolute impact on the business. Those are the things that are impacting those numbers.

I'd say what we're really proud of is we're gaining market share, and that's what we're really looking to hope to continue to advance. The Canada dealer growth is beating the competition, we're very, very proud of that.

UK traffic has grown 25% faster year-over-year than our biggest competitor in the marketplace. Our app is having tremendous success in the UK market.

What we look at here is what you just asked in the first question: the growth of the dealer base because we have so many new products we've launched in the US, our playbook is to say let's build those next set of products for Canada and the UK, and then you hope to see that will lead to the same kind of increase as we go forward. I'd characterize it that way, and I hope that gives you a sense of where we're head in the market.

Very helpful. Thank you, Sam.

All right. Good evening, and thanks for taking my questions.

I was just wondering, which of the products that you have launched over the last year or so are seeing the highest attach rate at this point? Then also, which ones are driving CarSight, because I think you'd mentioned about basically the add-on products supporting that growth and CarSight growth.

Yeah. Hey, John, it's Jason.

I'll reflect on some things that we have shared publicly in the past. PriceVantage and New Car Exposure happened to launch around the same time at Q4 of '25, and I think shortly thereafter, we estimated that they would be an eight-figure together they'd be eight figures in '26 this year, and we then reiterated a couple quarters after that that we were on pace for that.

That gives you a sense for those two products in particular. We've also then shared that VINMAX has been one of our fastest growing, fastest launched products, and so that's faster than either of New Car Exposure or PriceVantage, so that tends to put that in some perspective for you.

The other thing I'd mention to, and so those are all à la carte products, those directly help with CarSight. I'd also, though, I don't want to give short shrift to products that are not à la carte, and so Shopper Signals is one good example where, again, we mentioned in the script huge adoption of Shopper Signals.

Those who adopt it, and as a reminder, it's giving a full profile on the consumer, it helps the dealership sales team convert those leads much better, they sell more cars with it, it's a highly effective product, and that helps with a lot of things. That helps with retention, that helps with performance, that helps give dealers confidence that our data and our products are unmatched and are effective, and so that leads to sort of indirect upsells and retention.

When you talk about something that's going to add dollars to CarSight, there's the à la carte path to adding dollars, and then there's the create customer goodwill, create retention, create a sort of environment for us to sell other products as well.

That's helpful.

Just thought of that, as an example that I should have mentioned, including upselling to premium tiers of our listings product.

Appreciate the extra color. Then just my next question, I was wondering if you might be able to provide any early thoughts on '27, both in terms of strategic objectives and then also to the extent you can, any initial thoughts that might help us in modeling.

We haven't shared commentary on '27 yet. I think if you from a strategic lens perspective, we've got the three levers that we talk about in '26, and those tend, they shift from '25 to '26, they shift a little bit, but those are pretty central levers to what we believe is going to grow a long-term enterprise value creation over time.

I think those are foundational to what we're trying to do. What the environment calls for could change those a little bit, but you're not going to see us up and wildly pivot off of creating deeper value with dealers and creating a longer and stickier journey with consumers.

Thanks for the help.

Great. Hi, thank you for taking the question.

This is Ryan Powell on for Naved. We wanted to ask on Guru with usage up 24% quarter-over-quarter.

What share of US leads now originate from Guru or other conversational experiences, and then also how conversion compares with other channels? Thank you.

Hey, Ryan, it's Jason again. We have given some stats a few quarters ago in terms of I believe percentage of our users that were using, at the time it may have been called Discover, but Discover became Guru.

Then we've talked about some growth since then. We also, though, have innovated, built, created new AI applications that are not what the old Discover was, and we are always testing and evolving those, whether it's conversational AI or whether it's agents, whether it's in Dealership Mode.

It's getting increasingly hard to kind of define what does involve AI and what doesn't, because so many surfaces have AI now. We've shared I think this quarter that Guru unique usage reached new highs, it was up 24% sequentially, as you referenced.

We've also shared, so that's in terms of usage, it's growing, AI usage among consumers is growing rapidly because adoption's growing and we're introducing new use cases for it. In terms of its sort of efficacy, we have talked about Discover users, again now called Guru, were much more engaged on the site, they were submitting a growing number of exchanges, conversational exchanges, and they also converted better.

I don't know if we've given a stat on the improved conversion rate, but we've said that they convert better. Then the last thing I would say is you've heard us talk a lot about Shopper Signals, that is, as I just mentioned in the last question, a really effective product for dealers to help them improve their conversion rate.

Think about a Shopper Signal profile of a consumer that has simply used drop-down filters to shop for a car versus one that has exchanged with Guru, and there's such richer information on the user if they've interacted with it. The usage not only helps with conversion rate and engagement, also helps with the dealer products to help them with conversion.

Great, thank you.

Thank you. I appreciate you're not sort of talking about '27 yet, but if I think about this year, you originally sort of talked about some investment that was going to weigh on margins, and now it sounds like you're being a little bit more efficient on spend.

I'm just wondering how you're sort of thinking about the balance of those two into '27 as we sort of think about the cost side of the equation.

The balance of us gaining efficiency with—

Of like continuing to invest, but but of of wanting to invest, but doing so more efficiently. Like is it still is it should it be a net drag or or maybe not as bad as as, or not as much of a drag as you sort of thought earlier on?

We again haven't commented or guided on '27, and so I'm not going to talk about whether we think it's a bigger or smaller drag or anything that quantifies it. In '26, in some ways, I really do think that the product velocity and product innovation velocity speaks for itself.

In so much as saying, while we were more efficient, we sort of clearly didn't do that at the expense of not building new products. Having one to two pretty significant product introductions per quarter is a pace that we're really proud of and is faster than we've been in the past.

The fact that we were able to do that more efficiently doesn't mean that we held back on it, because we wouldn't be able to keep introducing products if we did. As far as how we think that continues, we expect to continue investing in '27 across our dealer pillars, and that is building new software products, that's building new data products, and in some cases, that's learning and executing on how to sell them a little differently.

It's also helping educate dealers and helping them with their behaviors and processes to get the most out of them. Shopper Signals is a good example of that, dealer performance partners is a good example of that.

We're going to continue investing across pillars because that doubles our TAM, it grows engagement, it grows stickiness, it does all these wonderful things. We're also going to continue investing on the consumer journey.

We just talked about some of the AI applications of that, but that is something that we are really focused on, especially as the search and agent and LLM landscape is so dynamic around us. We're committed to keeping product velocity high.

We also are committed to continuing to always be disciplined with our investments, run a good business, operate a good business. Margin, free cash flow, those things are important to us.

Product innovation and long-term growth is even more important to us, and we're going to continue, I hope, to prove to everyone that we can do both really well.

Thanks for that. Then this is a second question, and this may be somewhat naive, but I would still appreciate your insights here.

If I think at a very high level of your business, right, and sort of the marketplace is still the core, and I understand you're trying to sort of add all these other services and features to sort of help the dealer. If you think about marketplace, your importance and pricing and power there is really sort of tied to inventory size and, I guess, leads expected to deliver.

When I understand that LLM is small today, but if someone uses ChatGPT or Muse or Claude or whatever, and they find the listing, and it does source you, but they somehow click through, is there still some behind the scenes, I don't know, metadata or something where that's attributed to you? Or is there a risk that the dealers you're providing a listing, but then your customers don't actually know that you helped them get that customer to the dealership?

Hopefully that made sense.

It does. I understand it.

Sorry if this is sounding rudimentary, if a consumer learns about a listing on our site that gets sourced from us, and then they click through to us and submit a lead, obviously we're going to get credit for that. If they see that listing, and they see the sources from us in your example, and then they decide to go to that dealer website, then we would not get credit from the dealer.

We might get credit through some third-party attribution sites that might be able to connect the reference to us in the search, but that would be pretty indirect, and we would get a little brand exposure to that consumer, so that would help. Same thing with an agent.

If an agent were to work on behalf of a consumer and come to our site and submit a lead on behalf of that consumer on our site, then CarGurus will get credit for that lead. In terms of how consumers are going to think about that, consumers can find a listing on our site today and then go to the dealer site and submit a lead on the dealer site.

We know that that happens a fair amount. A lot of times they'll click out from our site to get there, other times they could open up a new browser.

We also know that a lot of times they go through our site because it's easier, and we provide a lot more connectivity to dealers, we give them a lot more context and so forth. Yes, an LLM can access data and represent it.

We also have an app in most of these LLM search platforms now, and so the consumer might go to that as well. They might see a reference to it and not use us.

What we're finding, and this was my reference earlier to higher funnel, is that a consumer is very unlikely to say, "I'm looking for this type of car, the LLM has returned two of those cars. I'm good.

I don't need to do any more research. I'll trust the LLM, and I'll just connect to the dealer on those two."

They just don't do that, because it tends to be a multi-week process. They want to do research, they want to compare, they appreciate our pricing intelligence and veracity, and that we have a Trust and Safety team and that we're showing all-in price, and that we can help them think about the totality of the deal.

The totality of the deal then starts to get into trade-in and financing, and a lot more than just which car and at what price. Again, just bringing it full circle back to your example, it's not that different, in some respects, from a consumer that searches on Google, and traditional Google returns a car listing in their VLAs, vehicle listing ads.

That often times is just the starting point for a consumer.

Appreciate it. Thank you.

Yeah.

Good evening, and thanks for taking my questions. Just starting with one quick clarification, could you maybe quantify the true-up benefit associated with the real estate tax reassessment in the quarter?

Thanks, and have a quick follow-up.

Sure. It was $1.3 million.

Understood. Appreciate that.

Then just as a follow-up, you mentioned Guru's transition to a full-service agent. Wondering if you could expand on what that would look like, what capabilities would it have, and how far it could extend into the transaction itself.

With Carvana's growth and activity across other marketplaces pointing to greater consumer acceptance of online vehicle purchases, how does that evolution fit into CarGurus' broader strategy for end-to-end purchasing, and is there scope to accelerate investments in those capabilities? Thank you.

Sure. Rajat, happy to speak to that.

The natural evolution of Guru toward more of an agent is focusing early on where I would say the pain point is for the consumer, and that pain point is in a lot of the research, the back-and-forth, the information validation with the dealer, and it's not about completing the transaction. 97%, 98% of consumers still want to do that in the dealership.

Our more vertically specific, automotive-specific experience with Guru is much more full-service that brings together much of what our site brings together today: inventory, pricing intelligence, and then connectivity with the dealer in order to get them to a spot where they can narrow down what visits they may decide to do, and then shorten the visits because they're equipped with more information when they get there. Dealership Mode helps with that, too, when they're on the lot.

I would say, it's not totally binary in terms of yes, it's transaction, or no, it's not, but I would say in the early stages, it's much more about helping them navigate the search and decision process, and then helping them in the dealership. The test drive and the finalization of the documents and the finalization perhaps of financing is still done much more human supported.

That's great color. Thanks for taking my questions and good luck.

Terrific. Thank you very much.

As always, we'd like to thank everyone who joined us on the call this evening. We certainly appreciate the support of our shareholders and our customers, and most importantly, we want to thank our employees and our Gurus for their passion and hard work.

We're extremely proud of the progress that we continue to make and the innovation that has helped expand our company so much recently. We're excited to share that with all of you.

Thanks very much, and have a great evening.