Good morning. Welcome back to the conference. Rob Coolbrith from the Evercore ISI Internet research team. Very pleased to have Yelp management with us here today. CEO Jeremy Stoppelman, CFO David Schwarzbach. Thank you much for being here. I think to start, Dave, we'd like to read the Safe Harbor and then we'll get underway.
Thanks, Rob, for having us at the conference. We'll be making some forward-looking statements during the conversation today that are subject to risks and uncertainties. Please refer to our SEC filings for more information on the risk factors that may affect our results.
Great. Thanks for that, David. Jeremy, just to start off with, maybe for those who haven't caught up on the story in a while, it has changed a little bit.
A little bit. Yeah.
Yeah. Take us through how Yelp has been evolving over the last couple of years, the AI transformation now underway and some of the products on the roadmap that you're particularly excited about.
Sure. You can kind of divide Yelp's history into a bunch of different epochs. Starting kind of pre-pandemic, we're really go-to-market led. Sales drove a lot of the revenue growth that you saw show up. Pandemic happened, obviously, that was a temporary hit to our business, but we really leaned into product-led growth, and especially within the services side of the business. That took us all the way through into the recent era of AI and the AI transformation that's affecting companies all around us as well as obviously, Yelp has a lot of opportunities, a lot of uncertainty. We're really leaning into the opportunity that AI provides. On the consumer side, we really need to think about transformation of the consumer experience. If you're in ChatGPT or you're in Claude, consumers are starting to have a different idea of what they expect from applications like Yelp.
We want to stay on top of that, and have been building Yelp Assistant now for over a year and just launched. First, it started in the services category. We focused our chatbot on services and have driven significant project growth through that. Now we've recently expanded that across all categories, which is really exciting. We're seeing strong feedback for consumers. On the anecdotal side, it's early. As we continue to ramp it up and expose it in all the different places, I'm sure we'll have more stats to share, but I think it's a really exciting moment for Yelp of reinvention. In addition to that, there's other greenfield opportunities that we identified and started building towards. Particularly, you have on the restaurant side, Yelp Host, that we've built. That's about answering phone calls using AI for restaurants.
Making reservations, now moving on to food ordering. The reception that we've gotten from business owners has been fantastic. Go-to-market there is looking quite strong. We're really pleased with the momentum there. It shows that we can innovate in an entirely different area. Obviously, we have a lot of in-house expertise built up leveraging AI technology. We're putting that to work there. More recently, we've also acquired Hatch. We had been building, when we started the effort on the restaurant side with Yelp Host, we also had Yelp Receptionist that got us into the services side of answering phones for small businesses. The other piece of the opportunity when you start answering phones, we were thinking about it in terms of, okay, well, we have a great presence at restaurants. We also have this big services thing, which is a huge TAM.
As we were building Receptionist, the opportunity came up to purchase the market leader in this space, Hatch, and it's been going really well. In the quarter, we reported 92% year-over-year growth. That's fantastic. Our focus is to keep that momentum going as well as augment their team with strong engineering on the voice side, as well as bringing our go-to-market and operations expertise. At this point, we've probably about doubled the team size. Being very thoughtful in trying to make sure that they can maintain their startup spirit. Obviously, it's really valuable features and functionality, but by having a small team, they've been really nimble, agile. That's helped them get ahead in this market. It's becoming a pretty competitive market. There's other players out there.
We want to make sure that they can continue to execute and capture a significant amount of the opportunity. It's a very exciting TAM. That's kind of a quick tour around the history as well as the AI transformation that's upon us.
Okay, great. Yeah, David, the flip side of the transformation narrative, I guess, and what you can do on the revenue side is arguably the investment intensity required to drive that transformation. FY 2026 EBITDA guide implies, what was it? Two, three points of margin compression. Maybe you could take us through the drivers of that across Hatch. Yeah, some of the internal transformation work. The incremental investments may be in SEM this year, which I think you have a new take on, right? Whether thinking of that as sort of a single-year reset or something that maybe extends out into 2027 as well.
We have definitely approached 2026 as a year of investment, given the significant opportunity immediately in front of us, as Jeremy was talking about. Part of that, of course, is operating expense, which is we have an opportunity to do more for consumers and for business owners. We want to lean in around that. That's the paid search side. We had leaned in in 2024, but didn't see the ROI we wanted and learned a lot. We put a lot of effort around product in order to really get that incremental lead to someone who would value having that incremental lead and would have budget to buy that lead.
We spent 2025 doing that, and as we've come into 2026, we've built that whole mechanism which enables us to go from paid search, create that project, and then direct it to a mid-market or enterprise customer who will value it. That is the whole investment that we've had to make there, and then closing the loop on ROAS for those advertisers so that they know they're getting the return that they expect from a lead. We see an opportunity there. We're still going to be very disciplined and look to deliver the ROI we expect from that paid search. As we've demonstrated, I think many times, that financial discipline is if we can invest more, we will, and if we are not getting the returns that we like, then we will invest less. That's the paid search piece.
Just historically, as you know, we spent very, very little on consumer acquisition. Just as an aside, that's part of the power of Yelp, which is this fantastic brand, trusted, high authority that attracts a tremendous amount of organic traffic to the site. That's the traffic side. At the same time, we deployed $270 million of capital in the first quarter, plus $30 million of retention that'll be paid out over two to three years to acquire Hatch. We think it's a fantastic asset. We've leaned in in terms of bringing both engineers and go-to-market folks over to Hatch in order to really seize on the leadership that they've already demonstrated to drive market share. We've been super pleased with that performance. As Jeremy said, what we saw was run rate revenue in March. Annual run rate revenue was $34 million, growing at 92%.
We really like that. We see a lot of opportunity not only to deliver product to those customers, and just as a reminder, this is an AI lead management tool that really enables folks to turn a click into a lead, but we see an opportunity extended to even more categories. There's a lot of runway there. There's a lot of product opportunity, and so we want to invest. This is a moment. You don't want to back off after you make an acquisition and underfund an asset when you see a tremendous market opportunity. We're leaning in around that as well. That has translated into guidance for 2026 that has EBITDA lower compared to 2025.
I would just say over the medium to long term, we see these assets as delivering, whether it's Yelp Host and that AI voice product extending into food ordering or on the Hatch side, it's not only increase in the products they offer to their existing categories, but expanding to other categories. This transformation that's happening for all businesses around how you work, how you build, how you code, how you bring product to market, even around the marketing techniques, using voice for customer success, using AI to coach sales reps to be more productive, coaching managers on how to coach sales reps. There's so many applications here, and that's why we said on the earnings call that over the intermediate to long term, we really do see a strong driver of margin expansion.
It's this combination of you need to invest today, you've got to learn how to apply these tools, you got to change how you work, and at the same time, I think we're all experiencing it. We have very high expectations for the productivity gains that we're going to get out of AI.
Got it. Going to come back and dig into other revenue. By the way, I think you need to change the name of other revenue.
We've gotten that feedback.
It's a very attractive name.
Exactly. Other.
I want to spend some time on RR&O. You've seen sort of accelerating declines in the business over the past few quarters. Lots of cost pressure right now on local businesses. Are you seeing this as a cyclical issue or more of a structural reset, or these businesses just sort of have less money available to spend on marketing for the foreseeable future?
Yeah. It's been a tough time out there to be, particularly in the restaurant space. That said, we continue to invest, particularly on the consumer experience, because restaurant traffic is good for Yelp. One way to think about it is it's a high-frequency thing that a lot of people do enjoy doing, and it's a reason to turn to Yelp. To the extent we could stay relevant and attractive for consumers as they're thinking about whether it's where to go out or even what to order through our DoorDash partnership. It's great marketing for us. It's great activation for those consumers. Eventually, that person who's regularly searching for restaurants or retail or other, is going to have a service need.
When they do, we want to be there, with Yelp Assistant, for instance, which they're already talking to for the restaurant needs. It's very natural to say, "Oh, my sink's broken. I think I need a plumber." We go straight into that conversation. A Request a Quote is created, goes out to a few pros, suddenly, that consumer is both delighted because they're getting a pro that's going to help them solve a problem, but there's also great monetization. With Yelp Assistant, we have seen growing projects over time and a significant amount of projects are now flowing through that experience. We continue to see that number rise. I think that's a great thing. Another nice thing about how we are approaching AI is we do have built-in monetization.
Even if we're monetizing in general less with the struggle that's happening with restaurants, we have built-in very strong monetization on the services side.
Got it. Do you find that just generally, your continued presence, your authority in restaurants, right, conveys authority across other categories? That helps you maybe even in paid project acquisition. I don't know. Maybe if you could complete the thought there for me.
We're obviously a critical source of human written content, trusted content, about local businesses, and a lot of that is in restaurants, and restaurants are a high frequency. It's really important, I think, to have great content in that category. Stepping back for a second and looking at this AI transformation that's going on, about a year ago, we pointed to the opportunity in data licensing to a lot of these AI search players. We knew that we had a critical resource for anyone that wanted to deliver a great search experience. If you're not Google, you essentially are going to want our content. Fast-forward to today, we've now signed deals with many of the major AI players. We continue to have high-quality conversations about new opportunities. I think that also points to just the quality and the caliber of the content that Yelp has.
All of these mega players are turning to Yelp when they need to be able to deliver a trusted answer about local businesses. I think that's very validating. Also points to the strength that we have in these higher frequency categories as well.
Got it. I want to turn to services for a moment. Obviously, it's gotten tougher as well. You were in the sort of low mid double digits, entering 2025, you decelerated down to the low singles. Maybe just to sort of unpack some of the drivers there, supply and demand dynamics is just a little more lighter. Demand for projects from the consumer side is an issue of budgets available to fulfill on the supply side. Maybe you can just sort of take us through the drivers.
Yeah. I think a big thing that we saw was the seasonal pattern got disrupted once again. I think last year it was tariffs.
We saw that show up in the data very clearly, then this year was conflict in the Middle East. I think threw consumers off course a bit, as they're contemplating some of these bigger purchases or remodel projects or any of the work that they were doing. So we did see slower growth there, but at the same time, we continue to lean in and invest. We have a lot to do on the services side in terms of product and engineering, improving Yelp Assistant, improving the business owner account. Lots to leverage in terms of AI with respect to what can we put in front of business owners to make their life easier, to help them create ads, to improve targeting. So that's really where we're focusing our attention.
Got it. On Assistant, I think maybe one of the benefits that's a little less transparent to investors is, you could probably put more ads in there. Maybe you could talk us through that, some of the broader benefits of Assistant to the business in terms of user engagement, in terms of ad delivery, in terms of advertiser satisfaction and what they're seeing in terms of conversion.
Yeah. There's still opportunity to drive increased monetization. Lots of the leads that are flowing through Yelp continue to flow not through Yelp Assistant, not through Request a Quote. That might be a person showing up, browsing around and calling a single business. To the extent we can convert more people into leveraging this AI experience that we're investing in and we're so excited about, that comes with better monetization. There is a real opportunity there, I think, for Yelp.
Okay. We're finally going to get back to other revenue. We still need to get to it.
Your favorite category
That's where a lot of the transformation initiatives, I think, are centered. You set a goal now to more than double your run rate there from, I think it was about $115 million in Q1 to $250 million by the end of 2028. Hatch obviously going to play a big role. Jeremy, can you talk about the shift in focus there on the core Yelp advertising opportunity? David, maybe you could take us through the three key drivers, I think, which are your Hatch, data licensing and your revenue share on the food delivery side.
I think in a transformation, you don't just want to focus on one thing. You want to focus on the big opportunities. First and foremost, we've got Yelp. We've got to transform that experience. We've got our big bet on Yelp Assistant. We're seeing some positive signs there. As we've got so much internal talent around AI, obviously the thought was, "Well, what could we build that's greenfield, that's not 22 years old, that has a big TAM that we can attack that's readily adjacent?" That's how we landed on Yelp Host, and Yelp Receptionist, which sort of morphed into an acquisition and combining forces with Hatch. I think both represent really exciting opportunities.
For restaurants, obviously it's a lot of work to pick up the phone all the time, and if people are on the phones, that means they're not doing other stuff within the restaurant. There's significant costs there. There's also all of the after-hours, and this applies both to services and restaurants. Like, you're not picking up your phone, it's going to voicemail, chances are that's business that's lost. What if that person after hours can make a reservation? Suddenly, you've converted that into better monetization, better revenue for your restaurant. There's a real opportunity there. There was clearly a need, and we felt like we had the talent necessary to execute on it. I'm really impressed by what the team has built. You can go and check out Yelp Host, you know, sample the voice, and see what we built. We think it's really industry-leading.
The team was just at NRA, not National Rifle Association, but the restaurant conference. There was a very strong reception there. It really feels like, you know, Yelp is pushing the envelope in the restaurant sector with voice. Voice is having its moment in that area. On the services side, you know, we've developed a lot of the capabilities. We had, I think a great voice experience, but we had the opportunity to partner and ultimately acquire the market leader that had been focused more on the text messaging side, which is also really important in the services sector. A lot of people want to communicate that way, as well as direct messaging. This is as a reminder across all platforms. This isn't specific to Yelp.
It's, you know, managing leads for services businesses regardless of where they came. It's, you know, those leads are really valuable. Oftentimes the business has paid significant money to acquire those leads, managing them better, more efficiently, turning more of those into bookings is super important to those businesses. You know, Hatch was there like from 20 I think it was 2023. You know, as soon as the LLMs became capable enough, they had attached it to the system that they had built and really saw the product take off. They had an early lead in that space, voice now is becoming sort of the other new thing that is, you know, part of what services businesses really expect and want. You know, Hatch has a great voice product.
We've augmented that with our voice talent as well. I think again, you know, given the execution chops of Yelp, we have in front of us now a very high growth opportunity in a very large TAM. We're going for it. It's not about, you know, a massive shift in strategy of, hey, we're going over here and, you know, not paying attention over there. It's that a lot is changing rapidly. To have an opportunity to continue to build Yelp, to continue to make it better, to drive value to consumers and transform the experience, while also pursuing these totally new and exciting high-growth opportunities, you know, leveraging capabilities that we already had, it seems like a no-brainer. We're very excited about the possibilities.
Granted, it's going to take time, as we pointed to, for a three-year, you know, timeline for how can we get this revenue to be substantial enough to really move the needle on the overall company. I think that's, you know, a bit of a high-class problem in that Yelp has significant revenues, and it will take time for these kind of new startup opportunities to fully bear fruit. I think looking at the traction that we've got, looking at the quality of the technology that we've had out in the marketplace, talking to customers, you know, any investor can figure out that like, "Hey, Yelp is onto something here.
The aha moment for me was listening to the voice, and it's shockingly good. You know, you've got the hint of Northern California uptalk, a little bit of dismissiveness from the Yelp Receptionist. You know, obviously, you also have distribution in spades. When you think about right to win, you know, to your point, there is competition in the market. I think we're probably seeing more come into the market today, arguably, if I look at my screen. You know, can you talk a little bit about right to win and how you're seeing competition play out, and how do you think about that?
Competition in which space?
In voice.
In voice. Right. Yeah. I mean, I think there's a bunch of startups, primarily, and I guess we focus on the services side, you know, a number of startups, you know, one of which recently raised a billion-dollar valuation. I think if anything, that's validation for where we've invested. You know, we got there, we made the acquisition. We had, you know, very quickly, have doubled the team size more or less. That's because of our conviction of the opportunity, you know, we can see that it's going to be a race for share. Ultimately, I think, you know, product quality and go-to-market is going to be the thing that, you know, causes someone to really be the, you know, the big winner in the space.
I think we come in extremely well prepared. In a way, you know, if I think about when I was pursuing Yelp, you know, as a young founder, many years ago, it's like we did have growth, and we did the best that we could, but there was so much that we didn't know. I think to be able to partner with a startup that is already there, that is already leading the market, and bring in our operational expertise right at the moment where they're scaling up, everything, you know, it's really working, and so they just need to execute on the go-to-market side, like health, so to speak. You know, that's Yelp's bread and butter. Not to mention we can bring You know, we're not building go-to-market from scratch.
We have all of these touch points with a lot of the same clients, deep relationships over many years, many of them advertisers. You know, there's a cross-pollination, a cross connection that really helps us on the go-to-market side. On the product side, you know, a lot of in-house senior expertise on AI that we were able to bring in and immediately start, you know, focusing on: How do we keep this voice cutting edge? How do we get latency down? You know, our priorities are focused on a product that blows everyone out of the water and a go-to-market that nobody can match. I think, you know, we're coming in, we're coming in pretty strong to this competitive opportunity. I look forward to showing the results.
Got it. You know, looking ahead, you know, anything you can share on any sort of, you know, key product milestones that you want to hit or, you know, key go-to-market milestones that you're thinking about? You know, anything you can offer there?
Yeah, I think it's really focusing on, given that this is AI tooling for these businesses, we're in constant conversation with the big clients and trying to understand what are their needs, what are their expectations, how can we get them even more confident in voice. One of the challenges is a lot of businesses, especially the big ones, don't want to hand over the entire call center to an automated system at this point, but they want to use it after hours. That is really about understanding, well, what is it going to take to unlock that opportunity? That's significant revenue expansion with those clients if you can give them the absolute confidence that your AI agent is better than the person that they have on the phones.
That's one example, is giving them visibility into the stats of here's AI agents that are answering the phones, here's how they're performing, here's the appointments that you're getting from that, and then also, here's the human side. Here's your sales reps that are on the phones. Here's how they're doing. There's also coaching opportunities. As you're getting all that data from the entire sales force that's on the phone, understanding what the humans are saying and how they're performing and which ones are saying certain things that perform better. We're gathering all this feedback from our clients all the time, and we are iterating on very fast cycles. This is not annual planning. This is more like every three or four-week kind of planning, because that's the pace that this industry is moving at. It's really exciting.
It brings me back to the old days of Yelp.
Yep. David, anything involving Gen AI, we have to ask about the gross margin profile, right? Anything you can tell us about that and where you think you need to be from an ARR perspective for this to become a sort of EBITDA accretive , if you will?
Yeah. Maybe just to step back for one second on other revenue and then come back around.
Yeah
to the gross margin and the cost of delivering these products that are AI products. For folks who aren't familiar, other revenue is composed of three components. One is transaction revenue. That's our DoorDash partnership that grew 88% in the first quarter. There is licensing. That's doing very well. You can now find Yelp licensed content, as Jeremy was talking about, on meta.ai, on Bing, Yahoo, Apple Maps, Alexa, other places. We feel like one of the things that has been true is that if you want to do local search, you partner with Yelp. The third component is the subscription revenue that we've been talking about, and we mentioned, of course, that Hatch had grown 92% in the first quarter. Overall, other revenue grew 75% in the first quarter.
Obviously, a strong start to the year, and we're really pleased overall with the opportunities there. Everything that Jeremy was talking about how do you deliver value at the end of the day to businesses, and so they want to spend more with you and they want to adopt more products. In terms of the gross margin for these products, we actually haven't seen a significant expense yet, and there's a couple of reasons for that on the product side. The first and foremost is that we have a very specific problem that we're solving. This isn't just broadly ask anything of an LLM and have it be able to respond to every type of query that you might get. It's very focused.
The benefit of that is we're able to stratify tokens, use the intelligent tokens for the hard part of the interaction, use less expensive tokens for the more routine parts. That's an element of it. The other thing is the way that we've architected the system is we can move between models quite easily, so that enables us to approach it with cost management in mind. The third is that we have a really, really good engineering organization, and they're able to architect these products in a way that's quite efficient. We all know that LLMs have a lot of error to them, so you have the risk that you're going to inflate costs because of rework or because you have to do a lot of pre-processing of the response in order to ensure that it's accurate.
I would just underscore something that we haven't yet talked about, is our data set enables us to fine-tune these models so that they are providing high-quality, highly accurate responses at the get-go. This isn't just true, for instance, in the voice product that we've built for Host, but I would just say for Yelp Assistant itself, one of the things that's really nice about Yelp Assistant today, and if you haven't tried it, I encourage you to do that, is it actually tells you why it gave you that response. That is very differentiated from other search experiences. I'd say that's the first thing. The second thing is it provides the evidence right there.
Here are the quotes that support the reason why I gave you that, and if you want to learn more, you can read more reviews, you can look at photos, so on and so forth. This idea that the Yelp data set enables us to be more efficient with tokens, I think is probably something that is not obvious to folks, but that's something that we've been able to leverage on the cost side. That's delivering the products. I think like everybody, as our team is adopting these tools, there's a lot of inflation in token use, and there's probably a lot of inefficient token use and a lot of rework that comes from results that may not be as accurate as you need them to be. I think everybody is going to go through this curve, which is, hey, let's get adoption up.
Let's help folks to learn how these tools can make them more productive. Then we're all going to be very focused on cost management around employee token consumption. You can see that from a variety of articles that are coming out now. Again, I think what's going to end up happening is you're just going to see very strict stratification of tokens for different use cases. You're going to see management of token use across.
You've taken down the token leaderboard.
We never did-
We never had that.
We never did the token leaderboard. Just this idea, use as many tokens as you can is just the wrong incentive.
The AB guys are sick of this. This is like the fifth time I've made the same joke in the last couple hours.
It's a first for us. Yeah.
There you go. I want to switch over to AI licensing. Yelp content now in a lot of places, including ChatGPT. You've highlighted the licensing era in the past, and just given the growth in other revenue, I'd imagine it's going to just continue to scale pretty meaningfully. Just given the trends that you're seeing, how impactful do you think that opportunity can be for the business as a whole? When you're striking partnerships with people like OpenAI, can you take us through some of what you're aiming to achieve, both in terms of direct monetization as well as referrals?
Sure. Yeah, we're really pleased with the traction that we've seen on the data licensing side, especially with respect to the AI players. Obviously, there's the revenue there, but as you alluded to, there's other ancillary value as well. For example, this is fairly recent. If you go to ChatGPT and pull up any business at this point, I think, you're going to see business data and information that's being pulled from Yelp data. With that, there's a Yelp logo, there's the rating, the star rating from Yelp, and you can click through on that, and end up back on Yelp. I think this is just the beginning.
If you think about what consumers expect, I think the way that these search services deliver information isn't quite ideal, in the sense of when they first started out, it was all about providing answers, not providing really their sources. Once they provided the sources, they were buried as little footnotes. I don't think that's actually the consumer-optimal approach. Now, the consumers understand that the information isn't always perfectly accurate, or maybe they want to do a little bit of digging and confirmatory diligence themselves. The interface is begging for change, which is just link to stuff sometimes. Yeah, just put in the hyperlink to the source material in a more prominent way. I think we're starting to see flashes of that in the various LLM providers and their consumer experiences. I think there's going to be a move towards that.
It would take the balance in general between these LLMs and the content providers to a healthier place. That's kind of a prediction. I think we could very well see that happen. In general, when we've done these deals, this isn't our first time licensing to search providers. Obviously, we have a long history, and an illustrious list of partners. In those relationships, we typically do get anywhere from healthy to very healthy traffic back to the content. It's very early. With the AI players, a lot of their experiences really haven't been built out yet. I don't know exactly the magnitude, but I do think it is a real interesting opportunity for Yelp to gain some incremental distribution as well.
Obviously, the revenue is significant and has grown rapidly as these players have adopted and needed Yelp content, there's the second-order element of like, and what will the distribution be? The other piece, too, is just the branding and to have all these prominent companies putting Yelp logo and sort of content right in front of consumers, I think also has a very positive effect that shouldn't be ignored as well. From a strategic standpoint, I think the fact that this is happening, again, cements the importance of Yelp. If you want to provide a great local search experience, local discovery experience, you really need Yelp data. Like, we are the key provider if you are not Google, and I think that's showing up both in the deals we've already signed, the deals we probably will sign.
I think that is very exciting and maybe an underappreciated element of the Yelp story.
Got it. Is that helping you? I mean, we perceive a little bit of a generational divide where for people my age, people love us on, there's only one answer for that, right?
Maybe for younger people, it's a little more fragmented. The distribution by LLM, is that starting to help us or maybe bridge a generational divide if there is one?
I think it certainly could introduce younger folks more quickly to Yelp. It's just very early, so I can't really say for sure. Obviously, younger people are using ChatGPT, using Claude, using all of these new search experiences. To the extent that Yelp is showing up very prominently, I think that bodes very well for the overall health of Yelp and cementing it into as like a critical piece of what's needed if you're building a search experience.
Okay. I want to spend a moment on that third component of other revenue, the food delivery economics. You added DoorDash as your sort of your key partner there, and you've already, as to your point, David, seen strong acceleration. I think in the past you had sort of an approach of checkout within the four walls of Yelp. Now you're passing people cleanly over to DoorDash. I can use my DashPass. I love it. Do you plan to get your sense of what inning we're in in terms of driving awareness of that shift within the Yelp user base?
People know that they can now make this seamless. Okay, that's the wrong word to use, transition to DoorDash.
I would say it's still actually very early in terms of awareness on Yelp. This is part of the reason why I think we're excited about the Yelp Assistant conversational experience is there's just it's so easy to do in-product marketing using a conversation. "Hey, I see you're ordering. Did you know you could use your DashPass?" Like that's just such a natural, easy complement to the query in the first place. I think we're super excited about that. Just going back to Yelp Host and answering phone calls for restaurants, there's a double-up benefit there. One of the things that we've seen is this rapid adoption, as Jeremy was talking about, and we shared on the earnings call that in April, we had an annual run rate just over $1.5 million.
Actually, just in May, that went up to over 2 million. Just from April to May, it increased by 33%. That adoption does a variety of things for us that's important. We are getting into the order-taking business ourself of getting those orders over the phone for restaurants. That makes us more intelligent about what consumers are searching for. It means that we can improve our in-product experience on the Yelp site, even as we cement the relationship with those restaurants and enable them to get food delivery through DoorDash. It actually fits all together very nicely.
Okay. Got it. David, we're finally going to dig into the numbers a little bit more. First I'll check and see if there are any questions in the room. Please.
I was going to ask about the numbers. I wanted to quickly on the data licensing piece, it just seems like there should be a lot of benefit. I'm curious how you're valuing the benefit that you're getting from the branding on OpenAI and the traffic that you're getting. It doesn't seem like you're getting that much money for what should be a really, really valuable data asset. It just doesn't seem like it's that many dollars.
I'd say it's important to view the licensing from a few perspectives. One is this is very, very, very high margin. A dollar of revenue is almost a dollar of EBITDA and cash flow. The leverage that we get from these licensing agreements, just from an EBITDA perspective, is very, very high. That's the first piece to bear in mind. The second, and I really wouldn't dismiss this, is we're balancing across several competing benefits from working with these very, very large platforms. One is, of course, the licensing dollars, but this ubiquity, having the Yelp logo everywhere is quite valuable to us. Ideally seeing a lot of traffic coming back to us. It's almost the ideal situation. We're getting paid for market awareness and for traffic, not the other way around. I wouldn't view it solely as the dollars.
The other thing is one of the things we have been absolutely unwilling to do is to license our data for training. We think that this is obviously Yelp's strategic competitive advantage. As part of those conversations, there's some back and forth about, "Hey, if you're not going to allow me to train, I'm not going to pay for that." It is this balance across a variety of competing interests as we work with them. At the same time, we still see a lot of opportunity ahead for licensing.
Got it. What, if anything, would change your stance on, other than a whole boatload of money, your stance on allowing access for training?
A whole boatload of money.
That works.
Yeah.
Speaking of which, David, you've flipped from a traditionally unlevered position in the business.
Yeah
to now being in a net debt position. Do you think of that as a structural change in the capital structure that you're sort of comfortable with maintaining? Or is this just a question of, hey, obviously we had an opportunity to Hatch. We have an opportunity with our share price.
Yeah.
We're taking advantage of those, and we're going to work our way back up into net cash position.
We did deploy a considerable amount of capital in the first quarter, $270 for Hatch, but we also repurchased $125 million in shares in the first quarter. Diluted shares outstanding from Q1 of 2025 to Q1 of 2026 declined by 12%. It's pretty substantial. We also shared publicly that we would be repurchasing at a slower pace as we move through 2026. It is different now for Yelp. We've run very large cash balances historically. We're also a very, very strong free cash flowing business. We generated $324 million of free cash flow in 2025. That's over a 20% cash flow yield, I believe, and that's exceptional. The way that we're thinking about it is fundamentally, we have this opportunity immediately ahead of us, which is to really expand our presence around AI tools. We've done one acquisition in that space with Hatch.
We want to execute against that to earn the right to do more M&A. If there's an opportunity for us to do that, then we don't want to be constrained by capital in order to pursue it. At the same time, we're going to remain very, very disciplined in what we're willing to pay for an asset. That's the first piece. The second piece, of course, is our longstanding commitment to return excess capital to shareholders. That's not going to change. We're going to share, we've gotten a lot of feedback and questions from investors over the past few weeks since the earnings call, and we're going to share more about our capital allocation strategy on the Q2 call.
Got it. David, maybe you can just walk us through the SBC reduction plan. Obviously, you're giving us a cleaner EBITDA year by year. Talk us through how that's playing out across changing compensation mix?
Yeah
versus aggregate comp and how it's a competitive market out there, right?
Yeah.
There's some big companies in San Francisco. How that's playing out among the engineering organization that's doing a great job for you.
These changes we made several years ago now.
Yeah.
What we shared at the time was that things had gone very well. Things have continued to go very well. I think you have to step back and recognize, first of all, Yelp's, we believe, a great place to work. We just did our annual people experience survey. Results were extremely strong. We were pleased with that. We haven't seen undue attrition in product engineering. We haven't seen undue attrition in sales and marketing or in finance. We're happy with the culture and environment that we've created. We think that we are very clear in the mission, serving local communities and economies. We think that we've created a great experience and place for people to work on interesting problems. Those are all very important to us and to our employees.
So far it has gone really well, and that's enabled us to reduce stock-based compensation as a percentage of revenue to less than 8%. At the beginning of last year, I believe it was 8% in the first quarter, and we committed to reducing it to less than 6% by 2027. We're on track. We don't need to do more structural change in compensation in order to achieve this next goal. It's going to be an outcome of the transition that we already made. First and foremost, we want to make sure that Yelp is a great place to work.
Well, I think I'll open it one more time. We've got about 80 seconds left. Any last questions or thoughts from the audience? Seeing none, that's a great place to stop. Thanks so much.
Thanks.
Appreciate it.