Yelp Inc. (YELP) Earnings Call Transcript & Summary

Thank you. Thank you. All right, we can kick it off. Well... Thanks to the team from Yelp for being here. I think this has become an annual tradition, so we appreciate the partnership.

Jeremy Stoppelman, Chief Executive Officer, David Schwarzbach, Chief Financial Officer. David, I think you have a safe harbor to read quickly. Thanks, Alex, 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. Jeremy, I want to start with you, sort of a big picture table setting question. The platform has been under a lot of evolution over the last couple of years. You guys have really leaned into what you call the product-led growth strategy. Maybe frame for us some of the key product developments that you are most excited about and how you see them driving growth going forward. Sure.

Yes, if you go back the early history of Yelp, we were really go-to-market led for a number of years. Then we entered into probably around 2018 or so, a product-led era, where we made a great bit of progress on the business, taking Yelp's ads business to greater heights, something on the order of $1.5 billion-ish dollars a year. And then now we're at the point that really it's the AI era. Obviously, a huge technological shift. We've had to rapidly adapt that technology both internally in terms of how we work, but then especially modernize the product, bring AI elements into it like Yelp Assistant, which we've launched in the past year. And then we also saw an opportunity, we didn't just want to necessarily focus on how can we transform only Yelp and what we've done historically, we also started looking for greenfield opportunities where we can ride the AI wave. And so that's led to some really exciting AI tools, tool capabilities that we now have starting Yelp Host, which was homegrown. So that's answering the phones for restaurants. And it's not just picking up the phone and giving directions, but it's integration with front of house, like our guest manager product, like OpenTable. It's also even taking food orders over the phone. And so that's commissionless, it's going straight into their POS systems. They don't have to pay the DoorDash tax necessarily, which businesses love. So the feedback we've gotten on this that product has been really strong. It even speaks, I think it's 17, 18 languages now. Really impressive the capabilities that the team has built out. And then we also started working on a product for services, a voice product for services called Yelp Receptionist. But as we were bringing that product to market, we also had the opportunity to partner and bring in House Hatch, which was years ahead in the digital lead management space. Great brand, great praise coming from their customers. Basically, leads come in to services businesses from all sources, so it's not just Yelp. Make sure to follow up on those leads, drive those leads to a booked appointment all through AI. And we took our voice effort and merged it with their voice effort. And so now we have a voice product as well as part of their suite. And that's been really exciting. And it's farther along the Host business, which was really a zero to one kind of endeavor. Hatch has been in the market for a number of years and has more substantial value.

That's great. I definitely want to dig into all those products in more detail. But maybe to take a step back and start off, just as you view the sort of, you know, competitive landscape and changes to the broader search and internet behavior. There's been a lot of updates to the search distribution landscape with AI. We've seen a number of announcements recently around agentic commerce and consumer agents and things like that. How do you view Yelp's positioning against that broader landscape and the value of the platform, both your installed user base and distribution footprint, but also the data and content that you could bring as well.

I mean, I think something that's become increasingly clear is that Yelp has a very unique data asset. I put it right up there with Reddit as another example, or maybe Wikipedia. But if you want to answer local questions, if you want to know who's trustworthy, if you want to know what to order, all of those things are housed and grounded by Yelp data. And if you're not Google, you don't have access to Google's data, you probably want to be talking to Yelp. And I think that's been borne out in, you know, the quality of conversations our business development team has been having and some of the deals that we've now announced, probably the most noteworthy being our work with OpenAI. And you're now seeing Yelp content integrated within ChatGPT experience as people dive into local. Just recently also there was a quote button, a request a quote button that taps into our request a quote system, showing up on relevant results within ChatGPT. And so that's part of our philosophy isn't just to have great content and hoard it with inside the walls of Yelp. Certainly we want lots of people coming to Yelp, but we've also always found that there's opportunity to bring our content to where the consumer may be, and that helps bring new people back to the platform. You know, it's often, the deals often take the shape of, you know, there's obviously licensing revenue associated, there's typically traffic elements as well as well, and you can find our content on a whole variety of different properties, not just ChatGPT, whether it's Apple or Bing or on the AI side with Meta, you can find our content there as well as Alexa. The list goes on and on, even in, you know, most autos have our data as well. So there's a lot of demand, I think, for trustworthy, high quality content about local businesses, and Yelp, I think, is the critical resource.

I want to dig into some of the products that you mentioned at the top, especially Yelp Assistant. I think that's the furthest along and probably most prominent across multiple of your services, RRO and services as well. What have been, you know, some of the biggest learnings from rolling out and developing that product, some of the use cases, and how that's trended over time?

Yelp Assistant is essentially a chat interface to Yelp. So taps into all of our great content. You can literally find a needle in a haystack now because it's searching across everything that we've got, which is phenomenal. I think the biggest challenge is at the end of the day, you're also competing against things like, frankly, ChatGPT, Claude, but even some of the new buzzy products. I know it's been going around at this conference about Instinct, what about Instinct? You know, it's a really exciting space and that the bar is constantly being raised in terms of like, how much can you push the capabilities? And so that's been the challenge for our team is like, as soon as we get something, we're like, wow, this is great. Probably the next week we're seeing something where it's like, well, we have to do even better. We have to add even more functionality. So it's a very dynamic, fast-moving space. But again, we have the trusted content. We have the critical resource to answer those questions, which means both we can continue to create a compelling experience right on Yelp, but we're also happy to partner and work with other players, be it Instinct or anyone else, and find ways that are creating value for both companies.

That's great. David, this might be a chance to bring you into the conversation around some of the investments that you guys are making to facilitate that AI development and growth. How do you measure sort of the return on some of these investments against the broader sort of monetization opportunity over time?

So obviously we're applying AI to the externally facing products that we're selling, whether it's for advertisers or the Yelp Assistant. So we care a lot about building those efficiently and using tokens in an efficient way. Internally, obviously there's been tremendous advance in the coding capabilities. So obviously on the engineering side, we care a lot about that. And then we're also looking at how to apply AI across the rest of the business, whether it's sales and marketing or GNA. And the way that we think about it, of course, is first to look at some of the core metrics that we measure productivity around. I think everybody's still wrestling with that. You don't want lines of code because it can be just inflated by writing more lines. Code is it really better to have more PRs, maybe, maybe not? You want the quality, but I think the most tangible is looking at avoided cost for the core and existing product. And that could be like doing a migration more effectively and so saving time engineering time would be an example of that and that's a that's pretty measurable, or you get a lot of velocity with new products. So when you look at something like Yelp Host, we were able to build that very quickly. And then we decided we could really increase the opportunity size if we added ordering over the phone, we were able to build that very quickly, get it in market and start generating revenue. There, the ROI is very clear. You have the team, that's the expense, and then you've got the clean effort, and then you have the revenue and the expected revenue in the future. So that feels pretty straightforward. And then there are just some things that have become possible that just weren't possible before. Why it's just self-evident. So on the GNA side, a recent example for us was, hey, we want to make sure that people have all the right permissions for a particular piece of the code, you know, you just send a bot through and you look at all those endpoints and you're able to identify where maybe all the permissions are not correct and then you can correct it. Like what's the value of that of an avoided problem? But it's cleared and that was a few hundred dollars to do. So we're looking at it in a variety of ways and triangulating around it, but what I think we can confidently say is that we are seeing significant ROI from our application of these capabilities within the way that we operate, yeah.

Is there an opportunity for you to adopt more sort of open source, open weight technologies as well? I would imagine what's valuable and a lot of companies have talked about Pinterest for example as well earlier today bringing in these open weight open source models and then applying your differentiated first party data and content to hone some of those models is that sort of an opportunity that you guys are undergoing as well.

Yes, I think fine tuning you can apply across all these models so that we're doing it's not limited to obviously open source. We're going to do that with other models as well. So I think I would really distinguish those. And I think that we are very effective at applying this great data set that we have to this question of, hey, how do we find local businesses or really make local advertising efficient? That's different than, hey, from a coding perspective, should we be using these open weight models? How can we apply them? When does the trade-off make sense? And I would just say broadly, there's a real focus on how effective is a model. It's not just whether the model scores better on a benchmark. It's do you have a lot of rework with that model? Does it actually code effectively when you give the prompt? And do you introduce more bugs that you have to fix later? So I don't think it's actually a slam dunk that you're just going to shift to the open source models. And at the same time, they obviously offer a significant cost advantage. So you've got to balance them, and we're certainly looking at them just like everybody else.

Shifting gears a little bit to the core business and especially the RRO side of things. You talked about some persistent macro headwinds, you know, mostly outside of your control. How would you frame the current advertising environment, especially for those core advertisers, and what's the outlook to reinvigorate that going forward?

Yeah, I would say on the restaurant, retail, and other side, it has been a tough period. Unfortunately, I don't see macro suddenly getting better. There's a lot of countervailing forces, inflation being one of them, the conflict in the Middle East, all the input costs for restaurant operators going up, labor costs, et cetera. And then on the consumer side, that shows up also in lower frequency of dining out. So it is difficult, but our focus is on continuing to keep that connection to the consumer, continuing to create the best product that we can, surprise and delight through things like Yelp Assistant, and how can we keep raising the bar on ourselves, on our team, on our product to try and make sure that, you know, eventually, like things are cyclical, eventually I think restaurants will thrive again. And we want to make sure that we have the most compelling product and have a great share of the opportunity as that arrives. So you'll see us continue to make significant investments, both in things like Yelp Assistant, but then also the core experience of how do we gather interesting types of content? What are new content types that are compelling to people? We have question and answer features that have started showing up on the site that people seem to be really engaged with. We have obviously lots of photos, video. All those things needs to get better and better. And at the end of the day, you're competing for people's attention. And so the product has to stay compelling, and you have to keep raising the bar on yourself, or you're losing portions of that attention to others as well.

And how is that sort of direct traffic engagement trended over time as well? I mean, that seems to be the North Star for you guys. And then once the macro advertising environment gets better more broadly, you guys have that direct traffic to be able to serve. How is that trended over time?

In the last quarter, we did highlight some positivity on the consumer traffic side. We attributed it to a number of different sources, certain partners sending us more traffic and downloads. We saw some positivity out of the SEO side of the business. And so I think that just shows that we continue to be really relevant. Our team continues to work hard to optimize the distribution opportunities that exist. And then I also would look to the future of like, hey, there are these emerging opportunities, things like ChatGPT, we're right there working with them to create a great experience on ChatGPT. But also, there's opportunities to bring people into the Yelp ecosystem. Things like the request a quote button that opens up an MCP widget that we are powering and all that information on the user comes into the Yelp ecosystem. We're then able to bring them quotes to solve their need. And they've also then been introduced or reconnected with Yelp. And so that's a bit of a template of how additional distribution opportunities can emerge through AI players. I think that's just the tip of the iceberg. We're very early days. And I think the market share of these agentic search players is still relatively modest compared to the classic Google market share, which remains at something like 85%, for better or for worse.

Shifting gears a little bit to the services side of things as well, you guys have done product development work and improvements to that product, both external consumer facing with request to quote and integration of Yelp for business and things like that. And then on the back end as well, on your matching engine and onboarding new service pros, maybe just frame for us what inning you're in or some of the key developments of that services business right now and then what the opportunity is going forward.

Yes, yes, request a quote has been great for us, driven a lot of growth over the years, it's digital leads. Are you know passed on to local businesses. I think the most elegant thing about request a quote especially in the AI era is its monetization that's built right in. So to the extent we're putting something like Yelp Assistant front and center, might ask, well, OK, great, you have a chat interface, but how do you even have ads in a chat interface that are compelling? How do you make that work? I think there's experimentation going on, but it's very unclear what ad format might replicate what Google has done, for instance, within Search. But for Yelp Assistant, and especially because we derive the majority of our revenue through search, services, when you're talking to the Yelp Assistant, you might be talking about restaurants or where to go this weekend or what have you, but when you come to a services need, you're then in a conversational request-to-quote flow. Essentially, you're creating a project, ultimately gathers the information necessary from the conversation and says, okay, I'm going to send this out to get multiple quotes. That moment moment is a very elegant monetization moment built right into a chat interface. So I think that's great for Yelp. It also could provide a template for others that want to work with us and tap into the Yelp business network and be able to deliver these quote requests to their users. So, you know, I think it's a very elegant solve to a portion of the how do you monetize chat problem, and it just happened to be kind of naturally built into what we were doing.

David, maybe some of the revenue growth drivers of services going forward other than, you know, request a quote and improving that, you know, over time. You talked about multi-location advertisers and service pros and really onboarding some of these large national service pros. You know, you talked about, you know, improving the matching engine as well. Maybe frame for us what you see as the key building blocks and growth driver for services specifically on the revenue side going forward.

In services, it's all about delivering the lead at the lowest possible cost and the highest quality. So we look at a lot of ways in order to ensure that. To the extent that we can gather more information, that's first and foremost. That's why Yelp Assistant is such an advantage, whether you're going through a request to quote in it or it's asking you other questions to discern exactly what it is that you're seeking or what the exact nature of the request is or the problem that you're trying to solve. So first it's gather more information, then it's continue to refine that matching algorithm. And then, interestingly, there have been some really useful applications of LLMs to the matching itself. For instance, synonyms matter a lot, and finding more synonyms and more relevant synonyms end up improving matching. And that's just, in essence, the beginning, because once you've gone through the matching, for instance, we bought Hatch, you want to work the lead more effectively. And then, of course, you want to be able to do the job and charge for it. So to the extent that we are able to support those advertisers in closing at a higher higher rate and getting more valuable jobs that are able, obviously, to charge more cost per click. So it's not one thing. You have to do everything on their behalf. And in general, I'd just say we've invested heavily in that communication platform, Yelp Message Center, and over the past several years to improve just how do I have that conversation? Can I interact with you over text? And I think there's a big opportunity now with voice, which is, can I respond to you? Can I reach out to you after you've started a conversation with me? So we're going to work across all of those dimensions. Broadly for Yelp, we've also seen some mixed shift as we're able to bring more categories in that deliver value at higher ASP, the CPCs for those are also higher. And we're always doing experimentation around all of that. So I think there's a variety of ways in which we can deliver more value to service pros, and then that leads to better monetization or higher CPCs for us.

And you've talked a little bit about in the past experimenting with paid search and SEM to bring in those leads and making sure you have the effective monetization potential of those leads in place before you make those investments. How is that sort of progressing?

Exactly, we've invested heavily in order to bring leads in through paid search and land them to the right advertiser.