Thank you, everyone, for attending the second day of the conference. We have David here from Yelp, who is the CFO of the company. David, I think you had a Safe Harbor?
Yep. Thanks 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. I think just to kick things off, Yelp reported two key results last week. It'd be helpful if you could just give us an overview of those results and any updates to your outlook for the rest of the year.
Sure. One of the things that we had said on our Q1 call was because of the policy uncertainty that emerged in the April time frame, we didn't see the typical seasonal ramp on advertising budget that we normally see. That was true both on our restaurant, retail, and other side, as well as our services side. That did persist as we moved through Q2 and into Q3 here. The dynamic out there is small businesses are operating under this set of conditions where input costs do continue to rise. They're also faced with uncertainty on the consumer. Because we're very much tied to the physical local economy in terms of our performance, obviously we connect consumers with great local businesses and services are done in your home, restaurants are where you go. That whole dynamic is something that's been playing out for us.
That informs, of course, the guidance that we gave for Q3 and for the full year.
Yeah, that makes sense. I wanted to spend some time. You outlined three strategic priorities for the year, the first one being leading in services. That's really been driving growth for you guys for the last couple of years. I was hoping you can maybe unpack what's been driving the relative strength in that segment of your business.
For advertisers, I think it always comes back to delivering value for them. They want a good lead.
Your second strategic priority is delivering advertising value.
Exactly. In terms of the services business, if you're not delivering that value, then for a service pro, they're obviously not going to advertise with you. The challenge, of course, is to ensure that you're doing everything you need in order to get the information that's required to make that a highly qualified lead. When we say lead in services, it's obviously with focus on being able to deliver that value. The consumer also has to find value. That's where our Request -a- Quote comes in. We'll talk about Yelp Assistant. This transformation that's taking place in the way that people access information is something that we've really embraced. We think that we're very well positioned around that. We'll spend more time on that in a moment.
We've introduced all these types of tools to elicit from that consumer the relevant information that enables us to match them with the right service pro. You really have to think of services not as an ad, but as a workflow. That workflow is the consumer has a need. They share some information. You find the right service pro for them. They still need to communicate. The service pro needs to provide a quote of some sort to them, usually through a visit. They need to come and do that work. The consumer has to be satisfied, and they need to pay. It's not just a one-step thing. You've got to build out everything that enables not just the matching to take place, but for the transaction to take place. We continue to invest in that.
Just as an ad, larger service pros, they do have a whole flow around converting leads. We've built out a lot of tooling for that as well.
You mentioned Yelp Assistant there. I think that's a really exciting product for the company. Maybe for those not familiar, can you just give us a little background on Yelp Assistant and the value it provides to both the consumer and the advertiser?
Sure. I think we're all obviously now using tools like ChatGPT and becoming used to having a conversation around whatever it is that we're seeking to understand or do. Broadly speaking, our perspective on search more generally is simply we're moving from links to answers and actions. That chatbot enables a conversation. We think that because we have such a depth of experience in this workflow, plus, of course, our review content, we're able to really build a chatbot for services that is very effective. What do we mean by effective? In the fewest number of questions, we get the information that's needed to most accurately match the person with the service pro. We've rolled that out for services. It's not in every entry point. It's only for logged-in users today. There's more surface that we can expand that to.
We're going to roll it out across all categories on Yelp. There's still a long way to go. Just as a stat, the increase in projects that have been generated through Yelp Assistant grew 400% year on year. We think that it's working. We think it's the way people want to interact largely with information going forward. We think we're very well positioned.
Yeah, that's terrific growth. I think it's an acceleration from the first quarter as well. It's good to see that. I guess, can you talk about maybe the monetization opportunity more longer term with Yelp Assistant? It seems to me if there's better matching between consumers and advertisers, that's a more valuable lead for the advertisers. Over time, that's a pretty compelling opportunity for Yelp.
Absolutely. One thing just for context, when you come to Yelp and let's say you type in you're looking for a plumber or a restaurant and you click on one sponsored link, that's one click for us. When you come to Yelp and choose to go through their Request a Quote flow and we provide you with four possible service pros to interact with, for us, that's four clicks. The monetization rate in Request a Quote is 4x a search and click on a single result. We obviously like that because that's much higher monetization, but it's not perceived as higher ad load by the consumer. What are you doing? You're eliciting the information. You're figuring out the right pro. You're bringing them together. You're delivering value the whole way.
The beauty of Yelp Assistant is that it's very similar to Request a Quote in that, in services, it leads into the Request a Quote flow. More broadly, it enables you to deliver value while also increasing monetization on the advertiser side. You can do that in a way that actually delivers value to both parties. You're delivering more value to the consumer because you're getting a better match, and you're actually delivering more value to the service pro because you're able to more accurately match them with people who want to use the service you provide. Yelp Assistant broadly, I think, is going to enable us to increase the monetization rate. Probably even more importantly, or just as importantly, is once you're in a conversation with a consumer, you can ask them more questions. Hey, I saw that you had been interested in an electrician. Did you hire them?
Hey, we found the right service pro. We found the right plumber for you. Have you considered gutter cleaning? Can you tell us about your experience on Yelp? Is there any way that we can improve this experience for you? It just changes the whole way that people interact. I think this will be broadly true, not just true on Yelp. It provides an opportunity for us to really engage the consumer and then obviously in product marketing.
Yeah, that makes a lot of sense. The learnings kind of compound over time, and you get to personalize the service for each consumer based on their history as you learn more.
Exactly. I think just make it by, you know, we say personalize. I think it's make it more relevant. It's making it, hey, we have that history. We understand something about you. Let's deliver more value to you.
Yeah, that makes sense. Another strategic initiative within services that you guys are pushing for is pushing more into the enterprise, larger services businesses. Can you talk about some of the initiatives you're working on right now to unlock some of those ad budgets?
Yeah, so multi-location, our definition being locations with businesses that have a multitude of locations. Multi-location did perform quite well there. Over the past year, we've really been building out the types of capabilities they need in order to manage leads. One of them is lead management using an API. Another is conversion API, which everybody's moving to in order to improve attribution. As part of that, there's also integrating with CRMs because they do use CRMs. Maybe a sole proprietor, a single truck service pro isn't using a CRM, but most folks who have multiple locations are using some form of CRM. We added Zapier as an integration. We've seen really good uptick with that. You start to build out this ability to gain information, connect information, enable service pros to be effective in finding the consumer, having that consumer be persuaded to pick them.
You're just adding value throughout the whole system through these integrations and improved experiences.
It sounds like it's pretty early days there, and there's a long runway.
Yeah, it takes time. Anytime you're working with enterprise businesses or folks who are using agencies, it just takes time. They have a workflow. You need to be able to make those connections. One of the things that's really pretty powerful about large language models is, let's say in a CRM, you want certain fields filled out. That's not part of our typical set of questions that we've built out or that Yelp Assistant typically asks. One of the beauties of large language models is that they can easily look at, hey, this is what needs to be filled out. Let me go ask that dynamically. Let me return in the format that's needed by that CRM that information. There's this middleware layer that large language models enable that I think is just starting to be appreciated broadly by companies.
It really enables this much smoother ability to move information around across various systems. That's actually a really powerful and interesting middleware capability that's emerging.
Yeah. We'll definitely talk more about AI and LLMs a bit more. Before that, just wrapping up on services, you did recently make an acquisition of a company called RepairPal, pushing more into auto services. Can you maybe just explain the logic behind that acquisition and how the integration has gone thus far versus your expectations?
Yeah, absolutely. In this idea that we want to be the leader in services, we have a lot of categories. Home services is certainly our largest category. Auto and local services are number two and three, basically. We saw this opportunity to add to the auto category or vertical for us. It is very, very much lined up, RepairPal is, with what we do. It's finding a highly qualified local business and matching the consumer with that business so that both derive value from that connection. The team is great. They provide a quote. We have Request a Quote. They provide a quote. Their network of service providers have high ratings on Yelp. That was nice and fit together. It just made perfect sense for us as an opportunity to go deeper.
I think the overarching point with this is that we are looking at ways to build Yelp using balance sheet and using inorganic to further increase the financial performance of the business. We are looking for additional opportunities there. We don't have anything to announce. The other part that was important about RepairPal was building that muscle, finding an acquisition that you want to do, negotiating at a price that both parties feel is fair, working through all the paperwork, and then actually onboarding that team and integrating the product. We have been very pleased with the acquisition itself. I think underlying that is just this capability for us to do M&A and deploy balance sheet on behalf of the financial performance, I think, is equally important.
Yeah, for sure. That's great to hear. The last strategic initiative you laid out for this year was transforming the consumer experience. I guess, can you maybe give us a couple of examples of what you've done on the consumer side, either website or app?
Yeah, so I spent a lot of time talking about Yelp Assistant. That's clearly part of this transformation. Also, just our home feed. We've really, really changed and improved and modernized over the past several years. We've made it much more visual. It's enabled us to add ad units there. There's a nice example in the letter that you could look at. Broadly speaking, what is important for any app? It's to be relevant. It's to provide something to a consumer that is meaningful to them. It could be engagement on social. That's not our primary mission. We're a search function. Can you provide folks with relevant and engaging content, posts that also lead to them being able to do something that's necessary and important to them? It could be going out on a first date. It could be repairing something at home. It could be a home remodel.
It could be something else. That whole idea that by providing information in an engaging way that's relevant to consumers, that's the intent. I think just as we all see these apps evolve and consumer expectations continue to rise, if you're not constantly improving what you're doing, then you're absolutely going to fall behind. There's just going to be this interaction between content and conversation that everybody is still figuring out, I think.
Yeah, that makes sense. I think we spent a lot of the conversation so far focused on services. I would like to maybe shift to RR&O quickly. It's been an area where I think macro and competition on the margin has been affecting that business.
Definitely.
Can you just delve into what you're seeing in that segment specifically? What gives you confidence that that business, that segment can stabilize and turn around?
Yeah, so restaurant, retail, and other does span quite a few categories from obviously the retail side to shopping to beauty and health. There's quite a few things, and each of those categories has different attributes to them. Obviously, they're a bit related, but there's something unique as well at the same time. I think Yelp, we remain relevant. We're very much known for restaurants. Content is fantastic. We think that our content really is the most reliable that's available. We've got great photos, we've got menus, and so on. Just that core thing, what is Yelp? Fundamentally, at the end of the day, we're a publisher. We think of it as human-generated content now. I think we've all talked about it as user-generated content. I think there's now this distinction that's very important that we're really embracing. Human-generated content is part of our purpose, obviously.
We're taking that human-generated content and we're really making it available in a relevant way. Part of our strategy is to make that content available beyond the four walls of Yelp. For instance, as I'm sure you're already aware, we're on Apple Maps, we're in Perplexity, we're on Amazon Alexa. Part of it is really ensuring that we have reach, and when people interact with that information, it's branded. That branding turns out to be extremely important because it conveys authority. Yelp is perceived to have very high authority. When you ask, like, will Yelp be able to participate as macro rebounds in restaurant, retail, and other, my answer is yes, because our content is relevant, has authority, and provides needed information to consumers. I've been really harping on this word value, but delivers value to them. We feel confident. We're continuing to modernize the experience.
We're using chatbots to make that even more engaging, and we feel good about that. At the same time, I think as you've seen a lot of the folks in the restaurant segment report, plus some of the data coming out on consumer right now, these restaurants are under tremendous pressure. Consumers are pulling back. Input costs are going up. It's a very tough environment broadly for a lot of these categories.
Yeah. Let's dive into the off-platform monetization a little bit more. Is part of that strategy diversifying where you monetize? Are you seeing any impacts? A lot of publishers have talked about changes in Google search algorithm or AI overviews, adoption of LLMs as impacting their traffic. Is that part of the strategy to expand the monetization, not just on Yelp, but outside of Yelp as well?
Absolutely. We see a large opportunity to monetize off Yelp. That's absolutely true. Again, because our content has authority, if you want to build a search engine and you are not working with Google, Yelp has, we believe, an extremely large directory of businesses. You need that. Of course, as I've already talked about, we have the content and the authority. One of the stats that we shared in the shareholder letter was that API calls for AI search providers had increased tenfold in the past two months. ARR that we're generating there through licensing, which is on a per API basis, had increased substantially to over a $10 million run rate over the past several months. That's clearly emerging. We see this great opportunity both to monetize through licensing, but also it puts the Yelp brand out there.
People do still click through because there's a lot more content on Yelp than can be surfaced off of Yelp. It's a big part of the strategy for sure.
Yeah, that makes sense. I think the AI conversation so far is centered around products and monetization. Maybe you could talk about how you're using Yelp internally to improve efficiency or product development within the company.
Yeah, just one product that we haven't talked about, which we're really excited about, is the AI voice product that we're coming out with. There's receptionist. AI receptionist is going to be for services businesses. AI host or voice will be for restaurants. It's really good. Again, in the letter, we have a link. You can listen to a sample of this. You're going to find that it's very engaging. It's fault or interrupt tolerant. It's very speedy. It enables, it can diagnose what the problem is on the service pro side. On the host side, it can handle very dynamic questions. That is super exciting. The reason I bring it up in this context isn't just to obviously mention it as a business-facing opportunity for us. It'll be priced as a SaaS product. Each of those components are building blocks for doing other things within Yelp.
For instance, customer service. Can you use a voice AI in order to engage on maybe simpler questions and queries that an advertiser might have? That's one example. Underneath all of that, though, is the matching technology, the diagnostics, the connections that you have to build between databases in order to support those products. When we think about the Yelp side of the equation, the enterprise applications that come with AI, they overlap because you're building capabilities that are relevant both externally and internally. We are doing a lot of work. We have dozens of projects running. I'm excited what's going to be possible on the post-sale side. There's ways to help sales reps to be more effective by having LLMs look at the transcript of the calls where a consumer has agreed to that. That helps managers to be more effective in coaching them.
Obviously, we all are very, very aware on the developer side, the application of these tools. You've heard a lot of stats around that. I personally still think it's early. The other thing that's very important that I don't think is talked about very often on the developer side is devs only write 5,000 to 10,000 lines of code a year. A huge proportion of their time goes into planning, coordination, debugging, QA, et cetera, et cetera. It's not enough to simply say that we have enabled a dev to be more productive by having an LLM write some of the code. That is only a portion of their time. I think this is really missed in the dialogue around dev productivity. It's how to use large language models (LLMs) and improvements in process and change how people work or devs work in order to increase throughput.
One of the stats that Craig Saldanha, our Chief Product Officer, shared recently was that we've actually been able to accelerate the time that we can bring a feature to market by we reduced the time by 60%. That wasn't just because we added LLM and magic happened. It's because we changed the process and had clarity on the priorities. We feel very good about the way that we're leveraging them. I think people have become a little bit distracted with the idea that the LLM is just going to write the code. Here's my feature. I can vibe code enterprise software. I can vibe code an at-scale consumer app. This, I think, is a little bit misguided. Vibe coding can help with a feature or a routine task. There's still all the planning and coordination that's required in order to deliver quality code. We're applying it.
On the G&A side, I would just say super early. There's a lot of conversations around how LLMs and emerging enterprise software products can really help to increase efficiency on the G&A side very early in my experience so far.
That's great. Sounds like there's a lot of exciting opportunities ahead.
Yeah, it's going to be, I mean, for sure, over the next five years, it's going to be remarkable. Again, it's just going to be this marriage between technology and process and training people up. It's the classic people, process, and systems. It's not just add technology and that's sufficient.
Yeah. It looks like we have about a minute left. I was hoping maybe you ran through a lot. Just maybe highlight one or two things you want investors to leave this conversation with.
I would really underscore, and hopefully it comes through from what I've shared with you. We're really excited about the potential of AI. We're applying it everywhere in the business, both consumer-facing, under the hood for those consumer-facing products, and across our enterprise. It's a very exciting moment for us. I just think the other point is we're really financially disciplined in the way that we approach this. We've committed to holding headcount flat this year. We obviously are on a $250 million share repurchase run rate. We think about ROI in everything that we do. When we're investing for the long term, we're not just thinking about how can we apply this technology to make Yelp more successful. We're also thinking about how do we do that efficiently and make sure that we're allocating resources to the best opportunity.
You've got this combination of an emergent technology where we think we're incredibly well positioned, plus financial discipline around how we develop those technologies to make Yelp both successful and relevant with consumers and advertisers, as well as financially successful.
Great. We'll leave it there. Thanks, David.
Thanks so much.