Yelp Inc. (YELP) Earnings Call Transcript & Summary

Good afternoon, everyone. My name is Trevor Young. I'm one of the Internet analysts here at Barclays. I'm pleased to be hosting the Yelp team, David Schwarzbach, CFO. Thanks for making it to the conference. Before I get started, I think you have some comments to make.

Yes. Thanks, Trevor for having us. Here 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.

Okay. Great. So to start off, we're in a pretty challenging macro environment overall, and yet you've managed to grow ad revenue double digits now for 10 consecutive quarters, really strong results there, on pace to finish the year in double-digit territory as well. Much of that growth is coming from the strength in revenue per paying advertising location. So what's driving that persistent growth despite that challenging macro?

There are 2 parts that I think are really important why we've been able to perform. The first certainly is that we have an audience that people -- that advertisers want to reach. So our audience -- more than 50% of visitors come from households with income over $100,000. 77% of folks have college or advanced degrees. Many are homeowners, and they are ready to purchase or they're high intent or transact by the time they get to Yelp. And they also tell us exactly what they want to do, they're -- type-in a query. So that's a great audience. There's clarity on what they're interested in. For advertisers, that means that we're able to match the right advertisers with those consumers. Also, it's cost per click. So they're only paying if the person clicks and it's very down funnel so there's demonstrable ROI. So that has worked really well for us. And through the third quarter, we did see robust advertiser demand as folks have concentrated on their advertising dollars against clear opportunity. So it's clearly been working well for us.

No. And it seems like that's likely to persist despite that macro environment?

So we've been -- one of the things that we do get is a question is if the macro environment deteriorates, what's the impact on SMB and consumer broadly? From our perspective, the folks who are advertising, they're certainly muted, they're not going to be rated. They do advertise. But for folks who have already been on the platform, when you go on Yelp, you're still going to see that star rating in an ad. So they can't shed their rating. And the folks who are advertising are typically 4- and 5-star businesses. So first of all, that advertising base is stronger than the average SMB. And then on the consumer side, what we think is folks have seen significant increase in home equity. And so they are doing projects. And they are also college educated advanced degrees and typically, unemployment will be lower among those folks. So that's also a positive. But I think probably, most importantly, it's just if you're spending money, you want to spend it well, everybody has lived through inflation, and so things are more expensive. So you're even more motivated to make sure that when you're doing that home remodeling project or putting up the fence or going out to eat, that you're going to get what you expect.

Makes a lot of sense, particularly on the consumer side. And you alluded to my next question a little bit, just on the advertising location trends, like the Q and P times Q framework. They had been generally up year-on-year in aggregate, while you've been driving that double-digit ad revenue growth, but more recently, last couple of quarters, trends that there have slowed a little bit, been a little bit choppy. And I think you kind of alluded to it before, where maybe you're getting -- churning some of those lower spending, less engaged paying advertising locations. Just talk us through that dynamic a little bit more and your comfort around letting these lower spend locations actually churn off.

Yes, yes. And we commented a little bit in the shareholder letter and on the Q3 call. But paying advertising locations, we have seen folks who are just not spending a lot of money with us, but might have a lot of locations, choosing not to spend anything that's paying advertising location, right? They have to be paying. So they fall out of the number but in general, the team is really focused on the larger opportunity. So I just think in terms of efficiency from enterprise sales perspective, you want them focused where there's the best opportunity. So overall, we're not at all dismayed about the level of paying advertising locations and we think over the past several years, just the quality has gone up, and that's reflected as actually we started out, that revenue per paying advertising location has consistently gone up. And we think fundamentally, that's because we're delivering more value because advertisers are smart, whether you're -- you have a single track or a big national, if you're not getting a return on your ad dollars, you're not going to spend. And so we think that, that's also in a way a reflection of quality, right? Because they can spend more, they do see return. We're delivering more value to them. They're increasing that spend with us, and we think that all is to [ the good. ]

Yes. Higher quality merchants to have on advertising.

And if I can just add, between the choice of spending to acquire the next paying advertising location or acquiring the next dollar from an existing paying advertising location. It's obviously much better from a leverage perspective to get that incremental dollar from an existing customer than a new customer.

Yes, yes, makes a lot of sense. Sticking with the PALs a little bit. RR&O, that's been trending a little bit worse than on the service side. Is there anything specific to like key verticals there? Like is it restaurants, aspects of retail as they're seeing their own consumer pullback? What's driving kind of that differentiation between services and RR&O?

So in RR&O, typically, you have the larger national chains, whether that's retail or restaurants. And so those are the folks that we are seeing decide not to spend anything. They weren't spending much. Now they're not spending. And that actually tends to be across the various subcategories within Restaurant, Retail & Other. Services is just not concentrated. It's not -- if you think about national chains, there just aren't that [ many ]. And so it tends to be your small- and medium-sized businesses there. And we've continued to do really well in being able to acquire them. We'll get to it, but one of the leverage points for us has been self-service, and that's been something that's for our SMB channel. And that has been very much around service process. That's been a positive and half our acquisition in the third quarter came through that self-serve channel.

Got it. Okay. So just tying the ad revenue piece altogether, how are you thinking about balancing growing wallet share versus adding additional PALs over time, you mentioned it's far more attractive to just get that incremental dollar, but it's also -- just to be clear, it's not that you don't want PALs to grow, you do, but it's kind of a balance of getting high quality, higher lifetime value. Help us understand that dynamic as we think about '24.

Yes, we'd absolutely -- obviously, we want to grow paying advertising locations. At the same time, we want to do it as efficiently as possible. We're incredibly ROI-disciplined. At Yelp, we measure to the greatest degree that we can, just what's the -- just like you would expect folks advertising on Yelp are looking at their marginal return on an invested dollar. We're looking at a marginal return on invested dollar, whether it's marketing spend or on the sales side or product and engineering. And so we have a lot of really good projects in flight that we believe enable us to continue to deliver more value and therefore, generate more revenue per paying advertising locations. So we're just -- we're spending more time there. But, we think that as we improve the overall quality of the experience, we differentiate between services, Restaurants, Retail & Other. We deliver value to those advertisers. That's a great pitch for the next advertiser. And also really informs us about, hey, if you're a service pro, let's say, you're a plumber, what are the things that are most relevant to you for a customer? And conversely, when we are doing really well, let's say, among plumbers, that's all feeding back into the matching technology. So we're vertically integrated. We have our own ad tech stack. And that enables us to deliver more value to that existing plumber, and then when a sales rep gets on the phone and says, plumbers near you are doing really well, that helps them convert if the person hasn't made it through the self-serve [indiscernible]. So that's all sort of ends up aggregating to the positive.

That makes sense. And you alluded to some of the work you've done on product, and I want to come back to that in just a second. But first, sticking with revs. One area of the business, I don't think gets enough attention is Yelp Audiences. Last disclosure, I think in 2Q, it was about a $45 million run rate business. How is demand trending there among some of the larger advertisers that might not otherwise have a need to advertise with Yelp? Have you ever -- have you seen any changes in demand given the current environment? And longer term, should we expect this to become a more meaningful part of the business?

So for folks who aren't familiar, we have a product that we call Yelp Audience. That's where we're taking visitors to Yelp, and we're creating an audience out of that group of folks. And then we are enabling advertisers away from Yelp to use that Audience in their advertising targeting. So it's a retargeting product, and it is CPM based. So -- and that is really an enterprise product. There are mid-market folks who use it. So it's not an SMB product at all. And we were at about a $30 million run rate in the second quarter of '22, and we're at about a $45 million run rate in the second quarter here in '23. So it's grown nicely and it has been an opportunity for us to both have a broader conversation with existing enterprise advertisers. So any of you who've been doing well on Yelp, here's another opportunity for you to take advantage of the audience on Yelp but away from us. So that's one thing, we like it, of course, because it enables us to monetize those visitors without increasing ad load, we like that a lot. And it also enables us to have conversations with folks who wouldn't otherwise have a reason to advertise on Yelp to come and advertise with us. So that could be, say, credit card. So we have seen interest really across the large national players in the product and we've been really pleased overall with the way it's grown. I think that product has a lot of opportunity to grow further. First of all, $45 million is truly a drop in the bucket, again, for ad spend for these players. It does tend to be a bit more brand oriented. So that is an element and there's the backdrop on brand broadly, but we do brand studies so that folks can assure themselves that they're getting a return on the dollars that they're investing. But we're not yet tapped into all the channels where we could deliver audience. So for instance, an area that we're continuing to work on is still very, very nascent for us would be CTV. And we do earn a lower margin on this product. So we're very sensitive to that. And obviously, advertisers want to be able to pay a competitive CPM for the product. And so if the -- it depends on the margin profile of a particular channel, may be less or more attractive for an advertiser to use directly through Yelp. So we do a lot of work in order to both get the best possible pricing from our perspective to acquire those display units, but also that enables us to be in a position where we can compete effectively with other display ad channels for those advertisers. So put a lot of work both in the [ selling ] side, measurement side and, of course, the product side.

And it's all really incremental from here because it doesn't cannibalize the base ad [ sales ], it's getting advertisers that otherwise wouldn't have been on Yelp. It's getting advertised off platform anyway. So it really is complementary to what you have in the core ad business.

Yes, we think it's super complementary. And again, I just underscore for us being able to tap this audience off Yelp is really interesting. And one of the things that we have been spending time on. And again, we'll come back to it, but is, we spend very, very little on traffic acquisition certainly compared to peers. And so we see a broad opportunity to increase -- well, actually really start spending in SEM. We have had a test budget this year. But just on this theme that there's lots of traffic away from Yelp that we can help to monetize that isn't overlapping with how we're already monetizing. It's actually very, very interesting.

Yes. Yes. Makes a lot of sense. Back to your comments earlier on product. Can you talk about some of the product improvements you've made, areas like Request-a-Quote, that's helped drive project volumes improving in the last quarter? What's the product road map or strategy that you're working with on Request-a-Quote to really get that back to positive growth? Because that's a key aspect for the service economy.

Yes. So -- and maybe just for reference for folks, in the second quarter, projects were actually down 10%. In the third quarter, they were down 5%. If people come through a product that we have, called Request-a-Quote, that's where you're going to go through a series of questions to really refine what it -- is it, what home services project do you have in mind to do. That's like -- as an example, is it pool cleaning or pool lining. That's a very different service pro between those two. And so we've done a lot over the past several years to optimize that tree so that the questions get better. And obviously, everybody hears a lot about large language models. We've used large language models to help us to improve the questions that we're asking. And that feeds then into the ad tech stack because the more information we have, the better we are at that matching. Clicks actually, despite projects being down 5%, were actually up 9% in the third quarter. So that reflects our ability to both monetize that audience at a higher rate, but it also reflects that click-through rate, if you don't present relevant businesses to the consumer, they're not clicking through. Yes. So when you get more information through Request-a-Quote, that enables us to do a better job. So of course, we'd love projects to grow, but again, against this inflationary environment, while people -- seems like the credit card data says, are spending about the same. If things are more expensive, then frequency is lower. And we've definitely seen that actually both on the services side as well as the RR&O side that frequency is down a bit because prices are higher. But the better we get at monetizing, we've obviously been able to still perform.

That's actually a good segue to my next question, which is more on the consumer side. Sounds like for a year now plus, we've been talking about or hearing about how consumers are pulling back on discretionary spend, trading more goods consumption for travel and experiences, spending more away from home. What are you seeing on the consumer side of the business? I think you flagged pretty healthy ad click growth just a moment ago. That's obviously encouraging. But just wondering if there's any changing dynamics as you dig in a little bit deeper on that?

So just to maybe, say the obvious, we do report the overall click growth. But if our choice is between another services click, which is only a fraction of total traffic, or another click for Italian restaurants, we prefer another services click because they monetize at a much, much higher rate than restaurants. And so -- and broadly, just -- if you think about the consumers we want to acquire, given that more than 60% of our revenue now comes from services, and we grew home services 20% in the third quarter, we really are lined up around on the consumer front acquisition of an engagement with services customers. Now we like the frequency that comes with restaurant. It's a reason to come back to Yelp or a broad-based consumer app. They're not one and done after a single project on Yelp. So we continue to invest quite heavily to make the experience more visual. We've got video-ed reviews. Reviews are a great way to drive engagement because people who review obviously do a lot more on the app. But overall, what we really want to see is that continued performance and engagement from folks who have a propensity to do a project. And who are those folks? They're typically going to be homeowners, they're going to be professionals, they are going to tend to be more affluent. And so the better we do at meeting their needs and then reminding them that they could do a search on restaurants or if they're only doing restaurants doing a search in services, that's something that we spend a lot of time thinking about.

So that's how you drive that continued engagement with the consumer even as they're being maybe more judicious with their spend?

Exactly. And so just going back to that. As we start to buy leads off Yelp through AdWords, one of the thing that distinguishes this for us is that it's not just buying a project. And certainly, we want that to work economically, but it's also acquiring that consumer who's already the most valuable consumer we can acquire because they're doing a services project. And if we give -- they have a great experience on Yelp, then we think that they're going to do another services project. And that's a little bit -- as I've said before, it's a little bit -- if Yelp traditionally has been come for the restaurant, stay for the services. This is very much -- come for the services, stay for the services and stay for the restaurant.

Got it. Got it. That makes sense. Shifting gears a little bit to the margin side. You made good progress this year, cost of revenue leverage. Can you walk us through some of the efficiency gains that you've made there? How is that possibly offset by growth in traffic acquisition costs as we think about something like outsized growth in Yelp audiences? Admittedly off that smaller base. And then realizing you're not giving '24 guide on this just yet. How should we think about leverage in this line going forward versus those incremental investments in areas such as like AI?

Yes. I mean cost of revenue for us is really low, sub-10%. I think that distinguishes us. And there's lots and lots and lots of projects under the hood on the infrastructure side that enable us to be more efficient. Now it's important to point out that as we purchase ads off Yelp to fulfill for Yelp Audiences that's something that is an expense there. So that's a component of it, and yet we've still been able to keep it, basically flat over the past couple of years, and that's just because on the infrastructure side, we've just continued to make excellent progress and be more and more efficient. One of the questions we do get is for you, as you apply large language models and AI broadly to those experiences, those are very expensive computationally. With -- that is true if it's an entirely open ended query. For us, it's such a narrow query. I'm looking for a locksmith that applying large language models in that context are just much, much, much less expensive. So for us, we just don't see a huge potential burden. Even as we look to make Yelp more conversational, we don't see -- or we don't have an expectation that, that's going to drive a huge increase in cost of revenue because it's so specific to Yelp, why I come to Yelp is so narrowly defined. So overall cost of revenue, we think that we can continue to be efficient there. But as we grow Yelp audience, that would definitely put some pressure there. And then there's still traffic acquisition cost, which is going to be more marketing expense, but nevertheless, we are increasing that. We're adding about $5 million in the first quarter as we move from what we call testing budget to experimentation budget as we continue to refine how we actually acquire the right projects.

Okay. And so then continuing actually with sales and marketing, what is the dynamic as we're heading into the new year, one with the sales force. Is the current size of the sales force, it's rightsized? Or do you need to flex that up or possibly streamline that more as you shift away from local to the more efficient channels like you mentioned, like Self-serve, which has been a big area of focus. And then how should we think about some of the performance marketing spend?

So on the sales force side, 2020, we did cut the team in half. We've kept more or less at that level. And that team has been performing incredibly well. It took us a while to figure out. We're now entirely remote, or remote-first, I should say. And so all those folks are distributed. It took us a good year, 1.5 years to figure out how to get them to a product -- a good productivity level, but we've actually found that by sales reps being able to live where they want to live, we've been able to retain folks and especially with the best sales reps, much longer. And that tenure as that's increased, has also led to overall improvement in efficiency. So we think that, that has certainly been the right level. We're very ROI-driven. If we could invest there, we would. We'll give guidance on head count in -- for '24 when we get to the Q4 call. But for '23 head count overall for the business has been about flat. That's also true in local sales and even in our -- what we call multilocation, which is mid-market and enterprise sales has been about flat. So this is a good tenured team. They know what they're about. We continue to refine the systems that surround them to make them productive, but we're overall happy when you match that or combine that with the fact that Self-serve is growing about 25%. And that's certainly a product effort, and that has performed very well. People tend to have higher, better retention when they come through and sell themselves on the product. And it's not just about acquisitions, it's also about being able to manage that head spend on the other side. So when you combine all 3 of those, so you've got the leverage that's coming from mid-market international, you've got a tenured, highly efficient local sales team and you have continued performance in Self-serve. That's enabled us to obviously deliver 28% adjusted EBITDA margin in the third quarter. And then performance marketing, we think that we're pretty good at measuring our own spend, and that's something that's enabled us to be very efficient. Also, [ TPAs ] have been softer over the course of the year. So we've gotten more for the spend that we've deployed. So we haven't -- we didn't mind that. But again, we'll give a lot more insight for '24 when we get to the Q4 call, but obviously, it's resulted in what we think are strong adjusted EBITDA margin, not only in the third quarter, but we've shared our guidance for the year overall.

Yes. Makes sense. And sticking with that, that good flow through to EBITDA margin. How should we think about potential further leverage in product dev from here. You've been hovering in this, I think, 18%, 19% territory as a percentage of revenue on a non-GAAP basis and generating some encouraging leverage year-on-year for now more than a year. Are we kind of at the end of life on that optimization, so to speak? Or will you have to lean back in at some point on that?

Yes. So Yelp's journey has definitely been this transformation from being a sales head count-driven growth model to a product-driven growth model. And we see lots and lots of runway ahead of us. Empirically, how to -- in the role that I have as CFO, how do I have a sense for where we are and how much runway do we have there, but we do a lot of experimentation. And so we're AB testing the improvements that we're making to the level that makes sense. And when you're at the end of that optimization road, the wins that come out of those experiments, they decrease, they become quite small. And we still see very meaningful wins out of the experiments that we're running and sometimes they are simple as copy changes. Sometimes it's an introduction of a new product, sometimes it's changing the flow. Sometimes it's the back end of the matching algorithm. It's all of that combined. We still see lots of good wins and lots of road map ahead. And one of the things that's exciting for a company like Yelp, we've been using AI for almost the entire history of the company, both for content moderation, but also, obviously, on the matching side. And when a new technology emerges like large language models, we're just super well positioned to leverage those. And it's not always obvious like people think ChatGPT and I'm going to have this dialogue back and forth, but large language models are broadly applicable, I alluded to a couple of applications before. But just for instance, in your search algorithm, really picking the right synonyms is important and large language models have enabled us to put together a better more relevant set of synonyms for particular queries. And -- so that's a very non-obvious under-the-hood kind of application for large language models, but we're really deploying this everywhere across the site, and we actually just transition our matching algorithm over to neural nets and we're applying neural nets broadly in your feed, what do you see. So you just -- the more you do in a sense and as new technologies emerge, it's almost like it's the ever receding horizon, like how far down, how far in are you. Well, we're always about like somewhere between the third and seventh inning because it's new stuff comes up.

So to actually just stick with that analogy, third or seventh inning, you're positioned to end this year around a mid-20s EBITDA margin. Probably a bit above where you were in '21. How far along are we actually in terms of optimizing EBITDA margin? Is there still room for improvement from here?

We definitely believe that there is additional margin potential from this product line strategy. Now, one caveat, we do have a commitment to lower stock-based comp as a percentage of revenue to less than 8% by the end of 2025. And we have just rolled out to all our employees what their expectations should be around compensation for 2024, and we're shifting from some of the comp away from equity towards cash on an equal basis here in '23. Stock-based comp would be down 20, cash comp would be up 20%. But I would just point out that means that your adjustment is lower by 20% between GAAP EBITDA and adjusted EBITDA. So you got to take that into account when you think about the margin road map. We've got a certain transition period because it takes a year or 2 to play out. But again, we unequivocally see continued margin opportunity for Yelp.

That makes sense. And I appreciate you hitting on the SBC piece because that's one actually we've been getting a lot of questions from investors. So just to be clear, you don't really anticipate a major shift in the overall compensation load. It's shifting from SBC as an EBITDA add back to now it's going to be cash compensation. So no longer an EBITDA add back.

Yes. We do think that there will be leverage over time. And just to underscore something that's quite important. When we say that we had the equivalent of a $20 million reduction in stock-based comp here in 2023. People are granted on a 4-year basis. So you're really saying that we didn't grant $80 million of stock. That is a lot for a company of our size, and that means it's not granted the next year and it's not granted the next year. So we're actually going through a transition. It will -- it should deliver incremental leverage over time from a compensation perspective.

Okay. And that's all kind of hand-in-hand with your goal to reduce SBC towards 8% for revs by 2025?

Exactly. And I would just say by the end of '25. Yes.

Last one for me before we open it up for any Q&A. Just on capital allocation, you've been very consistent in repurchasing about $50 million a quarter. You're now at a point though where annual free cash flow is set to exceed that. So kudos for achieving that milestone. Any updated thoughts on capital allocation as free cash flow more than fully funds that ongoing buyback? And then similarly, on the M&A front, I think it's actually been a number of years since you've done any sort of substantive M&A. Any thoughts on whether that's an area that's more interesting now that private stamps are probably more sensible than they were 18, 24 months ago?

So on capital allocation, we are committed to returning capital in excess of the target cash balance. Just to explain for a second, the way that we get to the target cash balance, operating cash buffer and some cash on the balance sheet for tuck-ins. I'll come back to the M&A question. And so because it's a target cash balance as we generate more cash, we certainly have the option, the freedom to return that capital as we stated is our commitment. So if we can be more efficient as a business and generate more cash, then it's something that we would look to return. That being said, we do have a team that looks at M&A. From a tuck-in acquisition perspective, areas that we would look at are things that could accelerate time to market or enhance the product. We're doing a lot more with photo and video, so a photo or video editor could be helpful. Under the hood for the ad tech stack, things like bidders and pacers are things that are of interest to us. So those would be sort of just to give a flavor for the types of things that we would look to do on the tuck-ins.

But obviously, coming from a strong capital position to begin with, tough to argue with that.

Yes. And I mean, I think we think of ourselves for the company that -- the size that we are, we think that we have a fortress balance sheet with over $400 million in cash and no leverage. We like...

Good spot to be in heading into '24. With that, I think we're up on time, but we'll open it up to any Q&A if we have any questions from the audience. And if not, then we can wrap it there. Thank you so much, David.

Thanks so much.