Angi Inc. (ANGI) Earnings Call Transcript & Summary

Good afternoon or good morning, everybody. Thanks for joining us. This is the first Internet session that we're having as part of the tech conference. I want to thank Oisin for joining us from Angi. So the format, you should all be familiar with this, I'm going to start with some questions. You have a chat box there. If you do have any questions, please put them in, and I'll either try to integrate them into the list of questions that I would like to talk about or we'll save them until the end. So Oisin, thanks for joining me.

Great to be here. Good to see you, Jason. Thanks, everyone, for joining.

So let's start with the plan to shift the business from recommending service providers to selling services. Big undertaking. Those of us who've tracked the stock for a long time kind of joke around that to some extent, you're the last one standing in the space trying to do this. How long do you think this will take? What should investors expect for this -- how long this transition will take?

Look, I would -- I'd take a step back and look at the 2 different models. And just to be clear, you got 2 models: one effectively where pros pay us to access customers, access homeowners; and another one where homeowners pay us to do work inside their home. And I don't think this is a shift. I don't think this is a shift 100% of the business from pros paying us to homeowners paying us. I think these 2 models work in concert, and I think they work incredibly well together. I think there are other examples out there of fulfillment companies or e-comm companies that have built really successful advertising businesses alongside their direct fulfillment business. The flip is we're doing one -- we're doing the other one first or we did the other one first. So we built the ad lead business to a pretty successful place before we built the fulfillment business. And the logic for that or the reason for that, if you think about it, is it's a lot easier to build capacity in a broad array of categories, in a broader range of geos with the lead in ad model than it is to build fulfillment. So we're now on a journey to build fulfillment across hundreds of different categories across the entire country, and we're seeing pretty clear signal from the market that it's the right time to do that. We're seeing pretty clear signal, particularly when you think about how large the gap is between demand from homeowners and supply right now, we're seeing pretty clear signal. You see that in the growth rate. Obviously, 127% of Q2, 166% of July. And that kind of -- on a pretty -- what I would think of a small but sizable base, so $73 million in Q2 and obviously significantly growing again in July. And I think the way I'd frame the question is probably what's the duration to get to critical mass by category, and that's the journey that we're on, where we're picking out individual categories, and we're saying, "Let's go after this category. Let's get to the first couple of cities that work, then let's get to major metros, and then let's get to the rest of the country." And I think for each category, it depends on the path we go. We've obviously bought a small roofing business. That's given us an acceleration in the roofing category. But I think each category, depending on the complexity of the category, is a 12- to 24-month journey and that it's just a matter of how many categories you want to bite off at each moment in time.

So to that, the roofing acquisition, is it too early to say, should that be like a playbook? So do you say, "Okay, a decking company. Now it's a fencing company, pool installation." I mean is that -- or do you need some time to kind of work with the business to see if that should be the playbook?

Look, we're 40 days into having closed it. So if we were in an office, I'd say the folks have barely found the bathrooms and found out how to work payroll at this point. So I think it's probably too early to call that as the playbook. I would say this is probably not the first of these. If you zoom out Handy, you could probably classify as the very first one of these, where we said, "Hey, let's get into cleaning and handyman and direct fulfillment, and let's go and do that model." And we're effectively now on the second one, and we've taken the Handy platform. We've expanded it more broadly past that. We've taken the team. We applied them differently. And there's a bunch of other stuff that came with Handy in terms of retail channel and other stuff, too. But if you think about the indicators we're looking for, the first is does the team integrate? Can we actually make this work? The second is can we copy-paste the cities they're in -- or sorry, the category they're in into other cities. They're in half a dozen cities in South Florida mostly. We got a couple more cities on the immediate horizon, a small one and a pretty large one, in fact. And I think after we do that, so after we spin up a couple of more cities, a small one and big one, I think we'll have a pretty good idea whether that model of entering through acquisition will work for certain categories. It doesn't mean it'll work for every category. Again, and every acquisition is different, unique in its own way. But this is obviously a large, attractive category, $10,000 AOV, lots of financing potential, a reasonable amount of margin. Like it's an attractive category. We're also going the organic path, right, where we're actively spinning up or have spun up half a dozen teams to work on individual subcategories internally. And that's a path that we also hope we can do in parallel. Look, we're very customer led here. So we're going to be very homeowner and pro led. And I think that's one of the keys to the roofing acceleration. We looked at the categories that we are doing ourselves. We looked at where the fulfillment rates were high. We looked at where the sat rates were high. We looked at where consumers were converting from online interest to purchase in a high -- at a high rate. And we said, "Wow, that category that we're not particularly great at today is still a pretty good category, and we're doing really well in it." And then we looked at other companies that were doing it. We looked at Total Home. And we noticed a key thing was they were buying a lot of leads from us. So this was a business where they were actively using our ecosystem to serve our customers, and they were just doing it. And obviously, the ownership structure is different. They own it versus we own it. But that was the primary difference. And we thought, "Hey, can we -- if we own this -- they're on a trajectory to expand one city, every year-ish or so. If we owned it, could we do 20 or 30 cities in a year? And what would that look like at its most extreme? And would it be a different business? What are the ways that we could make the product experience better if we integrated it?" And we went down that path, and we got excited about it, and we decided to take it in-house. But it's -- it really is being customer led as opposed to saying, "Hey, we have a definitive point of view on the market." It's -- look, it's the reason why we open up so many categories to fix price to start with. You could have said, "Hey, you know what, just pick one category, go real deep in it, but you might have picked an impossible category that had nuance and that was very, very challenging to get off the ground initially." So we broadened it out, and then where we see traction, we double down and go really deep. And I think you can expect to see us use that as the playbook and use inorganic/organic more as a lever to execute on that playbook.

And when you think about maybe the -- one spectrum, something that's incredibly simple and skewable, right, like change a lock, unclog a toilet to something that's many, many thousands of dollars, biddable type of job, is the -- is your understanding is the take rate potential -- is there a meaningful difference in take rate potential for "easy" kind of skew versus something more difficult? And to the extent while there may be less take rate potential in the more simple, is it very important to have that because -- are the bulk of consumers coming on looking for one versus another, right? So as you're thinking about top of mind -- how do you think about like those 2 ends of the spectrum?

Yes. So look, when I started Handy initially, my thesis was that dollar value was a good indicator of complexity of the job. I have -- I've increasingly come around to a point of view that it's more nuanced than that. And the things that go into complexity are degree to which you can standardize the existing situation. So how standard or nonstandard is like what's there already and the inputs, customer preference and the range of possible preference of the customer and licensing restrictions, local regulation, blah, blah, blah. And then the fourth dimension is like is the -- basically the size of the job. And in a weird way, you'd look at roofing, and you'd say, "Hey, just on pure size, it's a lot more complex than -- it's a lot more complex than changing a lock." You start to clip into the sub-bullets, and you realize, actually, customer preference in roofing, not that large. You largely either want the tiles, shingles, et cetera, that you had before or you have a minor preference for something else. Customers' preference on roofing is not actually that big.

They want a roof that doesn't leak, that is cost efficient to install and gets done on time.

Exactly. You look at smaller tasks, like you gave an example of a lock for a second, the preference on it changing a lock is actually more significant, bizarrely, because the person is going to touch it every single day. They might actually want it or not want it to be a smart lock. They might want it to connect to their other locks. They might want it to match or not match the other things in their building. It might have an alarm system associated with it, yadah, yadah, yadah. Like you clip in, and you're like, "Oh, actually, yes, on size, roofing wins, but on customer preference, changing a lock like wins in terms of it being a significantly more complex task in terms of like the customer has a very strong preference." So it really has allowed us, when you look at those 4 things, to break down tasks and look at a little more nuanced way and say, "All right, well, what are the true drivers of complexity?" And another one is the -- as I said, the degree to which the situation is standardized or not standardized. The ingoing assumption with changing a lock is that most locks -- or the most people will have is most locks are going to fit most stores, and that's like flat out incorrect. Most locks won't fit most doors. So you end up in the situation where you look at it and you say, "Oh, okay, well, the degree to which you can bake in a survey or bake in the cost of a visit into a roofing job, some $10,000 of AOV is like a no-brainer." Like that's an easy one. You do the initial estimate using LIDAR and area of photography. And then you go out, you do the survey, and you say, "Okay, yes, we're all 100% aligned." On a $250, $300 lock installation, you can show up with a wrong lock, and you can show up with multiple versions of the wrong lock, and you don't have that ability to do the survey in advance. Of course, you can get the customer to take a photograph, et cetera. But the reason I bring all this up is there are certain almost like fixed costs associated with remedying a job or assessing a job. And we're actually not that stressed on the margin percentages. We're stressed on the margin dollars. So we're really looking at it and saying, "What are the margin dollars that we can eke out?" And I know that doesn't help the financial model of folks who are running this on a margin percent basis. But we have increasingly come around to how do we think about Angi services on a margin dollar basis so that we look at it and say, "All right, what are the margin dollars? What are the take rate dollars we're going to take out?" Of course, it is the case that the take rate percentages on small jobs are larger or is larger than the take rate percentage on a large job. So the take rate percentage on a few hundred dollar job is obviously far higher than the take rate percentage on a roofing job. And the reason for that is because there are these fixed costs associated with doing certain things at a certain percentage rate that you actually just need to go and take care of. And it's a dollar amount, not a percentage.

And is that what -- is that why potentially the idea of a membership model would be more interesting? So when you're talking about lower-dollar projects, to the extent that somebody is already a member, whether it's you know, hey, look, this is somebody who is generally likely to follow through once they start a process, et cetera, and it helps you kind of recoup maybe some of that kind of loss, that fixed cost to start the process. I mean -- and then how do you convince the consumer it's worth signing up? We know people like memberships, right? We are the membership society here in the U.S., right? But I mean, how do you -- historically, outside of a home warranty, right, how do you get people to start thinking about Angi as the membership model for this?

So if you think about the economic model of the business for a second, there's 3 primary drivers of -- there's 3 primary drivers that have largely been, I don't say untapped, but we certainly haven't pushed them anywhere. We haven't pushed them to a place where they're impacting the business in a super positive way. The first is consumer repeat rate. The second is pro repeat rate, and the third is zero accept rate, so the rate at which we monetize the job. You move any of those 3 things, the business economically gets better. Now the way we think about what the consumer wants is the consumer just wants to get the job done. So anything we can do to help the consumer get the job done is a huge net positive for us and for them. So if you take that to its next conclusion, we spend a pretty sizable chunk of money in consumer marketing, hundreds of millions of dollars a year. We could just as easily take that money and apply it to helping the consumer get the job done, either through lower SP lead pricing, higher service levels, take your pick, or take rates. There are many ways in which we could apply those dollars to helping the consumer get the job done, which is highly, highly, highly correlated with increases in consumer repeat rate. So we help the consumer get the job done. They're going to come back. It's not like it's straight line, like it's straight-line correlation. So what we are focused on is using membership as a lever to get the consumer to engage with us so that we can commit to helping them get the job done. And if we do that, which we've shown again, the cohort data is early, it's light, but we've shown that if a consumer has the mobile app and they're a member, and we put this in the last letter, it's a 3x repeat rate on the person who just comes in and submits the average service request. So that's just like a massive game changer for the business. I think the rudimentary membership product we have now, the pay-to-save membership, it's fine, right? It's light. It's a first product that's in the market. It's got 140,000 members. There's so much more we can do with that product. And we're excited to expand it out. We're excited to expand out what it actually does, what it means to be an Angi member. I think there are complementary services we can bundle in. I think financing is an important part of membership. I think there's a lot we can do in terms of getting to know your home in a way that adds value to you and creates -- basically makes it easier for you to use Angi to do things in your home. I think, again, it's too early for us to say exactly what those are, but I think you can expect to see, over the next couple of quarters, those tests into other ways that membership will add value to people. And again, we're going to be incredibly consumer led. We're not going to say, "Hey, this is the answer. It is this thing." The same way we went out with membership. Whatever it was, a year and change ago, we went out with a bunch of different stuff, and the one that's took paid to save. And we've leaned into it. We've doubled down on it, and we're making it better. And as we think about the next couple of quarters, you can expect to see us take a point of view on probably half a dozen different ideas that we'd like for our membership that we think, ultimately, will be a part of it, but we'll try them all. And in small, a couple of thousand person tests, we'll see where we get the traction, and we'll lean in. And I think that's what you can expect to see us do across the board. So that's what we're doing in consumer. On pro, you can expect to see us do the same thing. We've got a bunch of pro tests out there right now in the marketplace business. You've got the pricing test out there. You got the credit onboard -- the onboarding tests. You got the credits. You've got the verticalization. You've got payments. You've got a bunch of different things out there. And some of them are actively changing pro behavior and actively changing the engagement or perception of Angi in the pro's mind. Payment is probably the one that's the largest scale at $26 million of payment volume in Q2, up 70% in Q1. And I think you can expect to see us continue to lean into where things work, which, frankly, probably is different to where we've -- how we've run the business for the last 5 years.

So let's talk about sales force a bit. How is the sales force organized today? And is this the most efficient way to sign up SPs? Or do we need to kind of test and work through another iteration of how the sales force is organized?

Definitively not the most efficient way to sign SPs. I think we're all aligned on that. We've got a couple of thousand people selling elite products. We got a few hundred people selling an ad product. They are largely not joined up. They're largely compensated in a way that creates behavioral problems. I think every sales force has challenges. Every sales force has opportunities, and what works at a couple of hundred people or a few hundred people probably isn't the same thing that works with a few thousand people. So we're in the midst of a pretty significant sales force change where we're going to go from having a couple of thousand people selling elite products to having teams of people selling by vertical. So that means there'll be an outdoor vertical, a home renovation vertical, you got it. So there's half-a-dozen-or-so verticals that we're splitting into. And we will be building category-specific expertise in those verticals. That means that the reps selling will be knowledgeable about their category. They should know their customers more. Right now, they go from selling a cleaner that has 2 people working for them and are buying leads at $25 a piece to an addition and remodel company in the afternoon that has -- or not even the afternoon, 10 minutes later, that has 500 employees and pays $350 per lead. That's an extreme example. But that context switching from one category or vertical to another, it stops you becoming knowledgeable in the category. But these are very different businesses that have different life cycles, that have different approaches. They have different -- you think about every different part of it. Like it's the difference between selling to a total mom-and-pop shop to selling to a small business that has structure, got gatekeepers, it's got a process and RFP. You're competing with someone who actually has a $0.5 million a year marketing budget. And you're trying to move some of it to the person who's taken out a small business loan and is basically figuring things out from day 1. So I think there's like real value from learning from the reps actually learning about the category. So real value from learning what does it mean to run X type of business. And I think you look at the sublevers, obviously, when you have reps that are in this loop of dialing, then rep attrition is super high. You got to keep replacing the sales force. We can't invest in them because they don't have expertise or they don't have relationships with the pros. If the model flips and we start to look at our reps as deep pools of knowledge about the category, then we can invest a little more in them, and perhaps the single biggest driver of rep attrition or correlation to rep attrition is what do we pay them. If we pay them a little more and they stay a little longer, and then instead of putting those dollars into training and recruiting more reps, we want to put it into keeping the reps we already got. And hopefully, we'll build some category-specific knowledge rather than just sales knowledge, which obviously is super valuable as well. But I think that category knowledge really can change the game. I think the other thing that's going to force us to do is -- it's going to force us to look at the profitability on a category basis. There are certain categories right now where you look at them, and you say, "Actually, this category doesn't actually make financial sense in the lead model." The cost to acquire the pro is too high versus the revenue that we bring in, and as a result, we need to think about what to do. And there's -- if you only have one sales force and one model of selling, one sales force and one model -- sorry, of monetizing, then there's no point in doing this exercise because you don't have any levers to pull, whereas now what we're saying is, "Hey, actually, there will be multiple ways for us to onboard these pros. Yes, we are going to go and have online acquisition. Yes, we will probably have a lower-cost sales force to take on some of the less profitable categories." And then the other dimension is to say, "Hang on. We have 3 different business models. We got ads. We got leads and we got services." Does every single category makes sense for all 3 models? Maybe. Or does it make sense in certain categories to go deeper into one business model versus another? So I think by verticalizing the business and by thinking about it using the lens of like what are the business models you've got, then you can start to become a little more surgical on where the opportunity is. And you bring it all the way back, all the pro wants is to grow their business. And where we have a single sales force and a single way of monetizing that pro, then we're kind of hamstrung, and we have to charge a super high price in a category that might not make sense to have a super high price in. And I think there's real opportunity there to say, all right, if we had -- let's just take an example, and this is not the -- don't get locked on this. But take cleaning as a category. Maybe cleaning doesn't have a high-touch onboarding model and a high-touch sales model. Maybe it's online enrolled. And maybe by doing that, it changes the economics for us, and we can reduce the price for cleaning. And the...

I mean, look, we know that there's plenty of local businesses as you go and they put in $10, $15, $20 campaigns for Facebook, for Google, et cetera. So it's not asking people to do self-service on their credit card. You wouldn't be the first one in this -- the SMBs to do that, so...