Good day, and thank you for standing by. Welcome to the Angi Inc. Third Quarter 2025 Earnings Conference Call. Please note that today's event is being recorded. I would now like to turn the conference over to Andrew Russakoff, Chief Financial Officer. Please go ahead, sir.

Good morning, everyone. Rusty here, CFO of Angi Inc., and welcome to the Angi Inc. Third quarter earnings call. Joining me today is Jeff Kip, CEO of Angi. Angi has also published a shareholder letter, which is currently available on the Investor Relations section of Angi's website. We will not be reading the shareholder letter on this call. I'll soon pass it over to Jeff for a few introductory remarks and then open it to Q&A. Before we get to that, I'd like to remind you that during this presentation, we may make certain statements that are considered forward-looking under the federal securities laws. These forward-looking statements may include statements related to our outlook, strategy and future performance, and are based on our current expectations and information currently available to us. Actual outcomes and results may differ materially from the future results expressed or implied in these statements due to a number of risks and uncertainties, including those contained in our most recent quarterly report on Form 10-Q, our most recent annual report on Form 10-K, and in the subsequent reports that we filed with the SEC. The information provided on this conference call should be considered in light of such risks. We'll also discuss certain non-GAAP measures, which, as a reminder, include adjusted EBITDA, which we'll refer to today as EBITDA for simplicity during the call. I'll also refer you to our earnings release shareholder letter, our public filings with the SEC, and again, to the Investor Relations section of our website for all comparable GAAP measures and full reconciliations for all material non-GAAP measures. Now I'll pass it off to Jeff.

Thanks, Rusty. Good morning, everybody. We know you're all exceptionally busy and working very hard in this earnings season, and we very much appreciate you taking the time to join us this morning. As you know, our mission at Angi is to deliver more jobs done well to our customers, our commitment to our shareholders to return to growth in 2026 and beyond, and generate more value. In the third quarter, we again posted the key markers for both. The most important metrics we look at to judge our customer experience are: one, our hire rate, the rate at which homeowners submit a service request on our platform Angi Pro paying for that lead on our platform; two, a pro win rate, which is the rate at which pros win the leads they pay for on our platform; three, our homeowner Net Promoter Score, which we survey on a rolling basis; and four, our pro retention. We again delivered improvement across these metrics in the third quarter as we have all year. Our estimated hire rate is up double digits. Our estimated win rate is up nearly 30%. Our Net Promoter Score is up nearly 10 points year-over-year and nearly 30 over the last 2 years. The pro retention continues to improve with overall churn better by 7% in the last 12 months year-over-year and up 26% versus 2 years ago. And we're not done yet. We're continuing to invest to get better in customer experience. We also continue to post the key markers for our return to profitable revenue growth. Proprietary service request growth accelerated in the third quarter to positive 11%, with proprietary lead growth at 16% and revenue per lead growth at 11%. The blue line to growth in 2027 is clearer and clearer to us and hopefully to all of you. Our network channel has gone from nearly 40% of our leads a year ago to less than 10% this year, third quarter over third quarter, making the rate of growth or decline there and impact on our overall growth. But that will change trajectory as we start to compare next year. Our strong proprietary growth is mathematically the key marker for 2026 growth. We'll likely talk about this a little bit more in response to questions later. We're also generating materially more value for the business with our sales channel in Pro acquisition. We have only about half the sales headcount we had a year ago, but we're actually producing more overall lifetime margins, meaning the margin for pro and the lifetime capacity for pro have materially increased. So with the step change that we've delivered in our sales effectiveness and our recent launch and now ramp up of online enroll, we have the key pieces to grow our overall growth capacity in 2026, and we expect to return to nominal active pro growth by the end of the year and the beginning of 2027. So with all these key markers in place, we're accelerating our platform transformation. Today, we operate on four platforms, bringing the United States to one internationally. U.S. platforms, in particular, have significant tech debt in legacy code, which has materially slowed the speed and efficiency of our product innovation and the business in the U.S. And with the rate of change in the landscape increasing with the rapidly growing presence of AI, we have to move forward and get onto a modern technology stack and get off pieces of software which, in some cases, are 20 years old. We've been progressively rebuilding key pieces of our architecture over the last couple of years, but we're now leaning in with the target of getting to a single modern global and AI-first platform by 2027. We've been and will be delivering new AI-first and AI-enabled software and improving the customer experience with it and our business efficiency as well as we go. So this is going to be a progressive improvement. There's no big bang here. And this effort isn't going to hinder our trajectory. It's all built into our outlook. And if anything, the platform work will allow us to accelerate our efforts in the business as we go forward and hit our milestones. Again, with all of this in place, we are looking forward very optimistically to 2026 and beyond. We're never going to be happy with everything, but we do feel very good about where Angi is, and we have even higher confidence that we're going to deliver against our mission and goals going forward. So with that, I think, operator, we're ready to take questions.

Today's first question comes from Dan Kurnos with The Benchmark Company.

Nice to see progress on the pro lead side. But Jeff, last quarter, you suggested you expected mid-single-digit growth in '26. So given that we're seeing much stronger trends in proprietary and obviously, the weaker in network, plus all the migration work you're doing, has anything changed regarding your 2026 outlook? And then I have a follow-up.

Daniel, let's go past. But we are tracking the same target for 2026 revenue growth, as we discussed on the last call. You referenced the mid-single-digit target and that's about right. We expect modest overall service growth with the strong performance in proprietary being offset by the network comparisons. I think you made the right comment that the proprietary looks a little stronger and the network looks a little weaker, and we probably net out around the same. We are delivering this all through very strong paid proprietary channel execution, and we're going to reinvest in branded advertising next year. We expect to double our TV spend given what we've seen on the strength of our branded traffic and our TV performance this year. If you look at overall brand search metrics, which is something some of the larger companies out there are looking at to gauge their overall campaigns, we were only down in the low to mid-single digits in the third quarter versus the prior year, despite year-to-date cutting our TV spend by 70%, which is not, I think, the relationship you often see, that will bolster our growth. And we think our significantly improved customer experience and the solid ROI there is, I think, attributing to that. I think revenue growth rates will likely vary through the year, likely a little lower in the first half of the year as we compare to higher network service request volume, and evening out over the course of the year. I think it's also just worth mentioning what we said on the last call that we expect a little leverage from revenue to EBITDA growth as we keep our strong fixed cost discipline next year.

That's super helpful. And then just, look, second, there's a lot of moving pieces on EBITDA in Q3 and Q4, including the shift to CapEx along with what you guys called out the resolution of two matters that could result in some slippage into '26. Can you just talk through those pieces and also how we should think about CapEx running in Q4 and next year?

Yes, this is Rusty. Our results compared to guidance were influenced by several factors, including better contribution margins, reduced expenses from less hiring, and some timing of expenses. Our international EBITDA increased quarter-over-quarter, which was mainly due to structural changes in the product organization that Jeff mentioned in the shareholder letter. We unified our domestic and international teams to focus on consolidating onto a single technology platform, a project we've been working on for a while. In Q3, the international team directed their efforts toward developing this new platform, resulting in lower expenses allocated to the International segment and more capitalized wages, aligning with our expectations. We anticipate that capitalization rates in Q4 will be slightly higher than in Q3 as we continue to increase the platform work, and this trend will likely continue at a similar run rate through the first half of 2026 before it starts to decline as we complete the platform work in the latter half of next year. Overall, we expect about $60 million in capital expenditures this year and a similar amount next year, but it will be more front-loaded next year compared to back-loaded this year.

Got it. And just, Rusty, could you just clarify what the two matters were? I know it's just timing stuff, but just helpful color on EBITDA, maybe some shift there?

Sure. So we have two vendor-related matters that are from prior years that we had high confidence would resolve much earlier in the year. Both remain under discussion and thus, we're not really at liberty to give more detail on them. There's a chance either or both of them might resolve in Q4, but we're obviously running up against the end of the year. So at this point, that seeming less likely. But we still expect to prevail ultimately, which might mean the impact will slide into 2026.

Our next question is from Andrew Watts with JPMorgan.

First, could you give us an update on what the response has been from service pros to the ads migration? And second, could you expand on what you saw in the network channel this quarter, and how that impacts your outlook going forward?

Sure. I'll take those. So first of all, the ads migration is more than half done this morning. It consists of two 30-day rolling migrations. We're doing them over 30 days because we want to match the contract renewal date. It just makes a lot more sense to the business and the customer. We'll be about 3/4 done on November 15 and then start a second 30-day migration. We've had zero disruptions or problems so far. We've received good feedback from our customers. As you probably recall, ad pros really have no choice as to which tests within a category they received, and no choice beyond their initial allocation of ZIP codes. So it's positive. It will make life better for them, and it will also improve our matching because you'll have pros actually receiving the things they specifically want. So we're getting good feedback. The migration is one of the planks to this all progressive global platform work that we've embarked on. We'll power down the legacy ads platform following the migration, saving money and allowing us to put resources elsewhere. There hasn't been any disruption of any kind of materiality in the P&L. And we really expect that on all of this work, given the way we're working and given our experience. This is our fifth migration. We did five in the European business. And so this is kind of a continuation of the work we've been doing for a while now. I would just say that having something like this come off seamlessly is still impressive that the teams that have been working on this thing tirelessly for over a year deserve a real tip of the hat on the effort and the quality work they've done. Eden, Yugo, Dave, Joe and everyone else. Thank you very much and if you're listening, tip of the hat to you guys. Let me go to the network channel. A year ago, just to recall, our network channel was almost 40% of our leads. It also had in the range of half the win rate of the rest of our channels. Today, the channel is less than 10% of our leads and the win rates have materially increased to be in the same range as the other channels. All of this was planned. We made a conscious decision to implement homeowner choice in January, which means that the affiliate homeowners who were previously auto-matched to available pros are now choosing each pro. Our data internally has said that homeowners who choose a pro were 60% more likely to hire a pro. Indeed, we've seen that kind of lift in the affiliate hire rates. So it has been a win for our homeowners and our pros. We also anticipated that as a result the volume of our leads would come down quite a bit, both because homeowners are going to choose fewer pros than they were auto-matched to and because there's less revenue for service requests to spend on acquiring more service requests. So we expected that. It was in our guidance. We're kind of on track there with a little bump here in the third quarter. We also expected volatility in the ecosystem. When we launched into this, we weren't sure exactly if everything would play out. I think net over the course of the year, we've gotten a bit less volume and a bit more profit than we expected. In the third quarter, we had three of our larger affiliates have bumps down in volume. One of them had to do with the quality of service requests in their affiliate network. A second one told us they had operational issues. The volume came down. And in the third one, we just didn't have as much volume available. Now we've gotten back a chunk in this volume, but not all of it. So we are at a lower run rate. Again, we didn't expect these things, and we also still expect that there will be some bumping up and down as we add network partners and some drop-offs. So at the end of the day, as we look forward, our current view is that we've come back off our bumps. We're at our new run rate. We're constantly farming and looking for appropriate partners. And we think that, again, we're stable. We could go up. We could bump down. We'll see. It's now less than 10% of our traffic. It's not a strategic channel. We're going to take the right traffic that we can match to the right pros and get jobs done well. But this is not something that we bank on as a source of future growth in particular. We're happy to have it and make it work and keep deploying there.

Next question is from Ms. Sergio Segura with KeyBanc.

Maybe starting with AI helper. I thought it was interesting that statistic you gave that it converts at a 2.7x higher level than the traditional flow. Now that's the default experience. Just how should we think about modeling the impact? And I guess, is that informing your view of maybe investing even more into marketing for 2026? And then I have a follow-up.

So let me step back. Let's just talk about our overall approach with AI generally. First of all, we commented in the letter that we made the move to AI first. And what we're doing is we're looking to implement AI across our customer workflows and our team workflows as well. We are looking to, as we build new software, build an AI data. The AI helper is really sort of one of the first prototypes where we are taking an LLM off the shelf. And our approach is to produce a fine-tuned LLM in each case. So this is the first application. A fine-tuned LLM means that we have a set of proprietary knowledge, which is structured in a certain way. In this case, it's our conditional set of service request questions by a task, which we can use to feed and change the way the LLM flows and the conversation with the customer. Secondly, we have a bunch of proprietary data on customer behavior through the product. And in terms of the interaction between the homeowner and pros, that we can also feed. And then as we deploy these products, we get new data. Through all of this, we've created a learning loop, which differentiates our experience from what somebody might get on an LLM with our proprietary knowledge, or proprietary data. So this is our core approach. What we've done with the AI helper is we first deployed it as an open box that effectively said, how can we help you on the side? Or tell us in your own words. When people enter that, and that's ultimately one-third of the customers who post service requests with us, they're more likely to convert. They're more likely to choose a pro, and thus, they're more likely to get a job done well. This started as a depreciation in conversion and the learning loop has sped up and now it looks accretive, and we believe we're seeing some of this in our proprietary growth. I think when you go to the next step, which is, how much work can this be? We don't expect that the other two-thirds of traffic will triple in conversion because there's causation and causality. So you've got to do a split test to actually see what the shift is. But we do believe there's upside in getting more customers through the AI helper. And we do believe that's important going forward. We don't have a big win baked into our numbers because we've actually just gotten the next phase in this test into play. The core of this is when you look at LLM technology, we think it's a huge opportunity for us because we can take an application like the service request path, which is fundamentally a conversation between Angi and the homeowner. We can deploy the LLM to have more effective natural language conversations against a larger body of data than our previously somewhat rigid conditional path. We could end up delivering a better match on our core asset, which is the 100,000 pros who are ready to get jobs done well for the homeowner. Because at the end of the day, we have always taken this conversation with a homeowner in the conversation with the pro, turned it into a conversation between the two of them because we have the largest supplier for us, and delivered the offline experience that people want. Done this on Google. We've done it on social, and now we're going to do it on LLM. We're doing it within our product as well.

The next question comes from Stephen Ju with UBS.

So Jeff, Rusty, I think I'll ask the AI question in a slightly different way. And I guess, Angi's relationship with the broader world, I suppose. So I think we're all looking at shifting traffic patterns because the usage of LLMs has taken up across the globe. So how does this change your traffic acquisition strategy? What's working? What's not working as you think about customer acquisition and service jobs acquisition? And narrowing down the scope of the question a little bit, I think as we've gone through the restructuring over the last couple of years, I think you've taken a pretty conscious effort to walk away from the traffic that was lower ROI. I would have thought that in the third quarter, we'd be bouncing off the bottom, but I think there's sort of a directional quarter-on-quarter decline here that we're noticing in terms of the overall activity. So I'm just wondering if you can kind of walk us through what you're seeing in the third quarter?

So the first question on traffic shifting. There are some indicators out there that traffic is moving around, statistically getting produced. There's also, what I would call the walking around research of everybody you talk to doing searches in places that sound a lot like LLMs or actually our LLMs. Look, our view on this is, again, what I said earlier, we think this is a great opportunity. We're in the middle of building our own proprietary app, to be deployed by the end of the year on one of the major LLMs, and we're in discussions with a couple of the others about deploying our current and then new technology there. So we think it's a great opportunity because we think that our domain knowledge and our proprietary data and the context we have is going to allow us to enter the chat midstream in the LLM and read the context from the customer and get them more accurately and with more expertise to the pro they want. So we think it's a great opportunity. Obviously, there are a bunch of cards. It's very early in the Texas Hold'em hand. So there's a bunch of cards left to come onto the table. But between our development capabilities, our AI team and the ongoing conversations we're having in the nature of our product, we think that we are very well positioned there. We're also, at the same time, kind of rebuilding our content approach. The structure of content that gets served in AI is a bit different, although there's a lot of correlations to the way it gets surfaced in Google SEO, but we're actively looking at what we do and how we do it to make sure we're in play there. And at a minimum, we get the brand impressions. I think then finally, we're actively working with Google on everything they're doing in terms of how they deploy ad space and AI mode and elsewhere. The AI MAX product, which is meant to focus on getting to the right spot against the AI is now over 10% of our spend. So we're literally trying to stay on the cutting edge of everything about where traffic is, where it is going and keep our team and our technology deployed in the right way there. We see this as an opportunity, not as something bad. We see this is actually very good. Your next question was about third quarter trends. And thinking maybe we should have been bouncing off the bottom. I think what we said is we get sequential improvement. We were minus 12% in the second quarter on revenue, and we said minus 8% to 11% on the third, and we came in at minus 10.5%. We had these bumps in the affiliate network, which is a non-strategic channel. Our core strategic channels are growing incredibly healthily. I think all of our proprietary service requests are growing at 11%, our leads are growing at 16%, and then our revenue per lead is up 11%. So if affiliate wasn't there, you had the lead growth and the revenue per lead, you'd have very healthy growth. In some ways, you could argue that our core business, the best part of our business is growing very healthily. It is well up off the bottom. I think the network channel is a quirky channel. It's a group of affiliates who we're working with to try and buy homeowners traffic that's going to match into our network and work well. It's not a big canvas. It's not quite a sort of algorithmically approachable as Google is. It's not as big as the social channels are. We got a couple of surprises at once. This will continue to be a theme. We do think we're going to offset it with this incredibly strong proprietary execution that you've seen growing every quarter. We do think our TV is now performing better than it was, so we're ready to lean in. We also think that our branded social organic is contributing to what we think is an incredibly strong performance in overall Google brand searches. So I think we feel pretty good about all the good parts. We've got a little bit of noise in affiliate. We got a little bit of noise in SEO. And again, nobody can bank on either of these as the key to their business anymore, I think, and they're both less than 10% of our traffic. Look, we're pretty optimistic. We actually feel very good despite a little bump. I take my family skiing every year at Christmas, and we have to connect because we're going to Idaho. Sometimes there's a delay. We've missed the connection, but we always get to Idaho and have a great time skiing and put on the matching pajamas that my wife buys, and have a family picture. So we are feeling pretty good right now.

Next question is from Eric Sheridan with Goldman Sachs.

Maybe one, if I can, against all of the investments you're making across the business. We noticed you also increased the authorization around the buyback. How should we be thinking about capital allocation back into the return profile for shareholders on either a linear level, or elements of you being more opportunistic against the stock price in deploying that authorization?

Great. Yes. Thanks, Eric. Since Q2 earnings, you saw we bought back the remaining shares in the authorization that was outstanding. That amounted to 1.3 million shares at about $20 million. So year-to-date, that takes us to $111 million representing just under 15% of the company. Then in mid-September, the board authorized us to repurchase another 3.2 million shares. We haven't yet repurchased any shares out of that authorization, and we'll utilize that as the Board deems that's an appropriate use of capital. Importantly, as we've mentioned previously, there are limits related to the amount of share repurchases in the two years following a tax-free spin-off. If we repurchase all of the shares under the current authorization, that would take us just under that limit.

And our next question is from Youssef Squali with Truist.

So maybe, Jeff, just stepping back a little bit, can you just talk about the broader picture, the health of the consumer right now, maybe just given the current macro? Has it changed at all on the margin? Maybe any difference between lower DMA versus higher DMA type of customers?

Sorry, can you just tell me what DMA?

DMA, similar to the distinction between higher-income and lower-income ZIP codes.

Thank you. Regarding the overall macroeconomic situation, we believe there was significant disruption in April due to macro events, which lingered somewhat into May. However, we observed a recovery in June, and we've maintained a steady state since then. While homeowner demand isn't booming like it was during COVID, it's also not dropping off like it did during the financial crisis. We consider the situation to be stable. We don't see any notable shifts in trends across different ZIP codes; some ZIP codes perform better than others, but we can't pinpoint any significant changes. So, that's how we view the macro situation—steady for now. On the platform integration, as I mentioned, we don't expect any notable wins from this integration currently. We're in the midst of migrating our fifth significant pro network, and once again, the process seems consistent with our previous migrations, with fewer post-migration issues than we've experienced in the past. This should enhance the customer experience and boost the efficiency of our commercial operations. We've already begun seeing improved efficiency by consolidating our sales force to focus solely on the new product, contributing to increased sales capacity and value generation for professionals. Will we see some benefit from this? Yes, it will be rolled out progressively, with the first impact visible in this migration. We expect net improvements on the homeowner-facing side in the first half of the year, as we introduce various platform components that should enhance conversion rates and the overall customer experience. Additionally, these changes will enable our team to test, develop, deploy, and iterate more swiftly. For several years, our progress in enhancing product and customer experience has been hindered by legacy technology and technical debt. Looking ahead, we plan to maintain our current run rate while incorporating what we already know, executing based on that. Our forecast for next year is relatively stable; we don't expect a significant lift from any new technology. Throughout our six migrations so far, we haven't experienced major disruptions. We have some internal technology specialists working on this, as opposed to our external construction and home service professionals.

The next question is from Matt Condon with Citizens.

Can you discuss the sustainability and the acceleration of service requests? I believe the acceleration is partly due to the shift in spending from the network channel to the proprietary channel. Is there a limit to your marketing efficiency and your ability to drive growth through that channel? Additionally, what are we seeing regarding competitive intensity?

We may see some acceleration in the fourth and possibly the first quarter, although we are not necessarily forecasting that growth in our proprietary segment. However, as we move into the second, third, and fourth quarters, we will face tougher comparisons in that area. We anticipate mid-single-digit revenue growth next year along with modest net service request growth across all channels, with some variability throughout the quarters due to differing comparisons. We also expect continued growth in revenue and service requests, depending on how leads per service request and revenue per lead are allocated between our paper lead and subscriber professionals. Overall, we expect modest growth in service requests and revenue per service request, but we are not asserting that we will accelerate through next year. Our online performance marketing team has had impressive success in recent years, achieving strong profit growth in 2023 and significantly increasing volume growth this year. They have various initiatives, and their collaborations with product and technology partners have been instrumental in that progress. We are focused on continuing to grow, particularly in increasing pro capacity next year. We expect to enhance the lifetime value per pro acquired as we transition from smaller to larger pro acquisitions by improving our prospect segmentation and targeting. We have added talent to that team and see significant opportunities with larger pros, as our penetration with pros who have 10 or fewer employees is much higher than with those who have 10 or more. With the rollout of online enrollment, we expect to tap into another pool of pro capacity. Increasing our pro count will lead to more available revenue if we can effectively manage service requests. We are optimistic about our online execution and the potential for growth in our network and demand. We believe we can continue to expand and are exploring new tools and channels to leverage our core strengths. While we have modest expectations for service request growth next year, ideally, we would exceed those projections, although we cannot guarantee that at this moment. We acknowledge our strong competition; on a revenue basis, we believe we are comparable to the combined size of our next two largest competitors, though we lack precise metrics. Google is likely our largest competitor with its direct-to-pro services, and their dominance gives them significant influence in the market. We recognize our competitors are a serious factor and are closely monitoring their activities. Our goal is to offer the best solutions to homeowners and pros, differentiating ourselves through high-quality experiences. We believe that by doing so, we can maintain our competitive position and remain the preferred choice. Our key strengths include our extensive network, the quality and skill within it, and our well-established brand, which has been built over 30 years by our founder, Angie Hicks, and others. This history of effectively connecting homeowners with pros is a unique asset that our competitors don't possess. Additionally, our marketing expertise and sales reach give us an advantage that others lack. We are committed to continuously improving our customer experience and feel well-positioned for future growth opportunities. If there are any further questions, I am ready to address them.

There is one more question that is queued up, if you'd like to take it?

Sure, let's go.

The last question comes from Ygal Arounian with Citi.

This is Max on for Ygal. Just one maybe on the 2026 EBITDA. I think the language maybe shifted a little better from similar to modest to that more modest higher end from last quarter. So just curious what's driving that? Is that some of the expected efficiencies from the platform migration, or some of those AI efficiencies from the internal tools you're using that you called out in the letter?

So I don't have our transcript from last quarter in front of me. I think we said mid-single-digit revenue growth and a little bit of margin leverage. I'm not sure if you said modest, similar or what we said. I think when we look at our margins next year, we're not predicting contribution margin leverage because we're going to invest up in the branded area. We think we get our leverage by holding our fixed cost discipline, which I think if you look at the P&L over the last couple of years. Rusty and the team have done a very nice job with. We do think we're able to get efficiency by being AI first. We think you put a multiplier on human productivity, whether it's coding or processing sales scripts or doing customer research. So we think we're going to be able to hold our headcount and keep our fixed costs down and realize the leverage at the fixed cost line as a baseline.

And at this time, there are no further questioners in the queue. This does end today's Q&A session and as well as today's conference. Thank you for attending today's presentation, and you may now disconnect your lines.

Thank you very much, everybody. We're very optimistic looking forward. Thanks for coming this morning, and thanks for listening to us. We'll talk to you all soon.