Welcome to the Angi First Quarter 2026 Earnings Conference Call. Please note, this event is being recorded. I would now like to turn the conference over to Julie Hoarau, Chief Financial Officer. Please go ahead.

Good morning, everyone. I'm Julie Hoarau, the CFO of Angi Inc., and welcome to Angi Inc.'s first quarter earnings call. Joining me today is Jeff Kip, CEO of Angi. Angi has published a shareholder letter, which is currently available on Angi's website in the Investor Relations section. We will not be reading the shareholder letter on this call. I will soon pass it over to Jeff for a few introductory remarks and then open it up 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 on 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 have filed with the SEC. The information provided on this conference call should be considered in light of such risks. We will 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 will also refer you to our earnings release, shareholder letter and public filings with the SEC and again to our Investor Relations section of our website for all comparable GAAP measures and full reconciliations for all material non-GAAP measures. Now I will pass it off to Jeff.

Good morning. Thank you all for taking the time to read our letter and join us today. We know everybody is busy. Just to repeat a little bit of what I wrote in the letter. We believe we're in the middle of the most transformational time in technology in a generation. We think AI agents and agentic coding present Angi opportunities that we did not have in the same way a year ago or even a few months ago. We believe it's incumbent upon us as good stewards of the company and its capital to move aggressively to take advantage of these opportunities, moving from our legacy platform to a new AI-native technology platform for our core business and flywheel much faster, building agents to multiply the effectiveness of our core customer experience and offer new capabilities to our Pro customers — what we are now calling the Angi Pro Chief Revenue Officer — is the second. And finally, leveraging agentic coding to build these agents in the platform twice as fast as we could before is the third. We have great assets. We're confident that our existing flywheel is one of the best in the industry, if not the best. We have 30 years of brand equity. We have nearly 200,000 active Pros across North America and Europe. We have the most powerful customer acquisition engine in the industry. Our flywheel is going to spin even faster as we deploy agents to improve our customer success rates and serve as a phenomenal distribution base for our new product that anyone building AI software would love to have. Thus, we think we have a tremendous head start and great leverage against the opportunity. For the last three years, we've been working hard quarter-by-quarter to incrementally improve our customer experience and business on an old brittle legacy stack and the resources we've been using to do this are really critical to moving forward as quickly as possible against the much greater opportunities I just described. We have made real progress on the customer experience. I won't list everything I've listed in the past, but moving NPS 30 points and improving Pro churn by 30% are key markers during that period. We've also made good incremental progress moving AI into our key revenue flows with 50% of our homeowners now touching our AI helper in their path. However, we've also not been 100% consistent at delivering incrementally with our legacy technology. The time and costs are extremely high. The incremental approach we've taken and we are taking is not enough, and it's not frankly worth the opportunity cost versus what else is in front of us. We just can't afford to keep our product development teams battling with the core technology to improve quarterly revenue and deliver against specific targets. So we're going to release our resources against the opportunities I just described. Getting to the new AI-native platform is critical because it's going to allow our core product to function more effectively and drive AI-first innovation, improve the customer experience and the efficiency of the business far better than we can on the decades-old code in our current technology. Our core flywheel is going to spin faster and our core experience on both sides of the marketplace is going to be better. Shifting and focusing on building the new Angi Pro Chief Revenue Officer is an incredible opportunity because first, we're going to generate materially more value for our core Pro customers by making sure they win more of our leads, driving retention, engagement, multiplying lifetime value, which in turn will spike acquisition opportunity of new Pros. It's the strongest bet we can make in this business. And then secondly, we effectively will have a new business because our Pros will be able to use the Angi Pro CRO for off-platform leads and the rest of their business, grow and enjoy more success, which is, of course, our core mission. We get more jobs done well for our homeowners and more jobs done well by our Pros. So we have a twofold market opportunity and a huge as-yet-undisrupted market where we have the leading assets and leading market position. So multiple things can be true at the same time. Our mission has not changed. We're focused on jobs done well, as I just said, and jobs won well for our Pros. Our goal is to deliver profitable and accelerating growth over time. And we are also making a clear pivot on how we execute our strategy, given, again, what we think is a remarkable opportunity in front of us in our space, and we think we are well positioned to win. So with that intro, we will move to questions.

Our first question today comes from Dan Kernan with StoneX.

Jeff, I guess the first obvious question, just to follow up on this. We're calling it a pivot, but it's really more of an enhancement the way I think you guys are trying to win business. And so with the reduction of guidance here or the pull of guidance, I guess, for the short term, maybe you can just frame for us how much this is going to impact, in your mind, revenue and EBITDA and over what time frame? And then I want to kind of follow up on sort of how you perceive the market opportunity.

Okay. Thanks, Dan. Good to hear from you. So first, it is — again, two things could be true. We have been going down the path we've been on. I think it's a material pivot in the way we deploy our resources and execute and I think it is a whole new opportunity that we are going to build as well. In terms of your question on revenue, EBITDA, cash flow, we've made a clear decision not to give guidance. We think that setting guidance and the pursuant distraction it is from executing its larger opportunity is not where we should be focused. But what I would say is our existing business and flywheel generate and will continue to generate solid operating cash flow, which we think of as adjusted EBITDA minus our CapEx, and we plan to continue to generate solid operating cash flow. We're not looking to destroy our EBITDA margins or take our cash flow anywhere near zero. We're effectively going to fund our platform and product strategy internally, meaning we're only going to add to our cost base where we see more opportunity. For example, our AI software and token costs will be several million dollars more than we anticipated a few months ago. But by taking resources off the legacy technology and acknowledging that we're no longer going to focus on quarterly revenue, there will be an opportunity cost measured in some amount of lower revenue implicit by not working on the core technology to deliver incremental revenue wins. But to be clear, we don't plan to use the cash on our balance sheet to fund the transformation; rather, we actually anticipate continuing to build the cash on our balance sheet by continuing to produce cash flow.

Does that — just to be clear on that before I ask the kind of TAM question, Jeff. It's obviously not a distraction. You're aiming for a bigger target here, some revenue opportunity loss, but you're — I mean we're still focused on the core business, and we don't anticipate — I mean, I don't — is there any way to kind of frame how big a disruption you think this might be to the core business in general?

Look, I think we plan to operate with a cash cushion. Without this being a commitment or guidance, I think we'd be happy with the cash flow cushion in the range of $50 million a year. That's adjusted EBITDA minus CapEx. That's not a budget goal, a commitment or a plan or guidance, but that's directionally how we think about a floor. And we think that that's a good number that allows us to internally fund the transformation and continue to deliver cash flow to the business. And we think that our core business will continue to generate solid profitability. We think that once it gets onto the new platform, we will have the opportunity to accelerate with innovation and efficiency there. And then I think we'll have the opportunity, as we put our agents in place and get penetration over the next several quarters, to accelerate materially following getting the new Angi Pro CRO infrastructure into place.

So with that, Jeff, I think in the letter, you basically said that your — the $700 billion TAM that you're referencing is just job value and for you guys to get to your $5 billion revenue opportunity, which you lay out there, it just seems like doubling your win rate. I mean, what you're suggesting here is that by building the CRO for Pros, I mean you have an opportunity for them to utilize this both on and off platform. So there seems like there's a software element to this. So maybe you could unpack for us how you think about getting to that $5 billion? And separately, is there a separate TAM that we aren't discussing yet today or in the shareholder letter that could be achieved or attacked from a software perspective, given that most pro marketing budgets are viewed as percentage of job value, but software is typically a separate expenditure line and kind of viewed as sort of a separate TAM when they think about cost of service?

A great multi-part but very smart question from you, Dan. I should expect nothing less. So yes, $700 billion TAM is residential construction, specialty construction, home services — total job value that we think is our target market for our platform and customer base. Today, we capture below 1.5%. The market is split 75% larger Pros, 10 employees or more, and 25% smaller Pros. We think we have 3% to 4% share in the smaller Pro market, and we're under 0.5% of the large Pro market. We have a strong view, AI or no AI, we can replicate the share of the small Pro market in the large Pro market. We think we've underinvested and not executed well there over time. Doing just that — and getting to that share would give us $2.5 billion of revenue with a 10% take rate, which is about our current take rate, where Pros pay roughly $50 a lead, they win 1 in 7 or 1 in 8. The average job is about $4,000. And so we think about it that way. If you look at our platform, 10 homeowners submit jobs. Seven of the jobs get completed, but only two of those are won by our Pros. If you look at our core long-standing strongest brands and businesses in Europe, which would be the U.K., Germany and the Netherlands, they win more like 3.5 out of 7. So we believe that doubling that too is well within reach. And so Pros winning twice as many — four out of seven instead of two out of seven — takes your share of the total job value in the market from 3% or 4% to 6% to 8% or 7% as a proxy. If you come back to the take rate, Pros are looking at their overall P&L and their share, they're paying to support their revenue. We think 10% is a pretty good marker where you're driving good value. By improving the win rate, we would lower the take rate unless we took lead pricing; taking lead pricing is one way to keep the take rate, another way is charging some for the software. So you're correct. There's two markets there. There's the lead market where maybe we'd like to be a little less than 10% to drive real value there. But there's also the software market and based on our research, if you take 10% of that $700 billion job value market, it's $70 billion, that's the potential revenue. We think that there is a comparably sized market, $50 billion to $70 billion maybe, in services and software to sell the Pros that is likely growing as software transforms with AI. So on some level, there's $140 billion of revenue out there. I think our focus is on delivering against the $70 billion opportunity for our Pros, but it is a product that while we're first focused on Angi leads, our Pros should be able to use for other leads and, frankly, for running their overall business. So you're not wrong. And when we think about our $5 billion revenue target, one way to do it is to get 7% of the market and a 10% take rate. Another way to do it is to get a lower percent of the market and effectively have software and services revenue. And then the third leg we have is actually accelerating growth even faster in Europe, which can be a material contributor because there's another $500 billion or $600 billion of TAM in Europe which we've been less successful at penetrating. We think that's tied a bit to market structure, and that's a different conversation. But we think we have multiple ways to get to the $5 billion. And I think you've hit well on a couple of them.

The next question comes from Truist. The line is open.

This is Robert on for Youssef. On the Q1 performance, can you just explain the levers relative to 90 days ago, which areas of the business outperformed and how sustainable is that outperformance? And then what are you guys doing in new LLM traffic channels?

I'll take the first question. So in terms of revenue, we had a strong January and February. Then March pulled us to the lower end of our revenue range, driven primarily, we believe, by macro factors. Service request mix shifted away from larger jobs in categories where we have the most extra capacity, such as roofing and HVAC, and towards smaller jobs. We survey thousands of Pros and homeowners and it's clear that homeowners backed away from projects more in March than in previous months. As a result, Pros reduced lead budgets because they believed they would get fewer jobs. Meaning, overall, we had lower capacity. On EBITDA, our EBITDA came in at about $23 million. That's above our $10 million to $15 million guidance range. There were two contributing factors. First, we capitalized about EUR 2 million more of engineering labor than we thought in our initial guidance. We followed our accounting policy here, and we went by the book. So it went a little bit higher. And second, we had a couple of one-time benefits on expense and some timing. And so we came out above our guidance for Q1.

Yes, I would again say editorially, we look at all-in adjusted EBITDA minus CapEx. So when we have these swings, you can blame Julie for following our accounting policy. But given our druthers, we wouldn't capitalize it — and I don't mean to speak accounting heresy — we just think it makes things more complicated and the cash ends up in the same place. So we're just calling out that benefit. Let's talk a little bit about LLM traffic. We have been investing a fair amount in making sure that we are there for the LLM traffic. We've been buying OpenAI API credits successfully. We're near breakeven on that buy. There's been a bunch of noise out there on it, but we're happy with it. We're in their beta test, and we know they are working on optimizing and we're confident that we're going to be able to grow value and expand there. We've launched our app successfully on ChatGPT. We'd like to see them move their app ecosystem into deeper integrations, and we're working with them on that. We're going to launch on Amazon soon. And we are live working on multiple other integrations with major players, which we expect to announce in the next couple of months. The overall share of traffic from these sources is pretty low right now, but I think we are all seeing consumer usage shift and it will increase. We think the platforms are going to figure out how to leverage this traffic, and we'll be very interested in working with them. If you think about — I wrote this in the letter, but if you think about our approach, our approach and our pivot is about making sure our Pros get better results. When our Pros get better results, our homeowners get better results. And when customers get better results the LLM platforms will want their users to go there. And so we think that in the same way that our results on SEO once kind of led us historically, and we have most recently taken leading positions in buying on paid social for the same reason, we think we're going to do the same here. So we're pretty excited about it. Our approach has been — we've developed technology where we can pick up the conversation in any part of the chat with the context in the chat. So if you were to say, "Hey, ChatGPT," or, "Hey, Claude," or whichever you're talking to, "I have water on the floor in my bathroom," we could effectively let the LLM know, and we will have the ability to pick up the conversation there and ask questions, which are LLM-driven, but with our proprietary domain knowledge fine-tuning the LLM chat. We also can pick it up somewhere in the middle or at the end when Claude or ChatGPT or Perplexity or whomever has diagnosed that, "oh, you have a crack in the base of your toilet," and we can say here's some Pros, Mr. or Ms. Consumer, and get the job done there. We're already taking the same approach with our core homeowner experience. We have in test an LLM-first chat that effectively mirrors this experience. It's currently a conversion experiment, which we want to narrow before we move it broadly. We do plan to lead with this experience when we're working with partners and new traffic channels because we do believe that ultimately, where we want to be is having a full chat with a homeowner, getting whatever information they're capable of, and homeowners aren't always very good at giving the information or assessing the information and being able to provide price estimates, advice, information and, of course, connect with our Pros through the experience. And that is where we see things going, and that being beneficial to the Pros on the other side as well. So that's how we're looking at it all holistically. I hope that answers your question.

The next question comes from Sergio Segura with KeyBanc.

First, I was hoping you could just provide a little bit more detail on what the Angi CRO is going to look like at the product level, any kind of required investment for that product? And just maybe a little color on why this is the right jobs to be done to focus on right now? And then secondly, relatedly, maybe how does your go-to-market strategy change with this new AI approach? And then if you could discuss any challenges or opportunities of targeting the smaller Pros that you mentioned in the letter for this product?

Right. So the reason this is the right job to be done right now is on a simple basis, this is, we believe, the best way to achieve our mission and deliver the best customer experience to both sides of the market. What we're trying to do is make sure that when a homeowner comes to our platform, they hire a Pro from our platform and the job gets done well, and the Pro feels like they've won a job well. Everybody is happy when that happens — customer NPS improves, retention and satisfaction jump, the Pros pay us, and we make more money. Our biggest gap, as I walked through earlier when I was responding to Dan, is the number of jobs that are actually completed versus the number our Pros win. So to drive that north-star experience, our Pros need to win more. There's been a dramatic change in the possibilities available to us with AI agents and agentic coding in just the last few months. We have been assessing and digging in and we believe that agents offer us the opportunity to close the loop and take that metaphorically from two out of seven to four or five out of seven that I referenced earlier, and double the win rate, double the effectiveness for the homeowner and for the Pro, and really grow value in the business and the ecosystem. In terms of how we're approaching this and how it's going to look, effectively what we're doing is starting with the core lead-to-close cycle. So the first agent would be what you might call an AI call center and booking agent. An outbound call can be made to the homeowner; if the homeowner doesn't pick up, the outbound tech can call back in. The booking agent gets more information, confirms the needs, books into the Pro's calendar, sends reminders to the homeowner and the Pro, handles rescheduling, and ensures the Pro shows up. Getting the booking is really the first key anchor in getting the job won. A lot of our large Pros look at booking rate as their key metric. But you can go from there and imagine that you can coach the Pro on the sale going in. You can record the visit and transcribe the meeting — analogous to tools like Gong — and send notifications with coaching advice to close the sale during the visit. You can also take the transcription of the call and have the agent put together a draft quote by the time the Pro gets back to her or his truck or van and is able to dispatch a quote right away. One of the gaps in the winning process is delivery of a timely and accurate quote. And once you have the quote, you have follow-up, checks on changes, closing the deal, and you can ask the Pro to intervene with a visit or a call to close the deal. And you can go from there. We're going to carefully assess the needs and the opportunities to make sure that when the homeowner submits a service request and creates a lead for our Pro, one of our Pros is consistently winning it. You can imagine the ecosystem will look like that. In our timeline, we should have our first agent in its first test in the next several weeks. We will then, as that gets going and we complete our agentic software development lifecycle — which is the platform on which you do your agent development — work on getting our second one out. We're working on prototypes and we hope to demo this at our Investor Day in the fall. We're pretty excited. We think the opportunities opened up here to really deliver value for our customers and then ultimately accelerate the business are significant for Angi.

The next question comes from UBS. The line is open.

This is Vanessa on for Stephen. I just wanted to ask a question on the guidance. So can you add some color on what forecast item is getting more difficult for you to resend guidance on and is it more on the cost side as you build out the product?

So I wouldn't say that we're having difficulty forecasting. We have high visibility in our business. We pay careful attention to what we're doing. It's just very simple: we're not going to give guidance because there isn't a reward for managing the quarterly or annual guidance. There's not any reward for hitting the range on our quarters. There's not any reward for dedicating resources to getting the next million dollars in the quarter versus the next billion dollars of value that's in front of us. And to be honest, the market is telling us that. So we're going to stop trying to invest and improve our revenue on our old platform, which is really just fighting the last war. We believe the upside of our AI-native strategy is, on some level, uncapped. So we believe that anything that distracts from the tremendous prize — management, engineering resources, anything that distracts from that — is effectively kind of a waste of time. We still plan to run our commercial machine and drive the business back to Pro growth and ultimately revenue growth. We're just not putting a timetable on that. Our milestones that we're thinking about are targeting getting onto the new platform in the next 12 months or so. That's a key marker in terms of getting into a place where we can innovate and work on the core business. And then secondly, as I discussed earlier, we're going to sequentially build, test and roll out our Angi Pro Chief Revenue Officer agents. As we get those into place and the new platform rolls out, we anticipate being able to accelerate our revenue in 2027, and we think it should be material. Otherwise, it's not really worth playing for. So I think without giving guidance, that's how we're thinking about it. It's not a problem of visibility or forecasting difficulty. It's simply a matter of where we're prioritizing resources and the feedback we're getting from the market on the value of doing that.

The next question comes from JPMorgan. The line is open.

For the first, Jeff, can you talk about what this pivot business strategy means for the consumer homeowner experience and how it may change? And then for the second, can you talk about the rationale for the debt repurchase in 1Q and 2Q quarter-to-date and maybe provide an updated capital allocation strategy?

So let me talk about the homeowner experience. I talked earlier about the development of LLM surfaces as traffic sources and our strategy there. Consumers are going to have personal agents more and more. Those personal agents will be able to go out and transact for them without them necessarily interfacing with a website. So what we strive to do is be the best place for a homeowner to come to get their job done well. We think the strategy we've laid out continues to be the best approach. We think the strategy positions us to deliver signals to the LLMs to make the LLMs choose us as a destination to get the job done, get traffic and demonstrate that we'll get the job done. When you think about personal agents, they're effectively trained LLMs trained on personal preferences. If we can train the LLM by delivering results and being the choice destination, the homeowners' personal agent will come and use us. So what we want to do is position ourselves not only to be a place where a homeowner can come and use us as their agent to get questions answered and find Pros, but the homeowner's personal agent will come and do that. We think we do that by delivering jobs done well for our Pros, which means jobs done well for our homeowners. I described earlier our thinking about the homeowner experience: the ability to deliver estimates based on information the homeowner gives us, taking photos, iterative conversation, requests for measurements, and so on. You can imagine the way the interactive experience can develop and ultimately that means the Pro has better information, we get better matching, the Pro can send the correct technician and equipment, and we can have a much stronger ecosystem. This can happen via a homeowner coming through an LLM, coming to Angi, coming through a partner, or from a homeowner's trained personal agent. We see the world evolving this way and we need to deliver against it.

In terms of capital allocation, as Jeff said earlier, we're confident in our ability to produce consistent cash flows. In terms of M&A strategy, we are conservative. We capped our share repurchase ability until next year. We repurchased about 20% of our shares outstanding at the time of the spinoff, which was the limit under the set of tax-free rules, and that restriction remains for two years following the spinoff, so until April 2027. As a result, we saw that buying bonds was a good use of capital. We bought about $100 million worth of bonds. That's about 20% of the debt outstanding, at roughly a 9% discount.

To follow on, we are clearly not against buying our shares at favorable prices, but we can't do that until next year. We're not against buying our bonds at favorable prices — as Julie said, we just bought a meaningful amount. We do have to be mindful of creeping tender rules and how that works. So we're not averse to it, but we have to be disciplined. We're not in an aggressive mindset; we're in a disciplined mindset about M&A. We would pursue a great value and a great opportunity that augments our strategy, but we're not trying to take the cash off our balance sheet and buy things outside of our core strategy. That's our thinking on capital.

This concludes our question-and-answer session. I would like to pass the floor to Jeff.

Well, thanks very much. Thanks for all the questions. I think we've laid out what our thesis is here, which is there are really tremendous new opportunities in front of us that are provided by AI agents and agentic coding. We think we're remiss to continue to work on the old technology, which is not easy to work with, nor is it productive to keep chasing quarters and revenue guidance. We are incredibly excited about what's in front of us because we think we have a clear line of sight on executing against our agentic strategy, and we clearly believe that we have the strongest distribution base between our brands, our Pro network, and our acquisition machine in the industry. We think we can spin the flywheel standalone. We think we can add to it by building our agents and then, as Dan pointed out earlier, there's another market opportunity here for us as well. So we're extremely excited. It's going to take us the next several quarters to put it all in place with the new platform and the rollout of agents, but we will be talking to you over time about our progress and how we're looking at the metrics. We just think that this is a unique opportunity and we haven't seen something like this in the last few years for Angi. So thanks again for joining us, and we will talk to you soon.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.