Welcome to the IAC and Angi Second Quarter 2023 Earnings Conference Call. [Operator Instructions] After today’s opening remarks there will be an opportunity to ask questions. [Operator Instructions] Please note this event is being recorded. I would now like to turn the conference over to Christopher Halpin, Executive Vice President, CFO and COO of IAC. Please go ahead.

Thank you, and good morning, everyone. Christopher Halpin here, and welcome to the IAC and Angi Inc. Second Quarter Earnings Call. Joining me today is Joey Levin, CEO of IAC and CEO and Chairman of Angi Inc. Similar to last quarter, supplemental to our quarterly earnings releases, IAC has also published its quarterly shareholder letter, which is currently available on the Investor Relations section of IAC's website. We will not be reading the shareholder letter on this call. I will shortly turn the call over to Joey to make a few brief introductory remarks, and we'll then open it up to Q&A. Before we get to that, I'd like to remind you that during this presentation, we may discuss our outlook and future performance. These forward-looking statements typically may be preceded by words such as we expect, we believe, we anticipate or similar statements. These forward-looking views are subject to risks and uncertainties, and our actual results could differ materially from the views expressed today. Some of these risks have been set forth in IAC's and Angi Inc.'s second quarter earnings releases and our respective filings with the SEC. 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 releases, the IAC shareholder letter, our public filings with the SEC and again, to the Investor Relations section of our respective websites for all comparable GAAP measures and full reconciliations for all material non-GAAP measures. Now I will turn it over to Joey.

Thanks, Chris. I appreciate everyone taking the time with us this morning and appreciate all the people across IAC companies who are working hard for our customers. I want to share part of a note I shared with employees yesterday, which I think is apropos some of the decisions we made in Q2. We win by building exceptional differentiated products that take the hard work off our customers' plates. We do it so they don't have to. Angi does the work to find the right pro to fix your garbage disposal while doing the work to help that pro grow his business. Dotdash Meredith does the research so you can choose the right product, plan the right trip or make a great meal. And Dotdash Meredith also does the work to help advertisers find the right customers. That's what D/Cipher is. Care does the work to help families find caregivers and caregivers find jobs. Vivian does the work so burned out clinicians can discover their dream opportunity and does the work for hospitals to get clinicians helping patients faster. The better, more efficient and seamless we are at doing the heavy lifting across IAC, the more time we can give back to our customers, and the more they appreciate and depend on us. But every day, every quarter, every year, we have to set the bar higher for ourselves, and we have to constantly do a better job for our customers. That's what this past quarter was about, and that's how we win. Let's go to questions. Operator, first question?

[Operator Instructions] Our first question comes from Cory Carpenter with JPMorgan.

On Angi, could you expand on what drove the larger-than-expected revenue decline in 2Q and how this impacts your outlook for the rest of the year? And then secondly, you were a bit less active deploying capital this quarter. How much of that was due to business terms versus other considerations?

Yes. I'll do the second one first quickly, and then I'll go to the first. The -- as you know, we buy back stock periodically, we deploy capital periodically, and there's not a -- we haven't historically and aren't likely in the future to have a consistent pattern on that other than buying shares when we think it's the most attractive use of our cash in the period. We deployed a lot of cash in the first 5 months of the year, and we thought that, that was attractive time to do it. And we took a breather these past few months, and we'll continue to evaluate that as we always do on opportunities for deploying our cash, whether it's buying back shares in IAC or Angi or any other opportunities in front of us. On the first question, I've been saying for a while that there were areas where I thought Angi had focused on optimizing for shorter-term revenue over the longer-term health of the business and our customer experience and that we were going to make changes along those lines. And for the past few quarters, we've been doing that. In the second half of this past quarter, we saw a further opportunity to do that, and I made the call to make that change, which I have no doubt was the right call in the business. I think the impact of that call, which was really restructuring some demand channels, ramping down channels pretty quickly over the course of the quarter. And this was revenue in the quarter, and it was also profitable revenue in the quarter. We ramped that down pretty quickly, and the impact of that would be most pronounced in Q2. I think we're already seeing July better on a -- July was better on a profit perspective. And I think we'll see the benefit of some of the changes we made in Q2 over the coming quarters. So that impact was most pronounced. But the substance of it was turning off channels of demand or reducing channels of demand that we thought can deliver the ideal customer experience. And that in exchange for that, what we're going to drive is longer-term retention of pros and better homeowner experience on our platform.

Our next question comes from John Blackledge with TD Cowen.

What are you seeing in terms of revenue trends at DDM Digital exiting Q2 and thus far in 3Q? And the performance marketing growth was good to see. How does that kind of momentum play in the back half for DDM Digital rev trajectory? And then just on the margins, what type of cadence could we see in the back half?

Thanks, John. I'll take those one at a time. So top line, just for us in June, it was a key moment in the integration and the performance of the combined Dotdash and Meredith assets. We had said since the beginning of the year, we were aiming to get to flat on Digital revenues and on traffic. We achieved both. We actually had 1% Digital growth in June led by strong performance marketing, and we were able to reach stability in sessions and in traffic during the month of June on Digital across the portfolio. That was led by the former Meredith assets broadly across the portfolio there. We feel good about the -- where those assets are on the migration and growth plans. We can continue to optimize. We're very focused on continuing momentum behind InStyle. And people we feel good about, but we'll always have more volatility just due to the entertainment category. It sets us up well for the second half. The third quarter, we -- as we indicated in the letter, we expect at or around flat, maybe slightly negative. That's due to a combination of ups and downs, continued growth in a lot of the Meredith properties, stability in certain Dotdash properties. We have very strong Amazon Prime Day. But also we've seen some softer traffic trends in the entertainment category as well as some partner sites. So Q3, we expect sequential growth, but year-over-year on the top line, flattish to slightly down, but we expect a very strong Q4. And the performance, the comps, the trends, we feel good about where we are, including the tailwinds of performance marketing, which we really weren't able to fully roll out to the Meredith properties last holiday as well as support from D/Cipher. Performance marketing was a key theme of the acquisition and really rolling out the Dotdash e-commerce integrations to the best-in-class Meredith brands. 12% growth in the quarter was great to see. We are heads down continuing to improve that and expect accelerating growth there on a forward basis, and there's a lot of opportunity. And then on margins, we've said we expect incremental EBITDA margins given how -- where we have the cost structure and the efficiencies we've driven to be in the 80%-plus range. And you can see in sequential Digital revenue growth and EBITDA improvement that or better. We expect to continue to see improved margins year-over-year and also on a sequential basis in Q3. And then Q4 seasonally is always a major quarter for Dotdash and Meredith, and both in terms of advertising and performance marketing revenues but also margin scale. And we see -- we forecast strong [indiscernible] in the fourth quarter.

Our next question comes from Eric Sheridan with Goldman Sachs.

Maybe 2 on Angi, if I could. First, following up on Cory's question from earlier, you really stuck out to us in the letter that you talked about lost revenue and used a phrase like good riddance. Can you talk a little bit about how much of this process of mix shift on revenue is inside your control versus outside your control and where we might be in the evolution of getting the type of revenue you want in the Angi business? And then the second along the same lines, it was interesting to see a paragraph on international and Angi in the letter. Can you talk about some of the key learnings from international? I think you've used the phrase that it's a good leading indicator for the trajectory the U.S. might eventually take, and maybe give us a little bit of color on that trajectory?

Yes. Thanks, Eric. So we did -- when making a lot of the changes in Angi, we did, obviously, the easiest stuff first. And maybe I'll break the revenue down in a few buckets. There is the revenue that came with either zero earnings or negative earnings. And that stuff is easy to get rid of especially if it has a bad customer experience or not an ideal customer experience. And you saw a lot of that happen very early in Q4 and Q1. So things like managed projects, the more complex services which we discontinued or other channels where we were doing unprofitable revenue. That's relatively straightforward, and I think we've cleaned up all of that. The second bucket, and I think we're pretty far along in the second bucket, though not completely, the second bucket is where we were bringing service professionals onto the platform that weren't really well set up for the platform. And so really didn't end up covering their sales costs. And that meant they turned too quickly. And we're far along through that. So those folks would have generated some revenue but probably would not have generated profit over the lifetime relative to their sales cost because they didn't stay long enough. The third bucket, which is probably the -- which was what you saw happening in Q2 and where we do still have some work is where we're bringing in demand, we're bringing in revenue. And that revenue -- that demand can generate revenue in the period but can drive higher retention -- or sorry, can drive higher churn among our service professionals or among our homeowners. Meaning that they'll -- that would be detrimental to the lifetime value of that customer experience. And that was really what the focus was of the…

The only thing I'd add on top of that is the second quarter last year was sort of peak for -- as we mentioned, for...

Empty calories.

Either low-calorie revenues or regretful revenues in terms of -- for pro experience or higher pro churn. So you can see some of that in the graphs on SP retention, where coming out of that period, you had some of the real drops in SP retention. You also have roofing, which had its peak revenue in the second quarter last year, which still not -- the business still isn't where we want it but has an oversized impact on Angi revenue declines in aggregate year-over-year. Thank you.

Our next question comes from Jason Helfstein with Oppenheimer.

I'm [indiscernible] Oppenheimer. So 2 questions. First, just a little more color on Dotdash Meredith, Dotdash. Are you seeing any competition for engagement given the massive increase in short video Reels, TikTok, YouTube shorts, et cetera. And then you did kind of talk about your outlook, but maybe go into some detail on categories, what you're seeing in the second quarter and what you're looking at the third. And second question just on MGM. Obviously, a very timely investment there. I think there were thoughts originally behind it to get exposure to online sports betting and iGaming while obviously participate in the rebound in Vegas. Is there a way to parlay that into a more direct play on online sports betting or iGaming?

I'll go first, and I'll turn it to Chris. Just quickly on the social media side, but we compete for attention with any other form of media, and we compete for advertisers with any other form of media. I'm not sure that the -- we could attribute anything that we're seeing directly to things like TikTok or taking share from the content consumption that we drive, the sort of content consumption that we drive. But as I say, we compete with all media in that context. On MGM, we are -- we've actually spent quite a bit of time looking at opportunities in this space. Came close on one, but we've learned a lot in our investment with MGM in terms of gaming, gambling and opportunities there and what's working. And I think that's an area that is interesting, but nothing immediately on the horizon there and continue to be excited about the progress that MGM and BetMGM is making, which is very impressive.

And going back, relative to building on Joey's point, the demos we have and where we excel, we wouldn't -- we don't see losing any traffic to those sites. And in fact, they are revenue and reader lead generation opportunities, given many of these top Meredith brands have limited social media presence. So Neil Vogel and team have been very focused on building the social media integrations as a channel for a lot of these brands, also expanding our own video presence there. We've got good infrastructure to do it, but increasing the short-form and medium-form video that is produced off of our properties for YouTube, for TikTok and Instagram and other channels. So we don't view it as much of a threat and more of an opportunity. Macro, we'd say the market is still soft from a premium demand side in brand, but it is a second derivative positive where if the programmatic market is down 5% to 10% as we said in the letter, that's definitely an improvement over what it looked like in Q1 and parts of Q2. Categories that are strong, retail, beauty and style, travel and auto as we said in the letter, some of that is due to just straight absolute growth in those categories. Some of it is due to how soft the second quarter was for them last year, whether be it supply chain or retailer backup. And then weakness, finance continues to be weak. I think that the category is trying to find its footing in a higher interest rate environment and lapping comps that were very aggressive a year ago. Telecom, we don't see much spend outside of one or two players, and big players are trying to figure out how they want to position themselves. And then entertainment and streaming, we think that is both a category that is -- streaming. The operators are focused on their model going forward and profitability. And entertainment and streaming, we're concerned about the strikes and ongoing strikes there leading to reduced advertising for new shows and fighting churn. So that's it. We expect the comps to get easier even if the absolute strength in it doesn't -- in the market does not improve, but it's one step at a time in the ad market today.

Our next question comes from Stephen Ju with Credit Suisse.

So Joey, I had a question on D/Cipher. So guess versus the environment in which we were still using cookies, how do you think advertiser ROIs compare before and after? Wondering if there's an opportunity for you to capture more wallet share as a result of this move? And looking a bit longer term, I think Google has been working to also move to a cookie-less world with an implementation target, I guess, sometime next year. So should we be thinking that this potentially shields you from potential sort of external risk?

Yes. Thanks, Stephen. That's exactly the goal with D/Cipher. So Dotdash Meredith has data now that says intent outperforms cookies like-to-like. And it's not surprising to us. I mean, we've been as an advertiser for -- across billions of dollars across many of products, we see the same thing. Google has historically -- Google, which delivers intent, has historically outperformed every other channel for us, demographic-targeted channels, cookie-targeted channels, et cetera. Intent is a very meaningful signal, and D/Cipher really mapped intent successfully. And so we're now proving that for advertisers. Key for us is just getting advertisers to give it a try, and then we can move the ROI. And yes, that also is very important right now on the cookie-less platforms like iOS and Google. As said in 2024, they're going cookie-less. And so we should be well positioned for that. And we're going to keep building the kind of content that chose intent, keep mapping that content and then proving that content with the ROI for advertisers.

Okay. And I guess, secondarily on Angi, reading between the lines on the shareholder letter, it sounds like you're looking to play a little bit more offense in 2024. You've cleaned up and shed some empty calories there. So what do you think still needs to be done from a product perspective?

Yes. We talked about a lot of the work we've done on the pro side. Obviously, the work is never done on either side in terms of product. But we -- we're very encouraged by the progress we've seen on the work we've done on the pro side, improving the pro experience. And that progress is most evident in pro retention, which we showed in the letter. The second half of the year is really focused on the homeowner side. Again, we're constantly doing both. But we're looking to launch more innovation from the product experience on the homeowner side over the -- of course, the second half of this year to set us up for offense in 2024. And there's a bunch of things that we're working on there, but probably the most noteworthy change is something I mentioned not this quarter. But I think the last quarter or the quarter before, where we start to surface the directory earlier. Meaning that homeowners can come to our platform and find and interact with pros more quickly on our platform. We think that has a great opportunity for the pros on our platform. We think that has a great opportunity for homeowners to drive engagement. We think that has the opportunity to drive conversion. And ultimately, what will matter on the homeowner side similar to what we've done on the pro side is driving retention and repeat rates. And that's what we're aiming to do and get to a healthy place to really start rocking in 2024. Thank you.

Next question comes from Brad Erickson with RBC Capital Markets.

Just 2 follow-ups on Angi. Talking a lot about the pros this morning, I guess, and some of the pruning you're doing there. Angi, I guess, always -- also had a tough time on fulfillment rates, so the demand that was coming into HomeAdvisor, for example. So how do you kind of balance this [indiscernible] on pros or reconcile it strategically longer term as you look to have the right SPs but also having enough to support growth? That's the first question. And then second, on Angi Services, where are you there from a category perspective? Do you feel like you're kind of fully baked in terms of the range of services you offer, and now you seem to get the demand going on that you're talking about? Or is there still work here to bring on new categories that can be kind of a further vector of growth there over time?

Thanks, Brad. It's great question and something that we are -- both are things we're very focused on. On fulfillment, we've been improving on fulfillment, notwithstanding the lower nominal service professional account. And we look at something which is a horrible internal acronym. But ZACBR [ph], which is zero accept contact or booking rate. And so it's looking at how often we offer no solution. And we've been shrinking the extent to which we offer no solution, and that's good segue into services. So one of the things we'll offer is matching with a lead pro or an ad pro. And one of the things we'll offer is the opportunity to get project done through our services platform. And not everyone will convert into services, but offering people services can be a significant value add and can drive repeat rate in the sense that sometimes customers are looking for pricing, and they can find pricing there. Sometimes customers aren't ready to make up their mind, but they want to see what the options are. And so giving them a fulfilling experience, meaning they know they have -- they at least have the option to get something done and they know what the price would be to get something done can drive customers back to the platform. We think services is enormously valuable in that context, and we think service is a significant competitive advantage. That gets to your second question, which is how do we expand services, and we absolutely believe we can expand services. Obviously, we've shrunk services through the course of 2023 because I think we grew it too quickly, but we absolutely believe we can. Right now, we're in services that we can price accurately remotely and where we can generate a margin that's a good, healthy place to be. And the way we expand from here is figuring out the next categories where we can -- first step being priced accurately remotely. And we have tools for doing that. We have specific categories which we think are the likely next candidates for that. But we're not rolling that out this year. We are -- that's something that we'll do in the future, but we do absolutely have the potential to expand that and do intend to expand that when we're in a place where we feel comfortable.

Next question comes from Ross Sandler with Barclays.

On Dotdash Meredith, as we look into '24, what are the puts and takes that would allow the business to grow faster or slower than the overall digital ad market? How do we think about that? And then on the SEO impact from these new AI search pages, one of your peers recently said that SEO traffic to their premium properties, which -- many of which are like Meredith sites are seeing a nice benefit actually from things, changes since earlier this year with SEO traffic growing a couple of orders of magnitude faster than the baseline. So what are you guys seeing? As Google now adds hyperlinks to these new search generative experience pages, what are you expecting to see, if anything, from that SEO downstream traffic?

You want to do the first one, and I'll...

Thanks, Ross. In terms of '24 or you think about the rest of '23, there's 3 main digital revenue streams: advertising, both premium and programmatic sales; performance marketing, e-commerce and services execution there; and then licensing. There's good reason to project both for all three going into 2024. On the advertising side, we've worked through the toughest of the pandemic-elevated comps. I feel like we've got the sales force in a good, structured, unified way executing across categories. And with D/Cipher, we think we can take share in both premium and programmatic over time at attractive monetization rates. The performance marketing has really been moving from strength to strength. We expect to continue to do that and create more content, more integrations and fine-tune them in a better way. And there's no reason why it shouldn't continue to be a strong source of growth in '24. And then similarly, licensing has passed some of its tougher comps, working through things at Apple News and other markets, and we should see growth there. So the -- and then on the audience side, both -- we expect both the historic Dotdash and Meredith sites to be increasing traffic. I'll turn it over to Joey. But relative to generative AI and be it Google or Bing, we have not seen any loss of traffic so far in our properties and are having active discussions about how we can be a partner, but I'll turn it over to Joey.

Yes. The only thing I'd add to that is those platforms are built on a healthy Internet ecosystem, and sending traffic -- their fundamental business model is sending traffic out to the rest of the Internet. And what they've said and what I believe they intend to continue is that doing exactly that and doing more of that every year. And so I think it is very easy to imagine a scenario, although we really don't know how this is going to play out, and I'm not sure they know how this is going to play out, it is very easy to imagine a scenario where they figure out how to use these tools to send more traffic out to the rest of the Internet. And there’s – again, that’s fundamental to their business model, but there’s also real value in that in the sense that we’ve talked about the fact that these platforms say and this technology is not yet reliable and not yet accurate. And so it can provide information or snippets or answers that users will look to platforms and brands like ours to for validation and which we think that the platforms can send to us for validation. And we view that as a – that certainly has the potential to be a positive. But again, I don’t think the user interfaces are hardened yet in a way where we know exactly what’s going to happen there.

Our next question comes from Brian Fitzgerald with Wells Fargo.

We want to ask about Care leadership change there. Given the new phase of the business, and the leadership change, does that imply anything about how you're planning to invest around Care.com?

Yes. Thanks, Brian. No, I don't think that, that changes anything along those lines. I think that Brad is very much focused, I think, on a continuation of the existing strategy and meaning execution against Instant Book and getting that product working, and fundamentally growing the caregiver base and family base in ways that delight both sides. But I don't think there's any fundamental shift in probably the core business strategy nor from IAC's perspective, capital allocation up or down from there.

It's -- we like the margins at that business. It scaled well as it's grown. And the incremental margins, gross margins are very solid. So we look to drive organic growth. And then as always with IAC, if there are M&A or inorganic opportunities, we'll pursue them as they arise and make sense.

Next question comes from Youssef Squali with Truist Securities.

So maybe just a follow-up to Brian's question on Care. Maybe can you talk about how big the transactional book offering is today? And maybe how should we think about that opportunity relative to subscription? And at Dotdash, are there any common characteristics causing some of the titles like Parents, InStyle, Shape to trail in terms of traffic? And what are you guys doing to reverse that?

Yes. I don't think we'll disclose a specific breakdown of revenue with Care. But the one thing I want to sort of correct in your question or how you're thinking about it is I actually view transactions long term as a driver of subscriptions. So the -- we can offer user a transaction on booking. But really the goal in that is to show the value of transactions, show the value of the platform and ultimately show the value of subscriptions to the product. So I think if we're successful on that, I'd like to see transactions actually ultimately drive more subscriptions through the platform and drive more retention through the platform on the existing subscriptions. Do you want to do the...

And the only other thing I'd add on Instant Book is it really opens up. There's so much traffic that comes through that doesn't sign up to a subscription. So allowing them at a higher rate to pursue a transaction prior to a subscription increases the engagement on the property, increases the experience and may lead to transaction-only customers or may lead them to execute a subscription, but very much increases the breadth of offerings to the traffic we get. With respect to the properties that we'd characterize as laggards in the Meredith portfolio, it is pretty idiosyncratic of why. The InStyle is a good property that was struggling when we bought it. It also -- in the theme of empty calorie revenues, they would have had the most empty or low-calorie impressions and sort of off-spec things that were driving clickbait-type activities. We really -- the management team has really looked to clean that out. We've also -- we're excited by new editor we brought on there and just repositioning it. So we feel good about the property and where it will end up. It's just a bit behind the -- or it's behind the curve relative to some of the others, but we feel good about the prospects there. Parents is, again, in a category with the family and a bit of mental health that weaves into it where very high traffic patterns a year ago, but there's an action plan there to improve. And then Shape is tiny. Neil and team would love us to remove Shape because it's noncore and small, but we keep it on there just for completeness. So each of them has their own situation. And definitely two of them, we feel good about the opportunities in those brands. We just got to keep executing.

Next question comes from Brent Thill with Jefferies.

Joey, on Angi, when do you believe it can return to a growth asset? And ultimately, what's it going to take? What are the pieces that make that recovery?

Yes. We said we think we'll get back to growth in 2024 and the -- there's a bunch of pieces to that. I think the margin part is easier. I mean, we're already back to growth on margins and profitability. At year-to-date, we've generated $55 million of cash flow. That's up $110 million per year. So I don't want that to get lost in all of this narrative in terms of what's happening there. But I assume your question is mostly focused on revenue growth, which is fair and certainly a focus of ours, which, as I said, I think, is in 2024. There's a few elements to that. One is SEO or traffic or top-of-funnel audience. So we still have the reality that the Angi brand, Angi.com is growing healthily and the homeadvisor.com legacy site is declining. Now the magnitude of that continues to, in a sense, get better over time, and that HomeAdvisor is a smaller and smaller piece of the total. But we're still dealing with that drag. As we get into 2024, that becomes easier. The other piece is we talked a lot about and we'll continue to talk a lot about retention on pros. So we're bringing a pro base into 2024 that is healthier, retaining longer, spends more. And I think that will be a help there. The other piece is driving conversion and repeat rate on the homeowner side. That's the focus of the second half of this year. And I think that can be a real driver to growth if we deliver on the things that we're working on right now. And the other thing that I want to mention, we talked about a little bit on this call, is services. I think services is a meaningful competitive differentiator. I think it's a great product with very high customer satisfaction, very high repeat rate. And we have to figure out how to start feeding more demand into that product because we're starving it a bit for demand right now. But if we can figure out how to feed more demand into that product and get that product exposed to more customers in a healthy way, then I think that can also be a driver of growth going forward.

Our next question comes from Ygal Arounian with Citigroup.

I want to go back to the gen AI question for a second. Two points on that. First, in terms of like internally on your content creation, you guys utilize gen AI, has it been a benefit at all? Kind of what are your thoughts around that? And then following up on the kind of search and LLM training piece, there's been a lot of discussions that there's going to be push back from publishers, new media lines around how they use your content or publisher content to inform their AI. Any thoughts around that and how you guys are approaching that?

Sure. I'll start, and Chris can add. The -- we're using gen AI in a bunch of places that -- actually across all of the businesses. But at Dotdash Meredith specifically, we're doing things that are sort of back-of-house with gen AI, meaning content briefs. So starting to put together outlines to that part of the -- our production process more efficiently. We're seeing increased productivity from doing things like that. We also think we can -- we're experimenting with some things on the user personalization side. So we've done, I think, historically a very good job of getting users at the top of funnel and sort of answering their fundamental questions. But we haven't done as good a job on keeping them, retaining them and sort of rotating them through our system. And I think that gen AI offers us a lot of tools to do that, which is figuring out what the related questions are and then figuring out how to use our content specifically to answer those related questions. We're starting to experiment on things like that and are optimistic about what that can deliver. The other piece is on the ad side, where we can set to customize ads and really expand creative infinitely and test creative infinitely for our advertisers, both to find better ad copy and find better targeting. And I think that can work very well. On the training side of the LLMs. Look, I think our position on this is clear. Ultimately, this will be sorted out one way or another. But fair use is a sort of very clear standard. I think we probably have a different interpretation of fair use than the folks with the LLMs do. But we are -- we don't think that our content can be repurposed for the same commercial use we use it for by others. And we're going to protect our intellectual property rights along those lines. And I do think there are productive conversations with everybody in the ecosystem because even if you want to argue fair use and even if you want to argue what intellectual property rights and things like that, the owners and operators of the LLMs know and understand that if they remove the economic incentive for creating the content upon which these things are based, that, that is not a sustainable ecosystem for everybody. And so you've seen some deals get done, and I think there's conversations happening that hopefully will be productive for everybody in the ecosystem.

Thanks, Joey. The only thing I'd add is in terms of real time, we're having a 2-day AI [indiscernible] CTOs going on right now. And when you've got the collection of companies we have, it's interesting to see the diversity of applications. So very much, as Joey said, things like application development, coding, customer service and onboarding functions where you can apply GAI and create new interfaces to engage with whether it's a nurse being onboarded or a potential customer, et cetera. And really access the depth of your database, clearly, anything where there's optimization oriented. And to be fair, many of our businesses have been using AI and machine learning for years, but that's continuing to evolve and then generative AI is -- prevents -- presents even more applications and intentions across advertising, matching tools for content development to your question, et cetera. I think broadly, with the CFO hat on, you can see the line of sight on both productivity enhancements and in many cases, what should be fairly deflationary cost impacts to businesses, be it customer service or people-intensive activities such as advertising, versioning, testing, those types of things. So it is, obviously, such a cliche but evolving rapidly. But just seeing our summit in the number of different applications across our companies has been definitely exciting for all of us. Thank you.

Our next question comes from Tom Champion with Piper Sandler.

Joey, on Angi, how important is getting SPs back to growth? It was flattish sequentially this quarter. And it sounds like better retention, but how are you thinking about the gross add side of the equation?

Look, we're definitely going to need to get the gross adds at some point. It hasn't been the focus recently. But ultimately, we need to get gross adds going, too. There's a lot going through the service professionals math. Remember, we lost a lot on the services side and moving out of the more complex services. And so there's a lot happening there. And still, dollar value is more important than nominal count, but we've ultimately, we're going to need nominal count to grow too. And I believe that's possible in 2024, probably more like late 2024, but I believe that's possible.

Our last question comes from Kunal Madhukar with UBS.

A couple, if I could. One more on the housekeeping side, which is based on the Q and the shares that you reported in the Q, you had about 85 million and changed in terms of number of shares. But in the press release, we talked about shares outstanding as of August 4, that was 82-point something. And so have you bought back more shares since the quarter ended? That's the housekeeping one. And then one on gen AI in terms of trying to understand how big and [barred] will be kind of behaving is for their content, they probably don't need to learn from multiple different sites. So if they choose a specific partner, maybe a couple of partners to learn from and maybe to pay for the copyright and for the learning experience, where does that place your properties in -- within that grand scheme of things?

Okay. I can take both, and Joey can jump in. The delta is the share between that share count as the shares associated with our CEO grant to Joey and not counted because of performance triggers related to those shares. So that is a -- that's the delta between the 85 million and the 82 million. With respect to your question, there's an inherent philosophical question, which is focused on having the best of the Internet or a subset of the Internet in terms of what they're learning and their models learn in. One of the key elements of search and the Internet to date has been identifying the best, as Joey likes to say, in surfacing incredibly rapidly for individuals and doing that in a constantly evolving way that is ideally in a fair and meritocratic manner to use the best of the Internet. If they choose a subset of content to train these models, they will have inferior results. They will have a narrower and less expansive body of information from which they are providing answers to consumers. And as evidenced by the inaccuracy of answers to date and hallucination, as they call it, greater risk of that. That seems to be reversing many of the best things of the Internet and decades of progress and information provisioning. Okay. Thank you, operator. Thank you all for the questions, and have a good day.

Thanks, everybody. So long.

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