Morning. It is really great to see you all in person here today. I'm Mary Ellen, Director of Investor Relations here at Zillow, and I'm really excited to welcome you to Zillow's Investor Summit for AI. Before we begin, I have a few brief housekeeping items that you'll all be familiar with. Quick reminder that today's remarks will include forward-looking statements. These statements are subject to risks and uncertainties that could cause actual results to differ materially from those expressed or implied. Please refer to the slide behind me, our webcast materials, and our SEC filings for a discussion of these risks. As a reminder, this event is being broadcast on the internet and is available on our investor relations website. A recording of the event, along with downloadable slides, will be available after the event on the investor relations website as well. We have also planned a robust schedule of presentations over the next two hours with no breaks for those next two hours, and I'm sure you're gonna have questions. I hope you have questions. Please hold any questions for the Q&A session immediately following the presentations, which will also be webcast. All right. I wanna thank you for taking the time to be with us today. Over the past couple of months, we've spoken with many of you and heard your questions about AI and what it means for Zillow. As part of those conversations, it became abundantly clear that we had an opportunity to turn on the lights about two main topics. First, why we believe AI is an accelerator to Zillow's business, and second, most investors and analysts only ever get to experience the top of Zillow's funnel. Many haven't gone through the transaction experience, either as a renter or a buyer or a seller, and certainly you don't ever get the chance to see what the professional experience is like. Thirdly and lastly, we have not done an investor day since 2013, and we saw this as an opportunity to essentially re-IPO the company in the era of AI. A chance to really introduce all of you to the operational AI execution underway and orient you to the financial discipline alongside it. Jeremy Wacksman will kick off today's agenda, sharing more about our strong position and strategy. Cynthia Taylor will then discuss how we are building the unified data layer across an incredibly complicated transaction to enable building amazing products that help consumers move and industry professionals better serve our mutual customers. We will then hear from leaders across the business who will demonstrate how we are leveraging AI for consumers and industry professionals. Jeremy Hofmann will then translate these discussions to our financial results and our financial discipline philosophy. Richard Barton will wrap the formal presentations before we transition to a live Q&A session with leadership. Following the presentations, you were all out in the gallery here before, we will have kiosks available so that you can experience all of the live demonstrations that we are showing today with our leaders to guide you. We will conclude with lunch, where leaders will also be available for conversations. All right. With that, let's begin. Between social media and Zillow, I pick Zillow. Zillow. Zillow. The Zillow's. We've got Zillow, a real estate application that lets you find all the houses for sale. Gorgeous loading state, and boom, we have a bunch of homes here. Please welcome to the stage Zillow Chief Executive Officer Jeremy Wacksman. Morning, everybody. Thanks for being here. I'm Jeremy Wacksman. See, I've been here for 16 years, and I can honestly say this is one of the most exciting times to be at Zillow and working at Zillow because of the technology shift we're feeling, AI. I don't think I'm overstating it to say that AI is changing everything. It's changing how you all work. It's changing how we find information. It's changing how we search, how we shop, and of course, it's changing real estate. Just like the last change everything moment with the mobile revolution, Zillow is leaning in and leading where AI is going in real estate. What's consistent for us across all of these technology shifts is actually very simple, our strategy. Our strategy is to build great products and services that consumers and professionals want and need. When we do that, we are successful on both sides of the marketplace. So much so, that as you saw in that video, consumers choose and search Zillow more than the category. This has been true for a long time. This was true when the internet came to real estate. This became even more true as we made the mobile shift, and this is true as AI comes to real estate. The question we most get asked by this room is really around the top of funnel, right? Who gets the search? Who gets the first click? How will that change? Well, the reality is, at Zillow, we've long been and believed it is much more than that first visit. In real estate, in fact, consumers have had thousands of options to choose from for finding a listing or answering that first question, right? They choose Zillow because we build the great product experiences that they want and need and tell their friends about. You see that in our results. 235 million unique visitors come to Zillow Group sites and apps every month, and 80% of that traffic is direct to us, right? That is people opening the app on their phone, typing Zillow into the browser, telling their friends about Zillow. That doesn't come from intercepting demand from a platform, right? That comes from building the best experience consistently so they choose us. Yes, we have this huge scale of audience, and we have this brand that they trust. But more important, the reason they choose us is we are a platform for them. We're a destination for their entire housing journey. Not just where it starts and the first question you ask, but the place to go to take action, to take the next step, to move forward with confidence in a transaction. That is our focus and has been our focus: make Zillow the best real estate experience end to end. When we do that, as you see, consumers choose Zillow, agents choose to build their businesses on Zillow, and as new horizontal AI platforms come, they will choose Zillow as the real estate platform of choice. Since the beginning of Zillow, our goal has also been very consistent: make the consumer experience more transparent, empower people with information, bring digital to this category. But what changed along the way was we figured out that finding the listing or answering that first question, that was the easy part, right? The hard part was everything that comes after that initial visit. Ten years ago, we moved into the transaction, right? We bought companies, we built services, we launched new products. We stopped thinking out of ourselves as a search or a dreaming and shopping site, and we started thinking about ourselves as a transaction platform, helping people buy, rent, finance, and sell. That meant investing in infrastructure, investing in workflows and businesses and platforms that the professionals use to transact business, the real rails of the industry. The reason we did that is because moving is not a one-click experience. It's not a category where you start and end and complete your transaction in one visit. It is a months-long journey. Imagine a renter coordinating with their landlord, their property manager, or a buyer coordinating with their agent and their seller and their seller's agent and their loan officer and their closing professionals. We have built the platform to support that entire journey, and for us, AI has been fundamental from the start. Our very first product in 2006, the Zestimate, was AI. Many of you, as Mary Ellen said, know us as the consumer brand, and you experience us at the top of funnel. What you may not know is the massive number of professionals that use our platform today. In fact, agents who use one of our products touch 80% of all transactions in this country. We help fill more than 60% of rental listings in the rental marketplace. Having the platform of scale for the professionals helps us build deeper relationships and be successful with our consumers to help more transactions happen. That shift to the transaction is a key advantage for us now because once you operate across the entire transaction, you get something incredibly unique and valuable, data that compounds, right? We don't just see the search or the initial question you ask. We see the tours you take. We see the affordability data you're using. We see the conversations you're having with your agent, the offers you make, the homes you close on. All of that data, proprietary data and context about our buyers, sellers, and our professionals is context. That context is what allows Zillow to build a far more powerful AI experience than any horizontal LLM can deliver for real estate. Increasingly, that context is a feedback loop. The more people that use it give us more context about themselves and their professionals that feeds the system. That is not something that is theoretical. That works in our system today, and you'll see a lot of examples of that in the products that the team's gonna show you. Let's imagine a real buyer or seller using Zillow. They've been scrolling for months. They've been saving homes. They have lists of homes. They've been comparing neighborhoods. They've been checking mortgage rates. They've been checking their Zestimate, right? Some people are deciding, "Should I buy or not?" Some people are deciding, "Should I list and sell or should I stay put?" Renters are trying to decide, "Should I take the next step and stay where I am or move," right? Every single one of those people at some point hits a wall and they say, "Well, what should I actually do next? What action should I take?" Right? A horizontal platform or a top-of-funnel platform. Well, they can serve up answers. They can give mortgage rates, they can give listings, they can give market trends, but they don't see all of those signals to actually help move the transaction forward. Right? Only Zillow can see things like all the homes you've saved and viewed, and the ones you're coming back to every day, signaling intent. Only we can see all of the available tour times on all listings, what tours are being booked, what homes are in demand, what customers think about those tours after they take them. A renter comparing lease options while also shopping for homes to buy at the same time. A homeowner repeatedly checking their Zestimate and looking at nearby comps in the neighborhood. A buyer's budget in BuyAbility, understanding what they can really afford, and the conversations with their loan officer. Only Zillow has all of this data and context, right? That is what helps them move through the transaction. Which means we can do more than respond to basic questions or have AI set a search filter, right? We can guide and actually help consumers take action. From understanding what a home might appraise for, to gauging real-time demand, to understanding why the Zestimate is what it is, to helping someone decide if they're financially ready to transact. We don't stop at guidance, we help them take action. Booking an instant book tour, getting pre-approved, hiring an agent. That guidance, that strategy only works if you have the full vantage of the category. You need live housing data, the content, you need all of that consumer behavior, the context, and you need to be an end-to-end platform, the transaction infrastructure. Zillow has all three of those things. That is why we can build the LLM experience that actually helps people take action and actually helps people move. Content. You're gonna hear a lot about content today. Zillow has the most comprehensive and differentiated inventory in the country, and that's because we work across all of the marketplaces, homes for rent, existing homes for sale, and new construction. Now, why does this matter? Because most consumers shop across these, right? The majority of renters are also looking to buy, right? The majority of sellers are deciding timing and whether to sell or list for rent. And many buyers are looking at existing homes for sale and new construction inventory at the same time. Because we work across all these, we see how people actually shop, right? And a growing portion of this inventory is unique to Zillow, especially in new construction and our rentals marketplace. Basic listing facts, they are the tip of the iceberg of this category, because it takes more than the address and the price and a few static photos to really understand the home and guide a consumer. That's why we've been investing for years in our content advantage, immersive media. Right? Our proprietary rich media is really twofold in its strategy. It starts with a consumer. We give a consumer a better shopping experience, but it also gives Zillow valuable insights. Listings showcase the 3D tours you can walk through a home, the interactive floor plans that are generated, the virtual staging capabilities, the Sky Tour video you saw. Today, nearly 10% of new for sale listings have proprietary rich media on them. That is a fantastic start, but we have a long way to go because we expect that to become the default experience that buyers and sellers demand all listings have in the future. That substantial coverage we have today becomes very differentiated data. It teaches our system crucial details. What's the light? What's the layout? What's the flow? What's the quality of the house? It informs detailed and accurate square footage, useful in a transaction, and conforms to mortgage standards. Most importantly, it helps train our AI models, right? It makes the consumer experience better, it makes the shopping experience better, it makes our Zestimates better, and it becomes an embedded capability across our AI experience. That content is what allows us to generate the unique context from our buyers. Right? I quoted the stat of 80% of our traffic coming direct to us. Traffic is a count of people, right? What matters is the quality of traffic that you have. We are incredibly proud that we have the actual buyers and sellers using Zillow through their transactions. 70% of everyone who buys and sells a home in America is using Zillow during that transaction process. What they are doing on Zillow is what matters, right? They're not just browsing or coming in and out for a quick visit. They don't pop in and leave for months. They are spending 2-3 hours a week over five months as a buyer. Yes, they're searching and they're finding listings, but they're also using our financing tools, setting their BuyAbility, shopping within their budget. They're comparing homes. They're collaborating with their co-shopper. They're using all of our rich media to take virtual tours and spending hours in those virtual tours. They're booking realtor on ShowingTime and meeting agents. Once they hire an agent, they're messaging with their agent in Zillow messaging. That is not casual traffic. That is sustained deep intent. That is the context that compounds for our AI platform. Every tour scheduled, every financing app, every transaction closed, every lease signed, all of it feeds the system, and that's what allows our AI to become more intelligent, more personalized, and more capable. We don't just answer questions, we personalize the transaction. Understanding the homes gives us the content, understanding the behavior gives us the context, but what matters is putting it together and giving consumers the ability to act, integrating all these platforms, all these software products together. I quoted the stat that agents touch 80% of transactions. That's across a variety of really incredible tools. ShowingTime powers tours for 90% of all listings in the country. Last year, 40 million home tours were booked on ShowingTime. Follow Up Boss is the CRM of choice for 80% of the top teams in this country. Nearly all Zillow Preferred teams are on it. 100,000 agents are using Follow Up Boss to run their business. It's quickly becoming the AI-powered workflow engine for the best agents. Agents are being delivered great high-intent customers because of our software products. Real-Time Touring, which many of you have heard about, those customers convert at 3 times the rate of normal customers. ZHL customers, folks who have been pre-approved with Zillow Home Loans, they convert at five times the rate as non-pre-approved customers. Our agents are using our platforms to run their business and work with high-intent customers. When they get ready to close, dotloop sees closing documents for nearly half of all transactions in this country. Taken alone, every one of those products is an incredible success story. You put them together, connect them, that's how you transform the transaction, right? When you integrate the shopping behavior, the touring data, the CRM data, the financing data, the transaction data, you create context that differentiates Zillow and value that only we can deliver. Sustained intent combined with proprietary rich media provides an AI platform that is directly integrated into the rails of the industry. That is a powerful combination, and you're gonna see many examples of that from leaders throughout the morning and the demos you'll get to get your hands on. Oh, by the way, we're not just doing this in any category, right? We are doing this in arguably the most regulated and complex category in the country. Structural complexity in housing shapes how AI can even operate. In fact, we have to have licensed brokerage relationships in all 50 states just to get listings, let alone answering questions that require a license to properly answer. The volume of complexity in this category is staggering, right? Every state is unique in regulation. There are hundreds of thousands of brokers working across over 500 MLSs, powering 1.5 million agents to do transactions. We have operated in this space for two decades, building deep industry partnerships so that we can innovate, but do it with an air towards compliance and trust of the industry. Doing all that is a massive undertaking. It is a structural challenge to operate in this industry and do it effectively and build technology to actually build a platform. The reality is that complexity and that regulatory snarl shows up as massive inefficiency. The best way to think about that is the best agents and loan officers that you meet will do maybe a few deals a month. The reason they can't do more is this inefficiency. Zillow is clearing the way for that inefficiency. Technology, increasingly our AI-powered technology, pulls away all the busy work, all the waste, so that humans can become super humans. Agents can become super agents. That is really our force multiplier in the industry, helping the most productive agents, the most productive loan officers and professionals become more productive and do more, more volume at higher quality. Right, LLMs are dramatically reshaping how all of us do our jobs. In real estate, that's gonna allow professionals to spend time doing what they do best so that technology can take away the things they don't wanna do that they have to waste all their time doing. We're already seeing that today, and you'll see examples of that today, and that productivity gap is only going to get wider. When we can connect the intelligence and the infrastructure together, you allow agents to do what they do best so technology can do the busy work. That is the real economic unlock in this category. Cameron is gonna share a bunch more about our professional tools and how you're seeing AI show up today to do this. AI is accelerating our strategy to become the housing operating system of choice. That is true today, the way the industry works. That will be true tomorrow as competing LLMs develop capabilities. Consumers and industry professionals choose Zillow. Agentic AIs will need to choose Zillow because we sit at the center of it all. We are building the operating system that makes buying and selling from the very, very first visit all the way to the closing document feel simpler than ever. Because at the end of the day, the problem we're solving, the people holding these phones, the buyers, the sellers, they are making the single biggest decision they'll ever make in their life financially. It is incredibly complex, it is incredibly emotional, and oh, by the way, they do it once every fourteen years on average right now. When you make a decision like that, you want clarity, you want confidence, and you want trust in the platform and the professionals that you are going to hire. AI can help the buyer understand what they can afford. AI can help the agent respond instantly instead of hours later. AI can surface the right insights at the right moment to move the transaction along. Tech will handle the complexity so that humans can focus on the judgment, the advocacy, and the trust. AI is absolutely gonna reshape housing, and the winners are gonna be the ones that empower and supercharge people with the best tools to take action. That is our advantage. I'm really excited to bring up Cynthia Taylor to take you through our product strategy to deliver all this. Thank you. You just heard Jeremy walk you through how Zillow is positioned to win in an AI world, and I am so excited to walk you through our product strategy for how we're building in this incredibly complex space. Zillow's combination of proprietary content, unique user context, and deep integrations allows us to build the products that you are about to see, that we are about to demo for you. We spent years building the infrastructure underneath the transaction itself, and that foundation is what allows AI to truly transform the experience that we provide to consumers and professionals coming to our products on a daily basis. We know we aren't the only ones building technology, and in today's world, it is easier than ever for somebody to spin up an app in the real estate space that does things like generate listing copy or answer a question about a neighborhood. The difference is that those apps are building at the edge of the transaction while Zillow is building at the core. Building at the core, as Jeremy said, means knowing how to build in a complex, fragmented, and heavily regulated environment. There's a lot of actors in this heavily regulated space, and that means that it limits companies without licenses in how they're able to serve their customers and the types of questions and the types of guidance that they can provide. Generic LLMs may not be able to answer questions that start to look like agency or fiduciary guidance or brokerage activity because answering those questions requires a license. Questions like, "How much is my home worth exactly?" or, "Which offer should I accept? How much house can I afford with my finances?" In addition to that, then there's the issue of fragmentation. Fragmentation is the other piece of the problem that Zillow and our products are solving for. A fragmented product has numerous limitations, numerous identities, fragmented data models, multiple workflows. Every single time you go and look at your user, you only get a portion of the picture, and Zillow is unifying the complex, fragmented real estate system with our products and our platform. We have spent years on the product side doing the harder work first. Zillow understands the regulatory landscape because Zillow has operated within this regulatory landscape for two decades. Zillow holds the licenses and has the infrastructure already built to actually participate in the transaction itself, not just observe it from the outside. Integrating the tools, the data, the workflows that power each step of the move, so consumer journeys and professional journeys and workflows are in sync, is something that only Zillow can do. Today we're here to show you how Zillow has already unified the fragmentation, creating a platform that allows AI to bring the integrated transaction to life for our customers. We're building solutions for consumers and professionals alike who are all participating in the real estate journey together. Taking a look at the consumer problem first. In most real estate experiences, as a consumer, every step you take, every professional you work with, every step of progress you make can feel like starting over because you may be working with fragmented systems that don't know the whole picture of your move, of your intent. You may be a renter where your payment history doesn't follow you and doesn't inform your financial readiness to buy. You may be a buyer where the platforms you're using doesn't know the homes you've toured, the preferences that you're unknowingly signaling by the homes that you're lingering on. You may be a seller who's not sure how to take that next step. Every interaction with every professional feels like it starts from scratch. What Zillow is doing in this space, Zillow is building one shared version of a mover's intent across the entire real estate journey. What that means is one persistent consumer identity, one evolving profile of what the consumer can afford across their renting and their buying journeys. One view into a user's preferences, a consumer's preferences, what they've toured, what they've saved, what progress they've made. One view into next steps and when that consumer should take them. Most people don't get stuck on inspiration when they're thinking about renting or buying. Most people get stuck on that action, knowing what that next step is and when they might be ready to take it. Why what we're building matters is because having this unified intent layer is what allows Zillow's products and Zillow's AI to move from answering really basic, simple questions to anticipating consumer action, understanding what a mover wants to do and what they are ready to do next. Zillow's unified intent layer is what creates these product experiences that ultimately drive success for a mover. From a renter's perspective, a buyer's perspective, a seller's perspective, that means that someday can actually start today because that consumer is finally going to be able to make the move. Similarly, on the professional side, you know, the real estate experience for agents and housing providers, the fragmentation can be just as real. In most real estate tools for professionals, each tool keeps its own version of the truth. The CRM has one version of the client, the transaction system has a different version of the deal. Sometimes fields that have the exact same label in different systems actually mean entirely different things. Zillow is building one shared version of the deal across the entire Zillow professional ecosystem. These are the products that hundreds of thousands of real estate agents use on a daily basis. One identity layer, one deal object, one contact model, one document system. What this allows Zillow to do is it allows Zillow to paint a picture into the deal and have one version of the truth. When something changes in one place, it's reflected everywhere. A consumer question isn't just a lead. It informs the professional where this client might be ready to go next. A showing isn't just a calendar event. A document isn't just a PDF. It informs the engagement, it informs the move, it informs the deal. When every part of the system runs on the same foundation, the same understanding of the mover and of the deal, AI built on top of this platform does not have to guess at what's happening. AI built on top of this platform understands the mover's intent, understands the transaction, and this is where it all becomes uniquely powerful. Zillow is building the unified platform for real estate. Our content, context, and integration are the three advantages that make the products you're about to see today possible. When this content, context, and integration are in the same system, the experience deepens in a meaningful way for our consumer and our professional customers. For consumers, Zillow remembers them, Zillow understands where they are, and Zillow helps them take the next step. For professionals, Zillow's products supercharge those pros with AI-powered tools and enable them to do better and do more. We're building at the core for real estate, allowing deep integration into the transaction itself and truly becoming the housing super app. Today, we've brought product leaders from across our rentals and our for-sale teams to show you what we've already built and what we're building for the future. First up, Michael Sherman is going to show you our rental product strategy and end-to-end experience. Nicholas Stephens is going to show you how Zillow's unified intent layer is coming to life today for buyers and sellers. Finally, Cameron Swiggett is going to show you how Zillow's unified understanding of the transaction transforms the real estate professional experience. Thank you, and I'm inviting Michael Sherman to the stage now. Thank you very much. All right. Hello, everybody. Nice to see everyone here. I recently started wearing glasses, but I'm gonna try and do this naked. Eyes are naked for this. The point that Cynthia was making around fragmentation is a really good starting point when talking about the rental strategy. The rentals space is really a microscope on that fragmentation. Our originating strategy was built around the idea of fragmentation in listings, right? There is no MLS in rentals, so there's no central repository, there's no database of all rentals across the United States. We took that as our first challenge. All right. How do we aggregate as much inventory on our marketplace as possible? The reason is simple. It is the number one thing renters want when shopping for their next home. They simply wanna see the inventory, right? First and foremost, far above anything else a marketplace can provide. Solving that, step one. Step two, three, four, five is everything that can now happen, right? Everything that happens next. To give scale to how we approach this segment, there are roughly three times as many renter moves annually as there are for-sale moves. The opportunity is huge, yet it remains the most disconnected, high-friction experience in housing. 64% of renters say that they looked at homes for sale during their for-rent journey. 64%. It becomes imperative that we have a platform that has both listings for sale and for rent in one place, allowing for easy navigation between the two. For most people, renting is kind of the entry point to housing, right? It's the first thing they do in housing. If Zillow can handhold that consumer throughout that first step of their journey and do a really good job, something really personal, very tailored to what they need at that time, we can set up a lifelong relationship with the consumer. We can help them through that first step. We can help them through maybe their second rental, and then help prepare them for financing, help them with that first home purchase. That opportunity gets reciprocated back to the rentals marketplace when they might want to rent again in the future. Our position of strength really comes on the back now of multifamily advertising. We've built the largest aggregation of listings in the United States, and now we have the largest audience of rental shoppers in the United States. Our rental business is expected to grow about 30% year-over-year in 2026, and we're well on our way to a billion-dollar revenue target that we've set out. The revenue outcomes are really just mile markers along the journey. What we are really excited about is the opportunity for a fully transaction-enabled marketplace, a place where renters can come and not just search, find, connect with a property, but ultimately secure, pay rent, and feel confident using our platform for the entirety of their rental life. I'm going to walk through some steps that we've built, some steps that we're exploring today, and I figure a really good starting point is where the renter starts, at search. As I mentioned, we've built the most comprehensive set of inventory in the country, and that's really important. It spans across both long-tail inventory, which we would define as listings, properties with 25 or fewer units in them, and multifamily inventory, so properties with over 25 units in them. This inventory, this aggregation of content is really important because the trust that we're able to build with the renter. When they come to Zillow, they have confidence that they're searching the entire market. Renters trusting our comprehensiveness of marketplace is important, but also the depth of detail on each listing. Real-time pricing, real-time availability, being able to schedule tours, detailed and rich descriptions about the unit or property. All of these evoke more and more trust and confidence in the marketplace that we're building. From search, we move to messaging. Renters, the rental landscape is wrought with fraud. Trust is at a low. There are scams constantly happening across the category. More and more renters are asking us if they can connect directly through utilizing our first-party tools and platform because they have the confidence that they are actually engaging with a human. With somebody that actually represents the property and can actually open the door and help them secure that listing. From that connection, we can move into tour scheduling. You can see how quickly we're moving away from search and find and lead generation into something that's much deeper in the funnel. At this point, we're really orchestrating the beginning of a transaction. This is done through both our first-party experiences as well as deep third-party integrations. These third-party integrations are critical because it's not just about, "Hey, we can display information on our site that maybe a third party has." It's critical because we can build a native experience. A consumer on Zillow doesn't necessarily need to know what's happening behind the scenes as long as it feels cohesive and the next step being right in front of them. We move into applications, and applications is a really pivotal point in the rental journey. It's the moment Zillow Rentals moves from marketplace connector to really a true infrastructure player. Annually, there are millions of rental applications submitted through Zillow. Millions. This gives us unique demand insights. It highlights the incredible value of our product, and I wanna touch on three components of our application product that are unique. One is we innovated on both the experience and the monetization model. Our application is purchased by the consumer one time. It's common across properties, and it's reusable over a 30-day period. This is inherently a better consumer experience. Renters apply to multiple properties before they secure their home, and reducing the cost and burden on that step is very important. Two, it ties monetization directly to the transaction. This is no longer about traffic and leads and volume. It's about high-intent shoppers actually wanting to secure a rental. Three, the application positions us to integrate more directly with both the property management systems who support property management companies with robust applications, as well as the property management companies directly themselves. This is really important. More and more property management companies are asking us if they can enable their application on our platform to create those efficiencies. Remove steps in the journey for the renter and help them swiftly get into their unit and their property. After application, we're in lease signing, right? Lease building, lease e-signing, and payments. Since launching payments, we've processed more than $10 billion of rent on our platform, and adoption continues to grow. In February, year-over-year payment processing was up about 20%. More and more companies are trusting us as their payment platform, and more and more consumers are looking to us to provide that value in their rental experience. Pausing on payments for just a minute, like, it's a really important product because of the value that it delivers, but it's also very important because it extends the relationship we have with the renter and the landlord or property manager, right? It's no longer about just that upfront lease signing moment. It's every month we have an embedded touch point with those parties. Going further in that direction, we launched a product around rent reporting, where we now report on-time rent payments to the three bureaus, helping consumers improve their credit score, right? To us, this is a no-brainer. We absolutely wanna do this. It advantages our renter, both when they want to rent their next home as well as when they want to prepare to buy, right? We've played a pivotal role in helping them improve themselves as they navigate through that real estate journey. When you zoom out, you're seeing something quite powerful, right? This isn't just about search and upfront lead generation. It's search, connect, tour, apply, leases, payments, renewals, the entire transaction increasingly happening on our end-to-end engine. We talked a lot about the consumer, but it's really important to also shift focus and look at Zillow Rental Manager, or shorthand would be ZRM which is our supply-side product and offering, right? The best walkthrough of this kind of pivots around the three legs of the transaction stool in rentals, which would be applications, leases, and payments. Applications, you have tenant screening, right, on the supply side. Leases, you have lease builder and e-signing. Then payments, you have, you know, rent processing. Jumping into applications, again, within ZRM, landlords view potential tenants in one consolidated dashboard, right? This includes credit reports, background checks, eviction history, payment history, right? All of their income documents. This inherently makes the operator's job easier, more efficient, and gives them access to our superset of audience. Two, executing the lease, right? Landlords and property managers can either upload a lease that they historically have used, edit it a bit, set it up for e-signing directly through our platform, or they can go through our lease builder tool, which is, you know, compliant with local laws and regulations and navigates them through the journey of building a lease, right? Giving them confidence that they're building something that will be usable that one time, as well as being able to come back time and time again to use this product. Three, collecting rent. All of these steps are enabled directly on Zillow. Managing maintenance tasks, ongoing communication with the renter, everything within one connected system. Zillow stays embedded throughout the entirety of the journey. Look, most companies in this space are doing one of two things. You're either building a marketplace that drives audience and you're selling advertising, or you're building tools and services and selling them to the supply side, right? Kinda like SaaS products. We see an opportunity and have already begun executing on both, right? Tying together this large audience to best-in-class AI-powered tools and services. Efficiencies, productivity goes through the roof for the operators. Everything I just showed you is built under one end-to-end transaction engine. It really is the reason why we kind of have the permission to go forward with this strategy is because of the ROI that we know we provide to partners. This is the best-in-class, most efficient leads and audience. They are transaction-ready, and they are ultimately coming to Zillow to find their next home, whether it's a long-tail listing or a multifamily property. No longer are we in the business of just generating leads. We're in the business of generating outcomes. Obviously today, LLMs, right, and SEO for 20-plus years before that can do a very good job of summarizing listings, right? Providing that very top-of-funnel information, driving traffic to our platform. What happens next is the really exciting step, right? It's about creating efficiencies in the process, building agentic solutions that transform the way that people are engaging on our platform. This is where Zillow AI can do far more than just listing summary and top-of-funnel navigation. It's only possible because of the data that we have and this interconnected system that we've begun to build. Listing fragmentation was step one, right? We had to go forward and aggregate that superset of inventory. It opens the door for everything that we wanna do next. We're building trust with the consumer and building trust with the operators. We're generating highly efficient, high ROI audience that turn into leases. There are billions of dollars in annual rent flow across the United States, and participation in that transaction pushes our sights well beyond that billion-dollar mile marker I mentioned at the open, okay? We want to build something that ultimately is truly transformational for this category. Okay. That is a good deep dive on rentals. I look forward to connecting with everyone, later in the day, and we're gonna play a short video before welcoming Nicholas Stephens to the stage. Thank you. Good morning. Two weeks ago, I actually moved, like real life moved. I've moved a bunch of times, bunch of different apartments, different tech jobs. This is the first time moving with two kids and a cat, which frankly I would not recommend to any of you. After the last two weeks, we've finally started unpacking. I have more than four spoons, I have Wi-Fi, I have coffee, and I'm finally feeling the gratification of going through that entire process of figuring out, you know, what the next home for me and my family would be. I've worked at Zillow for over 10 years now. I joined way back when to be the Zestimate and personalization product manager, and when you get to work with all the AI teams, 300 people building the experiences that you are seeing today, you get a lot of chances to sit down with customers and see what's on their minds. Actually moving for the first time in a while was a more visceral reminder that all these questions they have from the very beginning of the journey all the way down to keys in hand, get a ton more real when it's you. You're not just asking academic test questions on what the neighborhood is like. You are trying to figure out if that neighborhood is right for you. Building AI experiences, we think about the entire mover experience, not just the left of shopping, discovering, thinking about what you can afford, but all the different questions, the guidance you need as you start thinking about touring, as you start to think about finding an agent, how to sell, wanting to message with these professionals, and more. From an AI perspective, it's a really tough challenge. A lot of customers in the same session start with what I would call surface level questions. They're wondering, "Is this new listing worth more than 10 seconds of my time? Could you describe it in a way that I understand? Could you give me an overview of this neighborhood?" These are important questions. They're not hard questions. If you have that content, a generic AI could do a pretty good job. You have to make sure it doesn't hallucinate, but, you know, it's important, you get the content, you do a good job at the surface level. In the same session, though, those same customers usually want to go pretty deep. They start at market overview and descriptions, and they end up going, "Hey, maybe I actually wanna tour this home. Maybe I need to get pre-approved. Maybe I need to connect with an agent," and more. This top to bottom dream has been our vision from the moment I joined 10 years ago, but it's an incredibly hard challenge because if you make up any of the answers, if you get anything wrong, if you give the wrong guidance, that's an incredibly trust-busting moment. You need the AI to join the customer where they are in the trenches and really make sure it's appropriately guiding and directing, just like any good confidant would. When I started thinking about moving with my wife six months ago, I of course used generic LLMs. Who didn't? I said, "Hey, I'm looking for a three-bedroom home in Phoenix under $450,000," and a generic LLM does a pretty good job. All those homes are real homes. It found them on the internet. It summarized them. It was pretty good. Of course, I went and toured some of these homes and had follow-up questions, and when I came back and said, "Hey, hold on. One of these homes, does this have a fenced yard? I have a 5- and 7-year-old, and we're not trying to run into the road here," the generic LLM goes, "You know, you're gonna have to be more specific. Which one are you talking about? And I'm gonna go search the internet and sites like Zillow to try and figure out the answer to your question." A generic LLM could do that surface level well. It's a great start, but as soon as life gets more interesting, as soon as you wanna go deeper, that's where it has to hand off to a company like Zillow. So AI with content answers the question. We spent 20 years amassing rich, interesting content. That is no small feat. AI with content, with context about the users and professionals, and with integration, end-to-end integration, that's what closes the loop. As we sit here and look at this iceberg this morning, many of our Zillow teams have been trying to figure out how can we build an AI-based experience that truly guides our millions of consumers end to end, from the very tippy top of that iceberg all the way to keys in hand. Today, for the first time publicly, I get the privilege of demoing a new experience, we call it AI Mode, integrated directly into Zillow that will do exactly that: help you understand if this home is worth more than a couple minutes of your time all the way to guidance right before you make that offer, right when you get those keys in hand. We're thinking, okay, how can we demo this to you today? It's this brand-new end-to-end experience. We're gonna show it to you at lunch, but, like, how do I show you all the cool things it can do? We thought, why not actually take three customers who had real experiences with AI Mode the first day we turned it on for them and show you their real prompts. As we go through Beth, Peter, and Cindy, we obviously made up their names, we changed their pictures, but every single question we go through this morning is a real consumer question in the first session of them discovering AI Mode on Zillow. Let's start with Beth. She's in Phoenix. She's a first-time buyer. She's rented a bunch. She's used a lot of the products that Sherman was just talking about, and she's thinking, "Is it time for me to actually buy?" She opens up Zillow, the desktop view. She starts looking at a home, and because we turned on AI Mode for her, she sees a bar at the bottom that says, "Hey, you can ask Zillow." The first thing she does is say, "Okay, does this home have any recent renovations or upgrades?" Good news, this home in particular has an updated kitchen, it has the bedrooms that have been updated, it has a finished basement, and then a Lennox HVAC. Apparently, in Phoenix, that's an important thing that all Beths really, really need to know. She proceeds with these types of questions. She's saying, "Hey, what year was the roof replaced? It's a newer house roof. What type of roof? It's slate. Exterior material, brick. Year built, 1938. It has four bedrooms, three bathrooms." She gets all of this in a couple minutes. Now, this is useful information. I'd say it's pretty surface level. I'd say I haven't talked to this Beth, but I bet she's just trying to test out not only is this home worth it to me, but if AI Mode can do a good job summarizing the home for her. Is it gonna make anything up, or is it gonna give real content in its replies? She gets all of this in a couple minutes, and then something interesting starts happening. With that confidence, her next question was, "What makes this home stand out?" The AI Mode goes, "Hey, classic 1920s brick bungalow, flexible bedroom layout, updated kitchen, covered patio, large lot with expansion potential." Now, not only is she trusting AI Mode in this answer, but if you look at those details, those are answers that only Zillow can give, given our incredibly unique content. It's not generic location. It's down the road from a school. This is something that you could only answer if you have seen inside the home, if you've amassed that unique content. Back to the iceberg. She started with renovations. She's kind of testing what year was the roof replaced, what makes this home stand out. She's getting near the watermark of our iceberg and beginning to think about, "Hey, could AI Mode really help me end to end?" Let's go back to this Beth in Phoenix. Next up, and this is truly in order that she asked, "Can I afford to buy this home right now?" Now, this Beth was a pretty active consumer on Zillow before AI Mode popped onto the scene. She is a BuyAbility user. That's a product that we built. Roughly 4 million people have registered with BuyAbility. They've given us their unique financial information like debt. We've figured out their credit score with them. We've given them a monthly affordability payment they can use to shop. When she asks, "Can I afford to buy this home right now," we don't go, "Maybe," or, "Why don't you go and calculate that?" We say, "Hey, given your BuyAbility, this is somewhat within your range. It's a bit of a stretch, but it is possible based on all the stuff you've already told us." With that, she goes, "Okay, can you show me homes within 4+ bedrooms within $450K?" I love this prompt because she literally wrote in parentheses, "My BuyAbility." Good news, AI Mode at Zillow obviously knows about Zillow's BuyAbility, but she wanted to make sure we included that. We gave her a rich map of about 580 homes for her to pan through and look based on all that information. Unfortunately, she thought to herself when she was looking through these homes, "They're not really worth me stretching. $450,000 is a lot. Four-plus bedrooms, I might not need that." She adjusted in her follow-up prompt and said, "What about three-plus bedrooms? What about $400K?" She got a lot more options, 1,600 homes, and it started to be a little bit closer to what she was actually after in real life. She asked, "What are today's mortgage rates?" Instead of us going, "Okay, we're gonna search the internet. We're gonna look at rates from a couple days ago on random sites. Good news, with BuyAbility and Zillow Home Loans, we pulled up-to-the-minute rates that backed her BuyAbility number and made sure they were displayed accurately in the response. Okay, she's done a lot in one session. It was pretty impressive when we were watching on the other side. In that same session, her next prompt was, "I want to tour this property." At this point, we said, "Get out the yellow pages. Go to 1-800." No, we're Zillow. We have ShowingTime+. 90% of homes in America are booked with real-time availability given ShowingTime+, and that is already integrated into AI Mode. She within that session selected Wednesday at 1 P.M. We said, "We'll confirm that live. Before you get into that home that you're zeroing in on, if you wanna go back to get pre-qualified, if you wanna update your BuyAbility, go ahead and do your homework right before you get to see that in person. First session, Beth in Phoenix, renovations and upgrades, all the way down to actually scheduling and confirming a tour and the beginnings of getting pre-approved. An incredible surface-level to end-to-end AI session. Next up, we have Peter. I particularly like Peter because he's considering moving to my hometown of Seattle. He's probably excited about us potentially getting an NBA team in the next 3-30 years. When he is looking to Seattle, in his first session, he started zeroing in on one condo, and his first question that we're gonna see live is, "What is this home's estimated market value?" It's one of the most frequent questions AI Mode gets to date. Over the course of three weeks, he doesn't stop there. On that one condo, he looked at others, but on that one condo, he's doing due diligence, he's doing deeper analysis, and he's even thinking about his own lifestyle. What are the grocery stores nearby? What's the actual neighborhood like? He's basically getting a Ph.D. in this condo before he makes an offer. Let's actually see him live. His first question, as promised, is, "What is this home's estimated market value?" Remember, I joined a decade ago to be the Zestimate product manager, and at the time, we talked a lot about natural language processing and neural networks and the ability to see in homes. We didn't have the technology to actually have a conversation with the Zestimate. Consumers see that number and they go, "Why? Why is it that high? Could it be higher? Could it be lower? What, what produced that number?" Now, with AI Mode, we can actually have that conversation based on all the data that Zestimate uses behind the scenes. Because we told him about all the different factors, he turned into basically a Zestimate scientist and started asking questions like, "Does this other condo have any recent upgrades or renovations?" That's obviously something that helps the Zestimate going forward. His next question was, "Does this condo include a pool or a clubhouse?" In Seattle, Washington, if you haven't been there, pools are kinda rare, and so that does add some value to a condo, and he was wondering, "Hey, do other condos nearby also have a pool or a clubhouse? Is that gonna increase the market value of similar condos?" Then week three, this is when he gets interesting. He goes, "Hey, are there any HOA fees associated with this property?" We say, "Yes, as traditionally true with most condos, roughly $275 a month." "What about utilities? Are any of these included?" Oh, sorry. They aren't included according to AI Mode. In the HOA dues, you'd be responsible for paying electricity, water, and gas separately. This is a real prompt from Peter: "Actually, I'm pretty sure utilities are included in the HOA dues. The current tenant mentioned that." Zillow already gets this kind of feedback from consumers, but usually it's in a device kind of help center experience. Now, with AI Mode, we can take that information, obviously update it on our systems, but more importantly, tell the seller, tell the seller's agent, tell the landlord that, "Hey, some of what you're posting online might not be accurate. Do you wanna fix that so you can better market your property to the open market?" Okay. Peter has seen enough. It's probably week two or three at this point. He's going, "Okay. I'm ready to move to Seattle. I'm getting excited about this location." He starts shifting to lifestyle questions. Are there any highly rated restaurants nearby?" Canlis, my favorite if he's celebrating, is number one, all the way down to Pike Place Chowder. He goes, "Hey, how far is the nearest grocery store?" If you haven't been to Seattle, Pike Place Market, number one, but let's be practical. Safeway and Whole Foods also down the road. Third, "What points of interest are nearby?" Again, if you're new to the market, we go, "Hey, everything from Gas Works Park and Woodland Park Zoo all the way to UW is just down the road." Last, but certainly not least, "What are today's mortgage rates?" It's an incredibly frequent question from buyers who are near that waterline thinking about going end to end. Of course, because of BuyAbility and Zillow Home Loans, we give him an up-to-the-minute view of what kinda rates he might get if he gets pre-qualified. Peter is no longer searching, he's deciding. Over those three weeks, we learned a bunch about Peter. First, we learned all the traditional things that without AI Mode, we usually learn from saves and shares and clicks. We learned his target price, we learned what kind of neighborhoods he's interested in, some basic level features. With AI Mode, we learned a lot more. We learned what matters most to Peter, we saw rich behavioral signals. We know how ready he is and what the qualification for the best condo for him is in real life. If you're like, "Okay. Well, that's just useful for Zillow," no, that's useful for Peter. He asked 17 times about that one condo. If we woke up on, you know, 16th prompt and go, "Hey, what's your name? What are you asking about? What are you trying to do?" He'd be like, "I don't wanna use AI Mode." Because we save all that rich context, the next prompt, we're able to guide and point him in the right direction as he proceeds down his path. Okay, so we've covered Beth, we've covered Peter. Third, but certainly not least, is Cindy. Cindy is one of my favorites because I kind of identify with her in my own home search. She did a lot of rich comparisons that only Zillow can do, and I really wish I had AI Mode when I was looking. Cindy's all the way on the other side of the United States. She's in the Panhandle, she's in Florida, and she asks a whole bunch of questions as she discovered AI Mode in her first session. She's panning around in Florida. She opens up this home that has a pretty amazing covered pool, and she starts going, "Hey, I love this pool, but it's near a busy street. How big of a trade-off is that?" AI Mode goes, "Yes, the pool is strong. The street honestly is not that busy. Didn't really decrease the Zestimate as much." Her next question is, "Okay, this one, this other home has a higher Zestimate, but honestly doesn't look as nice. What is driving that difference in Zestimate?" Again, now that you can actually have a conversation with the Zestimate rather than a single number, we start breaking down in AI Mode what is actually producing that different number. Next up, is the location doing most of the work here? Again, she might be applying to Zillow to be on our science team, but or just understanding what is providing that number. Next up, this one checks almost everything, but is priced high. Am I stretching for the features? I love the, "Am I stretching for the features?" She's starting to ask for actual advice. She's relying on AI Mode to lay out what's in front of her and help her reason through it. She was looking through this and kind of getting excited about this specific area. Next up is really, is there any near grocery stores where she can actually be able to walk to versus have to ride the bus? Second to last, in the same session, "Do you think I'd actually use all of these features?" Last but certainly not least, "If I stay 5-7 years, does paying more here make sense?" That is an incredibly trusting first integration or first interaction with AI Mode. She went from, again, surface-level questions that any good generic AI LLM should be able to answer all the way down to stuff that only AI Mode can do authoritatively. Next up, she was wanting to compare two of the homes you saw flash in front of you. Anyone can make a good comparison table. I did this in Python in CS 101. What's unique to AI Mode is that how we are laying out is based on all the context and all the content that we have unique to her. We're saying, "Hey, Wintergreen offers more space. The layout is superior in one of these. Parking, that's important to you. There's one car garage in the Galino. Condition is move-in ready. The community, basically HOA fees are the same between each and every home. After all of this comparison, AI Mode on its own accord says, "Look, Cindy, sit down for a second. I need you to tell me between two clusters of home what matters most to you." Basically, AI Mode saw all the ones in her criteria set and said, on the left you can either go garage, gated, slightly newer, or on the right you can go extra bedroom, lower total cost, brand-new appliances. There's a fork in the road here. You can't have it all. If you tell AI Mode which one, Cindy, that you think is more important to you, I will be able to help you navigate which one to actually tour in person. Okay, we're gonna pause on Cindy. I see some of you writing notes. I'm gonna tell you, no spoilers, it works out well for Cindy. We're gonna pause here because I wanna talk about rich media before her next question. Rich media is something that Jeremy and Cynthia mentioned. It's been a 10-year journey at Zillow. We have been working hard to get the best 3D experience so that buyers, sellers, and renters don't really have to figuratively imagine what a home is like. They can see it from the comfort of their current home and really think about, "Hey, is this home worth it for me?" You can fly around outside homes. You can obviously zoom inside. We have the new Sky Tour product we mentioned earlier, where you can really get a sense of the neighborhood. More recently, if you want to reimagine the home, you can virtually stage it from all those rich listing images. All of this is not easy in AI land. You can do a pretty good virtual staging with a generic LLM, but usually that's about 60%-70% accurate. It's gonna imagine a new bedroom. It's gonna replace a wall by accident. It's gonna forget about your couch. If you are in real estate, you can't do any of that. That's a trust-busting move. When we do virtual staging, we have to make sure the models are 100% accurate. Every change is something that actually honors the room geometry and enables that buyer to really imagine herself in that home. Now, there's a bunch of stuff we've been doing more recently. We've vastly improved the realism and image quality of the indoor views when you're flying around the home. This is no small feat. I don't know if you've ever tried to take an indoor photo with your iPhone. To try and get this kind of quality both inside and peering out the window took a bunch of awesome science, and we're pretty proud of the result. The thing that we're gonna actually debut for the first time here is how we're improving the actual capture process of building our floor plans. Our floor plans are one of the most popular feature on Zillow. This took me by surprise. I thought they'd be interesting especially to click around and see different parts of the home. Actually, this is something that gives buyers and renters a deeper sense of how everything is laid out. For me, in my old home, my bedroom was right next to the kids' playroom, and I realized that was not smart as the kids scaled. When I was looking at new homes, making sure those two rooms were not right next to each other, super important. The opportunity here is that we can make the capture process for photographers easier. Today, if you're one of the 10% of listings that have rich media, that required a photographer to get a 360 camera, put it on a tripod, put it in the room, run out of the room, take a quick picture, run back, grab it, and have that capture upload to the cloud. It's a pretty great process. It's not that hard, but it still means running in, running out, waiting for feedback as it processed in the cloud. Today, for the first time publicly, we are debuting what we call Instant Floor Plan. This is something that we can demo at lunch, so if you think this is a made-up video, it's not. This is an actual Portland listing. The photographer in this case is using their own iPhone Pro, any traditional iPhone and starting at, in any room, pointing at the corner, and then proceeding to wave that phone around to capture in real time. What's awesome about this is twofold. One, it's accurate. It's actually identifying the corners, the room geometry accurately. Even more important, it's giving that feedback as the photographer is in the room. It's saying, "Hey, you might have missed a corner, so go back and make sure you point your iPhone there." Then at the same time, it's building that 3-D model on the iPhone, so the photographer can pan around and make sure everything is looking exactly as he or she sees it in the home itself. Again, if you think this is made up, we have this running on iPhone. Most homes don't have ballrooms, so it'll be interesting to see how much it captures of the full room. Please take a look at it. It's pretty awesome to see in real life. All right. We were wondering about Cindy. We said, "Hey, let's take a detour to Rich Media. Let's get back to her." Her next question, not making this up, is, "Does the kitchen get natural light?" Generic LLMs would say, "Maybe," or, "Why don't you look at the listing images?" Or, "Maybe you can Google search it." Because we have the densest and most awesome set of rich captures in the nation, AI Mode was able to go, "We have south-facing windows in this home, open sight lines. A window placement over the sink would give you that natural light that you are in the mood for." Okay, we've covered a lot of ground with our three customers, and really, I hope that you're leaving with the sense that Zillow sees that full arc. AI Mode needs to be good at that early stage. Is this home relevant to you? Do you wanna save it? Do you wanna view the 3D tour all the way over to actually booking that tour, getting pre-approved, and hopefully closing? It's this full arc that gives us a sense of where customers are at buying, selling, and renting at ridiculous scale. We already can predict if you are likely to sell because you are hammering on the Zestimate, you're thinking about renovations, you're looking at homes. With AI mode, that gets even richer because we are having a natural conversation with where you're at, what you're in the mood for, what you're trying to do as you think about your next home. That wheel is a pretty stunning one. If we start on the top left with high-quality media, all those rich media captures, basically being able to fly around the home, outside, inside, and then also look at a floor plan, that drives behavioral data. That drives more Beths, Peters, and Cindy to the site to really deep dive in every home they might be interested in. That deep diving helps AI mode understand what actions should they take next. Do they need to filter down? Do they have a left and right decision? Do they need additional guidance? That guidance drives more agents to our platform and essentially helps them show up at the right moment to be that superhero agent when they need it most. Before I get off stage, I wanna return to these three customers. You're probably wondering, "Okay, Nick, you're a product manager. You took the three best customers from day one, and they used it, they asked all the right questions, and then they probably never used it again." I would say two things to that. First, we're hiring product managers at Zillow. That is a fantastic question, so come see me after. Two, I checked on all three of these customers on Friday to see how they were doing. Were they using AI Mode more than once? Good news, Beth went from those 40 initial conversations to 116. She's getting pretty excited about 8 properties. She's done 11 what I would call deep dives on financials and what she can afford. I think she's the closest to actually making an offer. Peter, he's got a PhD in Seattle condos at this point. 45 conversations, four distinct properties. I think he's about to make an offer on that first one. Cindy's ready to compare every home in Florida. She's had 176 conversations, 24 properties. She's been using AI Mode 14 out of the 14 days we've made available to her. 43 comparison conversations in the process. All right, this is a tour de force of AI Mode. We started at top left, basically shopping based on affordability, thinking about touring, finding your agent. I didn't get all the way to the right because at that point, Cindy, Peter, and of course, Beth, need a real professional to help guide them across the finish line. What better to keep that story going than introducing my coworker, Cameron, who will tell you about the professional side of that very same journey. Thank you. Good morning, y'all. I would have hoped I'm more than a coworker to Nick, but alas, here we are. Nice to see you all. My name is Cameron Swiggett. As Nick mentioned, I am going to be talking about the professional side of this equation. We're gonna do a bunch of demos walking you through a day in the life of what it looks like to be an agent, particularly a real estate agent in this new AI-powered world. The thing I love about real estate is every single one of you in this room can relate to real estate on some level. You've either bought, you've sold, you've rented. You are likely one of the hundreds of millions of uniques that come to Zillow and scroll and search and dream and shop, right? I think the thing that is perhaps less readily apparent to people who aren't deeply in the industry is just how much of the consumer experience is actually driven by the real estate agent. Nicholas teed me up a little bit there. As soon as shoppers start to get pretty serious, whether they are about to schedule a tour, about to make offers, transact or close, or if they're on the sell side, if they're about to list, they're receiving offers, transacting or closing, an agent becomes the driving force and quite frankly, an extension of their consumer experience, as does the tools in which they operate. They are the engine for that consumer. I would argue there are a lot of engines out there. There are 1.5 million licensed agents in the United States, and I am not going to make you all raise your hands, but I bet you if I asked how many of you know someone in your life, a friend, a cousin, an aunt, a brother-in-law, who's a licensed agent and hasn't done a deal in a year or two, a lot of hands would shoot up. We see that in the numbers. It turns out the long tail of licensed agents is rather long. Said another way, 20% of the top agents in America drive 80% of the transaction volume. 20% drive 80% of the transaction volume. Which is now at the point I want to address the elephant in the room, right? We are here under the guise of the AI Investor Day, and we are talking about AI, how AI is inevitably disrupting so many industries, revolutionizing so many industries. We all read the same headlines. There are probably some of you in this audience are wondering, is that 20% gonna go down to 0%? Could a synthetic agent transact at scale for all of America? I quite frankly understand the intuition. I think it's a fair question. I would personally bet that is not gonna happen. I don't think that happens for a couple reasons. The research and the data informs that decision. But also, if you just take a step back and think about what it looks like to transact for the average American, this is a highly emotional thing that they're doing. Literally more than half of them cry during the transaction, and it is the most financially significant thing that they will do in their life, whether they're buying or selling. It is the vast majority of their net worth, and they have told us and we have seen that they want a human to hold their hand and to guide them and to coach them and to give them the confidence that they are doing the right thing for themselves and their family. That said, I'm not blind to the fact that AI will, has, and will continue to change real estate. The question to me is not does 20% go down to zero. The question is, how do we get more leverage and more scale out of that 20% so that they perhaps gobble up 100% of the transaction share? You could do the other side of the equation. Could fewer than 20% still maintain the same ratios of 80%? I think the answer to that is yes, and I believe that Zillow is uniquely positioned to make that happen. I say that for a few reasons. One, we have all the top-of-funnel advantages that Nicholas and team have talked about, the hundreds of millions of uniques that come to Zillow every single month, the context, and that brings massive advantages for our agents, of course, but also the top 20% who are driving 80% of the transactions, they're already on our tools today. They're in Showcase, they're in Zillow Pro, they're in dotloop, which touches more than 50% of transactions. They're in ShowingTime, which sees more than 90% of tours, and perhaps most importantly for this conversation, they're in Follow Up Boss. 41 of the top 50 teams, real estate teams in the United States, are on Follow Up Boss. These are the 20%. These are those who are pushing forward the most amount of customers, be that on the buy side or sell side to their successful goals. I figure, why don't we do a little bit of a day in the life of an agent, so we could see how we are infusing AI at every single touch point to make these agents even more efficient, so that they can offload some of their work to AI, so they can concentrate on the human connection, the thing that these clients are clamoring for, and they can delegate the rest to AI, so they can take on more concurrent transactions. I am making an assumption, as much as you all know about real estate, and you may love real estate, I'm assuming that none of you have been a licensed agent before. Even if you have purchased, you have not been behind the scenes and seen what an agent's day-to-day looks like. We're gonna do just that. We're gonna do a little bit of a day in the life of an agent. I should caveat by saying there is no typical day for a high-performing agent. They have chaotic, whirlwind-type days. In this fictional example, we're gonna talk about this agent servicing seven concurrent clients and then juggling hundreds, if not thousands, of people in their database that are gonna be kinda coming in and out of potential viability. All right. First step, agent wakes up just like you and me, brews a cup of coffee, and the very first thing that they're going to do is they're going to open up their CRM. They are going to open up Follow Up Boss. The reason why they do that, this is effectively their little black book. This is the place that they are getting information on what they need to do next, which clients need to be actioned on, what tasks are readily available for them to push the ball forward, who is the best potential person for them to convert into either an offer or a listing opportunity, right? This is ultimately how agents make their money. Now, I'm sure many of you have not seen the insides of Follow Up Boss. We're gonna walk you through the new version of Follow Up Boss, the AI-powered Follow Up Boss. So first things first, this is called Action Hub, and you will see here, no longer is this a unorganized, untriaged list, but you now have a fully stack-ranked opportunity list here. At the very top is an opportunity. We've got Ryan here. Ryan is a client that this agent worked with months ago. This individual said, "I think I'm knocking on the door of getting promoted, and when I do, I want to go ahead and start making offers." Well, their Zillow history is lighting up like a Christmas tree. They are obviously very active right now, signaling to our AI and Follow Up Boss that they probably got promoted. AI summarizes their activity. It generates a message, very actionable, personal, curated for that agent, allows the agent to edit it if they want, send it, they're off to the races. Next individual, Priya. Priya is a new tour request coming from Zillow. Never met this individual in their life, yet all of their information from Zillow is starting to flow in because of a My Agent relationship, which we'll talk about in a minute. Once again, even though I never met Priya, this agent now has a better understanding of the hopes and desires and aspirations that she has as it relates to real estate, generates a message, sends it. Finally, we have David. He is an active client. They've made an offer. An inspection report has come in. AI has analyzed that inspection report, surfaced a couple findings, generated a message for this agent, and then they are now scheduling a time to review that inspection report. I wanna talk about two things before we move on here. One, that last individual, David Park, the inspection report, that was a super simple task. Remember inspection report or David Park, we're gonna come back to that at the end of this demo 'cause behind the scenes, there's a ton of complexity to surface that what seemingly looks like a simple AI task. The two clients, you had Ryan and Priya, the thing that made that so insightful was the fact that we were able to surface all of the Zillow activity. These were not just cold leads. These were deeply insightful, actionable leads because of the activity that they were generating on Zillow, a lot of which Nicholas kind of walked you through a little bit ago. That is possible because of something we call the My Agent relationship, and that is formed on all Zillow leads, creating that bond between the client and the agent, allowing that symbiotic relationship where information is being passed forth, and Follow Up Boss, our CRM, go ahead and summarizes it. This notion of a My Agent relationship, while it happens automatically for Zillow leads, is also now available for any lead that this agent finds. If they're in the coffee shop, if they get a referral, if their cousin wants to work with them, they're able to send this My Agent relationship link, which you can see right here. As simple as one click within the CRM, it's prompting you to what that activity starts to ultimately look like when this accept occurs on the client side. There's a message that AI drafts, which of course has been fine-tuned to make sure that it is the most acceptable for that client. The client receives these terms, one click, and once they've done that, what you'll see is their customer profile in their CRM updates automatically. What have they been looking for? How has their search evolved? What are the types of homes that really pique their interest? What are no-gos? Do they really need an attached garage? Do they, for some reason, not wanna live in an HOA? What are the specific homes that they've been honing in on for the longest? Have they come back to them multiple times? Have they hovered on certain photos for a long time? Have they shared them? Have they saved them? This is massively insightful information for an agent 'cause it's effectively riding shotgun with them the entire time, learning their habits and allowing you as the agent to anticipate their needs potentially before they even know about it. All right, our fictional agent, he's had his cup of coffee. He's blasted through the top leads and tasks in his CRM, and that may have in the past either been not as fruitful or would have taken hours, but he's done this very quickly because of AI. He's about to walk out the door, get in his car for the very first tour request, very first tour of the day, and he gets this notification from Follow Up Boss. One of his clients, a somewhat dormant client, has now requested more financial information to ensure that they are ready to start making offers. What does this agent do? Before he gets in the car, he passes over all his AI summary information. The loan officer or ZHL loan officer digests that, adds that to the system, and within minutes creates multiple curated personal loan scenarios as well as a recommendation for that client. Our agent passes it back to their client, and this happened from the span of them opening the front door to getting in their car. In the olden pre-AI days, which was, like, six weeks ago, you would have gotten back in the house, whipped out a laptop, done a bunch of work. The loan officer would have done a bunch of work, and in best-case scenario in the afternoon, this client would have had that information. This is the kind of thing that's happening at scale, allowing this agent to continue on their day to handle more transactions and more volume. Our agent gets to the house, and now he's faced with a, I would argue, a two-part awkward scenario for buyer's agents. One, likely never been to this house before in his life, maybe never been on the street, never been in the neighborhood, has no idea what this house is about other than what he's seen on Zillow. He's walking in with the same amount of information as his client, but he is the real estate professional. He is supposed to be able to give guidance, right, on what to do. The other, also somewhat awkward, either has never met this client before or it's been months, and he doesn't remember 'cause he's been shaking 500 hands in between that time. Follow Up Boss is there for him, sends this notification, "Hey, you've got a tour in 15 minutes. Why don't we familiarize yourself with both the home and this client?" Opens it up. All of a sudden, what has this client been looking for in the past? What are things that have been of interest to her? How has her search evolved? What does her affordability look like or BuyAbility? Things that she has flagged that she loved about homes. She loves the fact it's in a specific school district. She loves that it's had a renovated kitchen. She really does not like the fact that it was either missing a garage or whatever, right? Also, by the way, this home was built in 1938. It's a little bit older. Maybe as the real estate professional, you should ask about things like maintenance or the home or the roof rather. Now this agent knows both more about the home, more about the client, and they go on, and they have their successful tour. He gives the coaching and the guidance that she's looking for. At the end of it, they have a conversation. What are we gonna do next? We are going to think about it overnight and potentially make an offer the next day. Follow Up Boss prompts the agent before he gets on the way back to the office. How did the tour go? He voice dictates into Follow Up Boss everything about the client, everything she said she liked about the home, everything that she wants to do next. Takes that voice note, synthesizes the voice note, and then updates her profile in his CRM, ensuring that we have the appropriate description of what her hopes and desires are, and it creates a discrete task. She flagged that she wants to think about it overnight and potentially make an offer. When he wakes up the following morning with his coffee, he's gonna have some at the top of his list, follow up with his client, and oh, by the way, AI, can you prep some offer documents just in case that she's ready to make that offer? Once again, reducing the mental load for this agent. All right. Our agent, he's back on the road, headed to the office because he's got more work to do. He gets to a stoplight, phone buzzes, and he has a new tour request. Great. He wants to act on this as quickly as possible to make sure he's locked this client down. Fortunately, AI's done all the work for him. Because of the ShowingTime integration, because of integration with his calendar, because we understand what the client's needs are, done all the conflict resolution, suggests at a time, pre-booked it, and then surfaced it while this guy is sitting at a red light and then says, "Hey, does this time work?" He has confidence this time's gonna work 'cause the AI has done all the work for him. A simple click as, "Yep, confirm. We're great." Notification back to the client. They're ready to meet Wednesday at 10 o'clock. He's now gonna legally and safely drive, and he gets back to his office. Once he's back at the office, I want to transition us a little bit here. I'm gonna harken back to what Nicholas talked about too with this listing soon model. You'll note that all the work that this agent has done to date has been all buy-side focused. They've done tours, they've done offers, they've done financing, right? They've done all the things that a buyer would need of them. Well, the top 20% who are doing 80% of the transaction volume have a diversified business. They do buy and sell-side, typically. I wanna show you how that predicted, that listing soon model, which by the way, this is a visual representation, would actually show up and manifest for an agent. Our agent is back at the office, he's on his computer, he logs back into the CRM, and what does he see? AI has discovered an opportunity for him. There's two AI signals have converged. A lead that occurred many, many moons ago with somebody saying, "Look, I've got this vacation home in Tempe, Arizona. As soon as this market starts to look like I can actually move this home, let me know, and I'm happy to list it." Well, great news. Zillow Insights are informing us that the market's getting hotter, and by the way, this client's already kinda beat us to the punch. His Zillow activity is flagging that he is, in fact, interested in looking at comps and listings. This now surfaces to the agent, and the agent goes, "Hey, by the way, bud, great time to start listing. Let's go ahead and get a booking on the schedule to talk about what our listing pricing strategy could look like and when we wanna go to the market." AI generated that message, he sends it, and he is now working on what would have previously probably been an undiscovered lead. AI is constantly looking in the background for those opportunities for him. All right. Our fictional agent, he's blasted through his day. He looks down, he goes, "Oh, shoot. My little kid's got her soccer game." He drives over to the soccer pitch, and he leaves his phone, his laptop, and his iPad in the car, and he's in pure dad mode. He wants to be the best parent he could possibly be. This, to me, is the greatest opportunity that AI has to allow us to scale to these agents to new heights, to be able to handle even more transactions. Because while he's being that good dad, AI is not a parent. AI does not have to watch a soccer game. AI does not have to eat or sleep, and it will work in the background for him. What I am kinda touting is this notion of a virtual transaction coordinator or a virtual listing coordinator. These are real-life humans that people employ, but this can be done at scale with AI. I wanna walk you through what we've been developing in a platform that we are calling Zillow Workspace. Zillow Workspace, you can think of it as the grounds in which we're putting the best of dotloop and ShowingTime and Aryeo and listing management, you know, document management, transaction management, the kind of deeper funnel back-end office tools that an agent needs in their tool chest to be able to complete transactions compliantly. He has this all now in Zillow Workspace. Oh, by the way, it's now doing things agentically on his behalf. Remember that listing soon opportunity in Tempe, Arizona? Why doesn't AI go ahead and generate all the collateral and assets and materials for him? It generates a pricing report, a comparative market analysis report, a listing price strategy, presentation materials, all things that he can then share digitally in advance of this meeting, or he can print it and sit down at the kitchen table and represent why he is, in fact, the best individual to list that home, and he has the best data informed by Zillow and the insights into how they're gonna have the most successful outcome. All right. Remember we talked about. I said the inspection report, right? There was a super simple task item that showed up as David Park in the very beginning. Well, here's what it actually looked like in the background. This is a very complicated, robust document understanding model that we've created. What happens is the PDF came in, that program then classified it as an inspection report. It digitally read all the information on the 47-page document. It extracted all the information. It put it in the transactional system of record as actionable activity, and then it surfaced insights, things that this agent should take action on. Now, what does that do? The agent doesn't have to read 47 pages, which is a great start, saving them time. Also AI, after all the decades of experience that we've infused it with, is giving really robust insights into what is the appropriate action to take on behalf of his client. That, of course, showed up as that super simple line item in Follow Up Boss, but you're seeing in the background how that actually came to fruition. Now, an inspection report is one complex document, but real estate is full of many, many complex documents, and we've built this model to be highly flexible and real estate specific. Let's do another example here. This is a listing agent receiving a bunch of offers. Those offers come in. Similar game, right? It gets classified as an offer, it gets programmatically extracted, all the relevant bits get put into the transactional system of record. Once again, this is licensable activity. The great thing is, at the end of it, the agent knows the key terms. They don't have to read the 20-page document. They know what matters. AI can also generate a really simple consumer-facing offer comparison sheet, something that they, as the agent, can put in front of the client and say, "You got four offers. Here's the thing that actually matters the most for you. Here's my recommendation. Which of these would you like to go forward with?" This document understanding, if you go use a generic AI, whatever one you personally use on a daily basis, you will find if you shove a bunch of PDFs in it does a reasonably good job at document understanding. What we've done is we've taken that great foundation, and then we've just added so much of our expertise and transactional knowledge on top of it to create a housing-specific version of it, which is flexible and scalable and derives really amazing insights on behalf of our clients. I'm hoping that you've seen now a little bit of the day in the life. I know we typically talk a lot about the consumer, but I wanted you to learn more about the agent a bit here. You've seen two things. One, just how chaotic their life can be as they're juggling numerous clients and trying to win business at any given moment. Also, where AI plays a pivotal role at literally every stage of the transaction life cycle. It is providing them scale and efficiency to allow them, if they're part of that 20%, like we have in our tools, to scale themselves and to handle even more transactional volume. What I hope, after listening to Nicholas's and mine, is that you see the harmonious relationship that both the real estate professional and the consumer have together. What I truly hope is that you see the commonality between all of that, of course, is Zillow sitting there at the middle. It is the glue between the consumer and the professional, and AI is simply making that glue that much stronger. For now, I'm going to wish you a good morning. I'll be in the Q&A session. We could talk more about lunch. I'm gonna welcome up Jeremy. There are two Jeremys at this company, of course, but there's only one Jeremy Hofmann. Jeremy Hofmann, all right. All right. I am really happy to be here. I have been excited about a Zillow Investor Day for a long time. I got less excited when I got smushed in between Cameron and Rich Barton. I will do the best that I possibly can to keep this interesting after all those awesome demos and before Rich talks about crossing the chasm. I'm thrilled to take you all through how all of this great product work ultimately translates to a great financial story. Before I get into the numbers, I wanna remind you all about the financial philosophy that we have that has underpinned my three years as CFO here at Zillow. First and foremost, we have to have strong revenue growth driven by superior products and services. Just as importantly, we have to control our fixed cost base to drive high incremental margins. With that, we will gain meaningful leverage on our path to strong GAAP profitability. We will use M&A as an accelerant to our growth strategy. Last but not least, we will be opportunistic with our share buyback to control dilution. That financial philosophy underpins the strong execution that you all have seen from us over the past few years. First and foremost, our revenue is growing well, and we're consistently outperforming a challenging housing market. Let's look at what we've delivered. We have grown revenue in aggregate since the start of 2023 by 43% and consistently outperformed the housing market, which has obviously been challenged. In 2025, we grew revenue 16% year over year against a housing market that grew roughly 3%. Thirteen hundred basis points of outperformance for the full year. In 2026, we expect to outperform the market again as we continue to execute on our strategy. While we've been consistently outperforming and growing revenue the way that we have, we've also meaningfully diversified our revenue base. On the rental side, we have accelerated growth as we have leaned into multi-family buildings. Our rentals revenue has grown 130% since 2022, and we have nearly tripled the number of multi-family properties that advertise with us from 28,000 back then to more than 72,000 as of the end of Q4. On the for sale side, we have grown revenue nicely versus the market over this time. We've introduced Real-Time Touring, Enhanced Markets, both of which are resulting in better conversion and more transactions. We have focused our efforts on Zillow Home Loans, becoming a consequential purchase lender, increasing our purchase originations by more than 6x over this period of time. We have introduced and expanded seller products, including Showcase, which is now on nearly 4% of all new listings in the country. We have acquired and grown Follow Up Boss as part of our expanded agent product suite. Follow Up Boss is being used by effectively all of our preferred partners today, and you saw all the awesome information and demos that Cameron showed you, which I think is among the most interesting stuff that we showed today because you don't necessarily get to see the agent workflow in your day-to-day, but you see how powerful Follow Up Boss has been. We've also expanded our new construction business to deliver more new home inventory to more consumers. Back to Jeremy's content-context integration. That content for new construction is quite important alongside existing homes for sale and rentals. All of this contributes to the sustainable revenue growth that you all have seen regardless of market conditions. We aren't just delivering consistent revenue growth. Our earnings are growing faster than revenue as we continue to expand margins. The way we expand margins starts with diligently managing every part of our cost structure. We split our cost structure into three distinct buckets. First is fixed costs. Our fixed costs as a percentage of revenue at the end of 2023 was 80%. Fast-forward to the end of 2025, the number is 64%. I'll dive in further on the next slide to take you through how we've been able to accomplish that. The second part of our cost structure is variable costs. We have been investing in variable costs ahead of revenue to support the rapid growth that we have seen in rentals and Zillow Home Loans, which collectively grew 83% since 2023. This period of forward investing in variable costs has been the right strategy to support the growth of these nascent businesses, and we expect variable costs to trend closer to in line with revenue in the second half of this year. The third bucket is advertising. We will dial up and down advertising depending on the opportunities that we see in a given year. For example, we stepped up marketing from 8% of revenue to 9% of revenue in 2024. That was to support the growth in rentals and make a clear campaign around Zillow Has Apartments. In 2025, we decreased marketing as a percentage of revenue based on the conditions we expected throughout the year. In 2026, we expect marketing to step back up as a percentage of revenue given the opportunities we see across our products in rentals and for sale. Just to give you a sense of how we dial up and down marketing depending on the opportunities we see, the last three years and 2026 gives you a good flavor for that. I do wanna spend more time taking you all through how we've been able to manage our fixed cost base from 2023 through the end of 2025. First and foremost, we have been consistently saying that we want our fixed costs to grow with inflation, and we'll fight that inflation where we can, but grow with inflation has been a place where we thought we could drive a lot of leverage in the business and continue to make investments along the way. I am pleased that on an organic basis, we have grown our fixed costs in aggregate since 2023 at 5% versus inflation of 6% over that period of time. We have also added fixed costs through four acquisitions we have done since then, including Follow Up Boss, Spruce, our closing service, Aryeo, and Virtual Staging AI. Underneath this fixed cost control is a lot of prioritization and discipline that our entire team believes in. Our entire senior leadership team believes in this discipline, and we go through it every year during our annual planning. For example, over this period of time, we have decreased fixed costs across corporate functions and overhead by 5%, and we have redeployed these savings into product and development, primarily in rentals, mortgages, and most importantly, in our AI efforts. These AI investments include talent, software solutions for our employee base, development costs, tokens, and many more. Additionally, we reduced stock-based compensation over this period of time by 14%. Using savings to reinvest in highly strategic initiatives while also gaining leverage to expand margins is a winning combination for us. Like I said before, we start every year challenging each team to reduce our fixed cost base. We make decisions on what we will deliver to the bottom line versus what we will reinvest into growth initiatives. You have seen our ability to expand margins and expand profits faster than revenue. We are proud of this, and we are also proud that when we do reinvest, we are shipping at a faster pace than ever before. The innovation cycles at this company, when powered by AI, have been impressive. We are shipping 40% more code per engineer. We are shipping products faster than we ever have before. One great example amongst many is the Smart Summaries you saw Cameron show earlier, where those text messages showed up for an agent. We went from concept to launch of that feature in less than a week. The opportunity to digitize and integrate residential real estate is enormous, and we are just scratching the surface. Of course, this is all in broader service of our broader mission, which is to make real estate more efficient by integrating renters, buyers and sellers with the industry professionals that serve them all inside our housing super app. This strategy and execution has allowed us to perform quite well over the past few years. We have grown revenue by 33% from 2023 to 2025. We have grown EBITDA by 59% from 2023 to 2025. We have grown adjusted free cash flow by 71% from 2023 to 2025. We've been actively managing our capital along the way. We have consistently delivered strong financial results, and share buybacks have become a key part of our capital management strategy and financial philosophy. We will continue to be opportunistic with our share buybacks, especially in times of market dislocation. Q1 2026 is a good example of this. We are very, very confident in our go-forward opportunity, and we leaned into share buybacks accordingly, repurchasing over $625 million of stock in Q1. Going forward, we will continue to be opportunistic as we grow our earnings and actively manage our share count. With respect to 2026 financial goals and mid-cycle financial targets, I'll take you through each of those. Our 2026 financial targets are unchanged. We expect mid-teens revenue growth. We expect continued EBITDA margin expansion, and we expect stock-based compensation to be down more than 10% year-over-year. With respect to our mid-cycle financial targets, our goals of $5 billion in revenue in a mid-cycle environment and a 45% EBITDA margin are unchanged. Importantly, today we are adding a 25% net income margin target to our mid-cycle financial targets. We are adding this target because of the confidence we have in our path to strong GAAP profitability. Our execution over the past three years has been strong and we expect it to be going forward. While we have been executing well, there have been a number of key debates that have emerged around our company over the last three years. I wanted to take some time to check in here on the status of those key debates. Debate number one. For those of you that remember, this was fall 2023 type timeframe. Buy-side commissions will rapidly decline, causing Zillow to be disrupted. Our perspective at the time was that independent representation has value. There will be differentiation based on quality of real estate agents, and we work with the best agents. Fast-forward to today, that has played out as we expected. Independent representation has proven its value. There will be differentiation based on the quality of real estate agents, and we work with the best ones. As a result, Zillow Preferred Agents have not seen any impact to their commissions because they are the best of the best. Debate number two: A large competitor enters the category with significant investment spend, causing Zillow to be disrupted. Our perspective then: Our brand strength and our unique strategy are differentiated, and we will continue to grow. Our perspective today: Our brand strength and traffic have both increased since 2023. We have added $638 million of revenue since that time, with rentals growth as a clear standout, doubling the number of multifamily properties along the way. Debate number 3: Distribution of for-sale listings will fragment, causing Zillow to be disrupted. Our perspective at the time: In any future industry scenario, we are well positioned to have broad listing coverage because of our brand, our audience, and the quality of our strategy, products, and execution. Our perspective today: In any future industry scenario, we will meet consumers where they are and be flexible about how we get listings. A recent example is Zillow Preview, which we announced last week. We continue to have a beloved brand and a highly engaged audience. Debate number 4: Recent lawsuits by industry participants and regulators will create challenges to our operations, causing Zillow to be disrupted. Our perspective at the time: The allegations mischaracterize our business, and we will vigorously defend ourselves. We build products and services for consumers. Our current perspective: Compass abandoned its lawsuit last week after a judge ruled their arguments had no merits. We will vigorously defend ourselves against litigation and continue putting the consumer first in our products and operations. The brand strength amongst consumers and agents has never been stronger. That was a lot. Before I wrap, I want to put our performance over the past few years and what consensus has for 2026 into context. This set of charts looks at our performance and expected performance across all consumer internet companies in the $10 billion-$50 billion market range. Our revenue is expected to grow 34% over that time frame versus a category median of 29%. Our EBITDA is expected to grow 56% over that time frame versus a category median of 34%. Our stock-based compensation is expected to decline by 22% versus a category median of flat. There are very few, if any, end markets that are more challenged than housing during this period of time. Now, of course, we're here today because we are discussing debate number five: horizontal LLMs will make vertical internet marketplaces obsolete, causing Zillow to be disruptive. Excuse me. Our perspective today, AI is an accelerator to our strong brand, our engaged audience, our unique strategy, and our differentiated workflows. AI will help us grow faster and will make us stronger. Today is intended to show you how strong our opportunity is, and how important AI will be to helping us further differentiate ourselves from anyone else in real estate. It's a really exciting time at Zillow. When I'm in recruiting pitches with prospective employees, I often get the question, "Why are you so excited about the company?" I answer it simply. We have an iconic brand, we have a unique strategy, we have huge goals, and we operate with startup speed and ambition. So much of the credit of who we are today is a reflection of our next speaker. With that, I'll hand it over to Rich Barton, our Co-founder and Co-Executive Chairman. All right. I get to kinda begin to land the plane right before we head into demo land and lunch later. It's really nice to see you all. There are some familiar faces. I really wanna thank you for coming out and spending all this quality time with us, especially given everything that's going on in the world and in our working lives right now. Thank you very much. I'm gonna speak for about 15 minutes, and then we're gonna bring a panel up to do Q&A, and that'll be the final stage session to take your questions. Okay. What I wanted to do, as the one of the more experienced of us here, is to tell two stories of personal experiences with companies that have crossed a technology chasm. Okay. I'd like to share a simple framework that I use to determine who in my world I think is gonna make it across the chasm or not. All right. Starting with story number 1. Let's go back, if you can, to 2001. Again, many of you may, some of you at least, may not have started your careers by then. In 2001, I was the founder and CEO of Expedia, which had been public for two years, was the leader in the online travel space, a burgeoning interesting brand, and a disruptor in and of itself, taking advantage of the internet chasm. I got an email from Reed Hastings, who's the founder of Netflix, and he was inviting me to dinner. I was a subscriber to Netflix DVD by mail. This was an amazing product. You know, but it was obviously a product that was gonna hit a wall. Reed invites me to dinner because he's getting ready to go public, and he's trying to put together his go-public board. I come in, meeting Reed, and I say, "Reed, this is certainly an awesome product and an interesting company, but, like, DVD is dead man walking. Like, it's not. This is just not the right form factor." Reed kinda smiled and leaned in across the table and he said, "Rich, I didn't call the company Netflix DVD-by-mail service. I called it Netflix. I paused for a second and I said, "Reed, that's a pretty good response." So from day one of the founding of Netflix, Reed had his eye on the chasm. He knew it was coming. He knew it was over for the DVD form factor, and he was confident that he and his incredible team would find a way over the chasm to what became streaming. Basically, a delayed internet shockwave for his business. I joined the board, and I enjoyed a front row seat on Reed's trip across that chasm, and I learned a lot from that. I am still a board director to this day. All right, so that's story number one. Story number two, let's fast-forward about six years. This chasm that we're facing is the mobile chasm. It's 2007. zillow.com is one year old. We launched version one in 2006. Sorry. Confusing my time zones. All right. In 2006, it was a year old, and we were well on our way to category leadership, but we were still about half the size of the leader in the category in 2001. We got there very quickly on the backs of an incredibly provocative product anchored by the Zestimate, which was a completely innovative feature that drew lots of people to try Zillow out. Something happens. Steve Jobs goes on stage at Moscone, and he introduces the iPhone 1.0. I remember Lloyd and David Beitel, our CTO, who was also with us at Expedia. You'll meet him in a sec, and myself, we all got iPhone 1.0 on the first day, and we kind of disappeared into it. We had this incredible epiphany together that it was obvious that this device was going to change computing. It was going to change everything. We immediately threw our BlackBerrys away, which we were addicted to. I'm sure some of you out there had those BlackBerrys then, and I'm sure you complained about the soft keyboard of the iPhone 1.0. We figured out how to use our thumbs and use that soft keyboard because we saw this as a brand-new platform. We decided it was gonna change everything, especially in real estate. As cool as zillow.com was at a computer, at a desk in your house or in the office, so cool, it was gonna become 10 times cooler when you could have it in your hand walking around a neighborhood that you were shopping in or driving on your Sunday drive for shopping. We knew it was gonna be 10 times as interesting. We called a code red. We decided to pivot the whole company and go all in on this new platform. We decided if we didn't do that, we were gonna get run over by the company that did. All right, it worked. We were the first real estate app out on the iPhone platform, and we were by far the best. There was no map rendering service offered by Apple or Google later, initially. We had to build our own map rendering on iPhone. So much so that Apple took note and invited us down to Cupertino to tell them how we did it. Okay. We formed a relationship with Apple. We were very excited that on a future launch, Steve Jobs himself actually demonstrated the Zillow app at a product launch, which is pretty huge for geeks like us. This app propelled us. We were kind of half the size of the leader in the space on the web, and the Zillow app propelled us into the number one spot in the category, and we've never looked back. All right. Moral of this story is we saw the obvious future. We pivoted fast and with skill, and we crossed this mobile chasm to glory. All right, we're facing a new chasm right now, all of us, and trying to figure out who's gonna make it across and who's not. In addition to being founder and co-chairman here right now and an occasional CEO here at Zillow, I am a venture capitalist and an investor like you all. Your questions that you're asking are the same questions that I am asking. I'll share a quick kind of rubric that I've been using, the criteria that I've been using to begin to grade companies' prospects. All right. Here's a summary of the criteria. There's not gonna be a lot of surprises here. I'm looking for companies where the founders are leading the company or really highly engaged in the company. I wanna see founders. I wanna see big, complicated transactions. I wanna see unique content and data, truly unique. I wanna see hidden depths of industry integration, and I wanna see brands that command their customer or their consumer funnel. I'll dig into each one of these briefly. All right, founder engaged. This is an obvious one. The founders created the mission, and they caretake the mission. They are usually engaged and paranoid. They're usually engineers, which means they can hire great engineers and all of the terrific product and support people that surround great engineering teams. Founders tend to be product first. They tend to lead product-led organizations. Founders are much more willing to hit the big red button and pivot the company when a new threat or opportunity emerges. Okay, so how does Zillow rate here? Well, Rich and Lloyd, you'll meet Lloyd in a sec. Some of you know Lloyd. We actually met at Stanford as engineers and studying engineering and economics. It's kind of trite that we met at Stanford, but it is actually true. Lloyd, in fact, started working at Microsoft when he was 14 years old because his mom and Bill Gates' mom were buddies, and they both had geek sons. True story. He went to Microsoft after school, and he had a hand in recruiting me into Microsoft. While at Microsoft, we put a great team of people together on a project, Expedia. Together we did that. We spun it out of Microsoft and did quite well in creating a really iconic consumer brand there. When Barry Diller bought Expedia, the two of us combined took some of our winnings from Expedia and rolled it into the founding of Zillow. We took the first $5 million that we put in ourselves into Zillow. Okay. This is all a way of saying that we here, we care, we're involved, we're experienced, we're motivated, and we are really excited about this AI acceleration that we're seeing right now. All right. Next criteria for making it across the chasm, big transactions. I want to see high-dollar transactions, ones that matter. I wanna see ones that take a long time. I want to see transactions that are complicated and recursive in their process. I like visually engaging shopping experiences that create emotional connection and entertaining connections. All of these things are defensive against kind of horizontal platform assault or being meta'd. Okay. You guys can think of companies that tick the boxes here. You know, this reads like, from a Zillow perspective, a laundry list of the real estate transaction. This is one of the reasons I was attracted to this transaction initially, and it is the reason I love it as we are looking, staring down the AI chasm. All right, next criteria for winners from losers. I look for unique content and data. This is an obvious defense against horizontals and being meta'd. Okay? It's also an obvious audience attractor and differentiator as well. Some companies really do have this. I've put a few examples here. Most do not, however, and it takes a discerning investor to figure out what content is actually proprietary and unique and protectable and what is not. Okay? How does Zillow rate here? You... A big chunk of the day that you've seen, the impressive demonstrations and speakers that you've seen today, I hope have affixed in your mind that Zillow is sitting on a giant and rapidly growing trove of unique content. Unique content and data that we integrate with more commodity content to create a unique alloy in the form of listings, media, and actually this double feedback loop of consumer data from the consumer side and the professional side that intermingle to reinforce each other, that feed this proprietary user information into an insatiably hungry AI that just makes the experience better and better. The breadth and depth of the data that we have at Zillow and that is accelerating in growth will widen the lead that we've established in the industry. All right, almost done. Number four differentiator or criterion that I look for, hidden depths of industry integration. Of all the stuff you've seen today, Nicholas's demos were amazing. Cynthia laid out a great framework. You know, probably you hadn't seen or known about a lot of the stuff that Cameron talked about relative to the professional, the agent and the loan officer as a user of our software. A daily, multiple times daily user of our software. Okay? That's probably been the most eye-opening for you. Generally speaking, I'll just summarize it. I think you get this point, but looking for companies where the. You guys probably know this, but a physics fact, 90% of the mass of an iceberg is actually underwater. Okay? I'm looking for companies where 90% of what they do is actually underwater, where their software is insinuated and pervasive in an operating system-like way into the digital workflow of an industry. Okay? I'm looking for massive use and dependence on the pro side. Companies that are springing to mind here are, you know, I don't know them intimately, but I would assume Carvana is this way. Expedia has a lot of this. Airbnb is absolutely this way. Shopify, Uber. Okay, this is where the user experience that we experience on a daily basis when we use the product is literally just the tip of the iceberg of what the company does. Okay? How does Zillow rate here? It's been on great display today. There is a lot going on under the surface of the water at Zillow and more every day. Zillow, I think you've heard it, people have said it a couple times, but I have been, we have been talking strategically about building the digital operating system for the whole of the industry, and we are well on our way to doing that. All right. Finally, my last criterion for the separating the AI winners from the AI losers, who's gonna get across that ugly chasm. That image was probably produced by AI, and AI is in the base of the canyon. Command of the consumer funnel. I wanna see companies that really do command their consumer funnel. They have huge brand strength as measured by third parties, not themselves. I wanna see low churn on every churn metric or low relative to competitors on every churn metric that we can think of. I wanna see almost all, most of the traffic and or customers coming to the brand directly, not via paid or even SEO channels. A way to measure this that I particularly look at is I'm impressed by companies that have low sales and marketing spend as a percentage of revenues relative to their competitors. It's a pretty simple one. I'm not telling you anything that you don't already know. Some that I'm familiar with, Netflix and Airbnb, their sales and marketing expenditures are 20%-33% the size of their competitors. One-fifth to one-third of their competitors. Guess what? Zillow's are the same. We're highly efficient in our sales and marketing expenditure as a percentage of revenues relative to our competition. Zillow is in rare air. This is one metric that I'm grabbing that's just happens to be a third-party metric, Comscore. It's correlated to brand strength. But it puts, this is just unique users, Zillow in a category of brands that are fortress brands, powerhouse brands that command their own funnel and command their own destinies. It's a little surprising almost to a certain extent, given the size of the company, honestly. I've always said we cast a much longer shadow at Zillow than we are tall. Now what that says to me is tons of opportunity and potential to grow, right? Herein lies tons of opportunity to grow. We have a lot of TAM in front of us. Okay. Zillow has no dependence on anybody upstream, and we never have. It's another proof point. Another one is that Hofmann just talked about. Zillow has withstood a massive and violent competitive attack in the last few years and done so with aplomb. That should be telling you something. It tells me something. Okay. Summary, my little checklist, my scorecard. Zillow checks my boxes. Founder engaged, sometime founder annoying. You're gonna meet a couple other founders, right now. Check. Big complicated transactions, obviously a big check there. Unique content and data, check now, and it's gonna be double check in the future. Hidden depths of industry integration has been on display today, and it's impressive to me, and it's been our strategy right from the start. Obviously, Zillow commands our own consumer funnel. All right. You can tell I feel good. We have the team, we have the experience, we have the tools, we have the vision, we have the motivation. Personally, I love technology chasm crossings. I believe the teams that I've been a part of and led perform at their very best in these moments. We've navigated across the internet chasm, we've navigated across the mobile chasm, and now we are navigating across the AI chasm. We see it as an accelerant. We see it not as a problem, but as an opportunity. As evidence of how I feel, you may have noticed in the last few weeks, I've been pretty serious stock buyer in Zillow, personally. I guess I'll summarize by saying I'm more fired up about the opportunity in front of Zillow now and our position relative to that opportunity than at any point in our 20-year history. Okay. Again, thank you for coming. I'll invite Brad Berning back up onto the stage to bracket it while we get our Q&A panel together, which is the last session. Thank you all very much. All right. I get to officially kill a few minutes while they try to reset the stage for us, with some chairs. First of all, thank you, Rich. Thank you to all of our speakers, and thank you to all of you for being here, today. I think you got to learn a lot, but I also wanna make sure you get a chance to have some time for Q&A here with investors and ask follow-up questions to the day. During the Q&A phase here, I'm gonna have a few ground rules to try to help out a little bit. We're gonna continue to do this live, so people can actually hear this online. Raise your hand, I'll call on you, and you'll be able to get a microphone. We have three microphone runners that will run around, that are gonna be able to help you out a little bit. Sorry, I'm just killing a minute. Our speakers today are gonna be Jeremy Wacksman, it's gonna be Jeremy Hofmann, it's gonna be David Beitel, it's gonna be Rich Barton, and the rare appearance from Lloyd Frink as well. I think you'll be excited to get a chance to ask all the direct questions from our teams. With that, I will invite our Q&A guest speakers to join us here in about, probably about 30 seconds, I'm guessing. Awesome. Thank you. Thank you, sir. All right. While we're getting queued up to get some questions started, I'm gonna start with one question to just get us warmed up a little bit. I'm gonna have, Rich, you talk a little bit about, we've talked about the future of AI, but maybe draw the through line about the vision of Zillow from inception to how AI is looking forward for you. Yeah, as if we haven't heard enough from me. You know, for me, a simple strategic insight and the through line to everything we've done, you know, in our careers, at least the three of us here on the end, is, you may have heard me preach about it, I call it power to the people. It's a sure thing to bet on technology and products and services that empower people, that give people more power. The history of technology has been the history of giving people more power and more control, tracing back to when we were kids and the PC came out, and that was really radically empowering in a really eye-opening way for, you know, for a subset of us that were using those products. That PC got connected to other PCs via the Internet, and that opened up. That was another technology chasm. That opened up a whole new host of personally empowering services like Expedia and like Zillow, that enabled regular people to kinda tear down the walls of the Bastille and access information and tools themselves by which they could make better decisions. That whole system got packaged into a tiny little iPhone, you know, and an Android phone later, and it became kind of a mobile version of that magic thing that ended, has ended up turning into a remote control for our lives, a magic remote control for our lives that we can control right from where we are standing, sitting, driving, wherever. Another empowering layer. I'll wrap by saying we see AI in this same way. It is the next evolution of this personal empowerment. It's a much more kind of intimate. It's almost like an intimate and a much more kind of action-oriented away from kind of research and control into an intimate and action-oriented technology. We see it as super empowering. It will be another layer of super empowering consumers. I'll stop, and I'll actually turn to Lloyd, 'cause we promised that Lloyd would get some air time, and we think you'll enjoy hearing from Lloyd. I'll ask him, what was my setup? All right. Lloyd is our co-founder and lovingly our most super geek. Okay? His role relative to AI at Zillow has been the largest token consumer, the largest single token consumer, and I would say the chief AI provocateur to every organization in the company. He's Captain AI Code Red. Lloyd, tell me about how do we set this up? Tell me about November 30th, 2022. That was the date that ChatGPT 3.5 launched. I first heard about it from Rich. He said he'd seen Bill Gates at a Kraken game, and Bill was engrossed in his phone doing this chat thing. He's like, "What is that, Lloyd?" I was like, "Chat? I don't know. We've seen chatbots all over the place. Like, I don't really know what you're talking about." Later that week at a dinner, somebody mentioned this, you know, ChatGPT thing, and I was like, "Ooh, second source. I should maybe go figure out what- I wasn't enough for you to go look it up? Bill Gates wasn't enough for you? I felt like a little guilt. I like, "I gotta go check this out at some point," but when somebody else says it was like, "I gotta go check this out." After dinner, I go home and open up my laptop, and I open this thing up, and I'm like, "Okay, write me a program that prints out the prime numbers between one and 100 in language BASIC." This is the first program I ever wrote on a computer. BASIC isn't this language around anymore. I went, "Okay, can it really do that?" It spits out like five lines, and I'm like, "Whoa." I'm like, "Oh, that's not quite right." I'm like, okay, about ready to dismiss this. I see a bug in it, right? I kinda go, "Hey, you know, are you sure that's right?" It went, "Oh, yeah." It fixed it. I was like, "Hmm, okay. Just a second." I was like, "Okay, do that in the programming language C." It like did it. I said, "Okay, how about Fortran?" It did it. I'm like, "How about Swift?" Yes. It did Python. I'm just like, "Oh my God." I knew that I wasn't gonna get a lot of sleep that night. For the last three years, every time a new model comes out, I lose sleep just playing with it and just so excited. I've been vibe coding, you know, for the last three years, doing all different sorts of little projects. A lot of them are kind of video games, Pong games, different Atari games, et cetera. A lot of them are work-related things too. I can imagine any app, and I do all apps for the iPhone. They're actually just little websites, but they work really well on an iPhone. I can imagine any of those things, and I can make it happen. It's just a magical experience for me. How do you work? Okay. What is your cockpit? This is my geeky thing. What does your cockpit look like? My mission control. I call it my mission control. We work from home. I have three different monitors up there. I work on a handful of different projects at any given time. Three of them I can have open. I work on these during meetings. I can kinda start something here, monitor what's going on there, what's going on there, and, you know, paying attention to the meeting. It's a lot of work that's going on in the background. Inside of each one of these projects that I have open, it's Cursor is the main control panel that I have. I use Claude Code and ChatGPT. They're working together. They're criticizing each other. They actually work pretty well. Their personalities are pretty fun when you get them to collaborate together. I have Databricks, where all of our data in Zillow is. You know, anything I could imagine, any question that I have, my lens onto our data, I can just whip up a little app and get the answers to it. It's just the most fantastic experience. I remember some of you actually may have come to our Zillow Unlock event. You know, how many have people done that? A few of you. Okay. All right. Nobody wants to raise their hand. I remember last Unlock, you running around talking to the agent team leaders and the agents and the brokerages with an iPad open running your vibe-coded data app. Yeah. Let me explain to you, this is one- You didn't wanna talk to them to learn about their kids. You wanted to talk to them to show them how they were performing. Go ahead. The way that, you know, when we're looking at our business, we have a funnel of the business, all the users that come in, the connections, and then the funded loans and the transactions and the integrated transaction that come out, right? We have our funnel of all the data, and I have my particular view of that on my iPhone that I can look at by day, by week, by month, all the different charts, all the different metrics, and I have a really good sense for our business on my phone. I wanted to do that for loan officers and for real estate agents and for the teams of those real estate agents. I have this little app that's like their pipeline. You know, all the data that you see on on Follow Up Boss, right? We have a similar CRM for loan officers. They can see each other's when they're working on mutual clients. They can see each other's pipelines. I can see their pipeline. I have whether or not they're green, yellow, or red in terms of performance, and that's measures how much the allocation of leads that we're gonna give them. They just are blown away that I can just instantly say, "Okay, tell me about this lead. Tell me about what happened in this month. Oh, you were gone for these months here. What happened to your conversion there," right? That's the main metric that we look on. you know, they're just like, "Oh my God, you can see all of that?" They know it, and it's a much richer conversation that I can have with them, and try to figure out if they have a strong relationship with a loan officer and- They realize if they do that, you know, those leads that have been pre-approved by ZHL are five times more likely to convert. When you get somebody that actually viscerally understands that, you can see that their conversion is much higher, and they're much higher up on those leaderboards than other people are. Those are the kinds of conversations that I can have. Yeah. Conversations. He's kind of the data terrorist for inside and outside the company. All right. Yeah. Let's get David. Let me set you up, David. David's our founding CTO. He was the founding CTO of Expedia almost. You know, when we opened up. I was a founding member. you were the CTO. We worked at Microsoft together before that as well. Yeah. I mean, you could sense Lloyd's excitement, right? Having a founder that is so engaged and has had his aha moment, the kind of culture that creates at our company, right? It's not a fear, it's an empowerment. It's a way of thinking as an employee that if there is something you need to know about what Lloyd's describing, how our business works, how our products work, how we should think about that next feature, you don't wait around for someone to bring you that tool, you go figure it out. That's just really a huge unlock. Maybe many of you in this room haven't had that personal unlock, that personal aha, but once you do, it's just. There's nothing you can't do, and there's certainly people around the company that can help you. That is the kind of what happens within the kind of confines of Zillow. You are empowered, you're given the tools, the models are changing. That's awesome, right? Every week the tools get better, and there is nothing we can't do. That's the mindset. I appreciate that. Not everybody in this room gets a chance to Major to meet our founders and get to have this conversation. Hopefully it gives you a little bit of context behind the culture. Exactly Right, on AI that we're driving behind this and why we're excited about the opportunity set. I appreciate you guys getting a chance to share that with the group. It also I think explains to Rich's checklist in which you have engaged founders, which I think becomes very obvious even if Rich wants to call it an un-founder checklist. All right. We can wind up the AI section and you can take the ball back. Yeah. I'll conclude with for those who were at Lloyd's launch at table or you run into him out in the demo zone out there, ask him about a couple things or don't because you won't get a lot of I got lots of different little apps to show you all maybe. One is this incredibly cool solar system app that he vibe coded that he's very, very proud of. You'll be getting- The other one. Yeah, that was the solar system. One that's more useful is all those conversations that Nick was demonstrating, that Nicholas was demonstrating when he was up there. Lloyd wanted to see all the raw conversations. He didn't wanna see the AI summaries of the conversations that were happening, so he vibe coded a little iPhone app that literally has real time Zillow AI mode conversations that are happening right now, you know, flying through, so he can actually go inspect everything himself. Ask him about that. All right, Brad. Awesome. All right, we'll get to your questions now. Who's gonna raise the first hand? Carl, if we can get a microphone over here. If I could ask, state your name and your firm names, please, so we maybe get a little bit of understanding of who's asking questions, please. Yeah, of course. Matt Cost, Morgan Stanley. Thanks so much for taking the time. This has been great. I guess you're clearly spending a lot of time thinking about what the next generation of tools on Zillow might look like, and experimenting with a lot of the technology. I guess first question would be, what are one or two moonshots that you're working on right now that are the most exciting? Are there virtual agents that are coming up? I mean, what's happening in the bowels of Zillow using this new tech that could be transformative? Then the second one is on pre-marketing and Zillow Preview. I guess what are the opportunities that you might get from the data of seeing people at that first point in the exploration journey of what homes are available? when you think about, EXP signing a deal with Homes.com and Compass and Redfin, is there an execution challenge there as well? Because the first step in some people's journey if they wanna see that inventory now takes them away from Zillow. Yeah. Before we get started on that, I probably should have set one ground rule. Start with one question at a time. Trying to remember three at a time would be a little tough. One of us will remember. We'll get to all of them, Matt. You wanna start? I can start. Yeah. Lloyd can jump in. I mean, I think you saw a lot of the moonshots today, honestly. I mean, if you think about the categories of where we're focusing our efforts, you saw Nicholas take you through AI Mode and rich media experiences that become just much more immersive, much more realistic, and drive a lot more consumer engagement. I think that's kind of one class. You saw moonshots on the professional side. Everything from what we're doing with agents, but also what we're doing with photographers and letting people do things without specialized hardware. These are things that will drive more engagement from the buyers and sellers, but will drive more engagement from the professionals that are building their business with us, and that feedback loop just continues. I think you'll get to spend some time with some of the demos this afternoon, and I encourage you to play around with them. Then on Preview, you know, as Jeremy said, we don't really get too concerned about listings flow. Listings always flow to Zillow. Preview is really a response to the market wanting to be able to bring coming soon listings online. There are so many markets where there are coming soon listings. They're a small percentage of inventory, but they start some listings, and many rules and MLSs don't allow those things to be displayed publicly. We looked at that and said, "Okay, brokers want a way to have that coming soon inventory on Zillow." The response, even in the first week of announcement, has been really tremendous. I mean, the broker partners we launched with, we've got hundreds more that are beating down our door to get into the program, and we'll talk about those as we're able to get those onboarded. I'm sure there'll be partnerships and fragmentation around some of this listing supply, but one thing we are highly confident in is most all buyers and sellers, most all agents who represent buyers and sellers are gonna wanna find that content on Zillow to get their home sold. Yeah Great, thanks. John Colantuoni from Jefferies. I wanted to ask about the potential evolution of the role of agents as it gets easier for consumers to do research on properties using the AI products that you're creating. I'd be curious to hear how you see the importance of agents and the fees that they're charging evolving over time, and how this potential change could impact the way that Zillow generates revenue. Thanks. Awesome. Thanks, John. I think we'll start with you. Yeah on that one. Yeah. I mean, I think you heard Cameron, I think, put it pretty eloquently. The long arc of technology coming to real estate agents has been to elevate the good ones and empower the good ones to be better at what they do, and I think you should expect AI to continue to accelerate that trend. You're already seeing the rise of the more professional agent, the more, you know, team-focused agent, the one that can respond to modern clients, and you saw some examples today even of how technology lets them do that more. That allows them to do more deals at a higher quality, right? More volume and more quality at the same time doesn't normally happen. That's a real economic unlock, and the value they provide goes up in that case. The best data point to look at there is through all of these technology shifts that Rich talked about, we ask buyers and sellers, you know, "How do you wanna hire professionals? Do you wanna hire professionals? What do you wanna hire them for?" The percentage that want to hire help has gone up through the technology shift of the internet, through the technology shift of mobile, and now with AI. Now, what help they want is what's changed, right? If you asked that question in 2006, "Well, I gotta hire an agent 'cause he's got a book of listings that I don't have access to, so I'm basically just paying for access." It's not a great experience, right? Now, I'm hiring my local market expert, the person who knows my street better than anyone. I'm hiring a negotiator who wins listing or wins offers more often to help me win in a competitive market, right? That's what they're hiring now. They're able to hire a better professional. That's why you heard Jeremy talk about through all of the changes that have come, the partners we work with have seen their businesses and their pricing power go up because they're good at what they do, and they're unlocked to do a better job. I'd just echo that. If you think about what we're building with all that Nicholas showed, it's going to better empower a consumer to understand more before they get off their couch. That means it's going to be a higher intent lead. On the other side of the marketplace, you're making these agents far more productive because you're delivering higher and higher intent leads. So much of our opportunity is not just to grow the funnel that we have and grow the brand, but also mine the group of people that are already in our sites and apps today. 70% of all buyers and sellers are on Zillow today, and our transaction share is far beneath that. We think about these AI enablement tools as way more understanding from the consumer, creates a higher intent lead, creates higher productivity for the agents, and the whole marketplace spins better as a result. That's right. Thanks. Taking the question, Bernie McTernan, from Needham & Company. I'm assuming the goal of all this AI technology being rolled out is to keep the share gains going or even accelerate them. Can you talk about what's the more important part? Is it the consumer technology to get that more funnel? Is it the professional side? If you could just talk through those dynamics, and maybe the follow-up is just when this could actually start impacting the financials in a more material way. Maybe I'll take the financials, and you hit the- Yeah. David Beitel should jump in. I mean. Yeah I'll start. I mean, both sides are important. It's kind of a cop-out answer, but you sort of saw it in how we organized the day, right? The consumer side is really important for driving deeper engagement, more rich conversations, as Jeremy talked about. That just results in higher quality leads if you're on the professional side. But you saw examples of that from Nicholas. I mean, the depth of conversation and the tools, the depth of interactivity with rich media, those are just a couple examples of pulling the consumers below the surface of that, of that iceberg. Then on the agent side, it's similar. It's tools for productivity, and it's tools to unlock them to do a better job, whether you're an agent, whether you're in a team lead's office doing other roles, whether you're the loan officer, whether you're the closing professional, getting that transaction done more efficiently. That lets you close more deals for the same cost. As a small business owner, that's the number one thing you're focused on. For us, both sides are important. I think it's a great question 'cause if you take one thing away from today, back to Rich's checklist, it is you have to work across this whole thing to build the real estate experience. There are a lot of companies that are gonna attack a little piece of it, are gonna attack just the top of funnel of it, and I think that doesn't actually solve the problem, right? Like, solving the problem is covering both sides of what you asked about. Yeah. From a financial perspective, it's already showing up, right? Like, we started, to Lloyd's point, using AI tools as soon as we possibly could in late 2022, early 2023, and it's been part of the cost structure that we fight inflation against and then look to invest in these AI tools. Those are delivering a ton of value to us already because we're seeing more growth in conversion transactions, and we're seeing that done with leverage. I'd say it's already there. We look and say we're still scratching the surface comparatively, and this should be more of an accelerator on the product strategy because of the way that we're deploying it. Mike Ng from Goldman Sachs. Two questions for me as well. First, for the Zillow Preferred agent business, I was just wondering if you see an opportunity to increase the number of leads that get served up by, you know, selling leads to preferred agents that may not necessarily be buyer-initiated, but, you know, by using all the cues and intent signals, you know, that you see from browsing data. I just have a quick follow-up after that. You? Sure. Sorry. You're getting all of them. That's more of a conversation about the overall philosophy of kind of the lead qualification funnel we have, and I think AI obviously plays a huge role in that, but so does providing software and tools to help agents mine their entire business, not just their business on Zillow. I bring that up because I think you can take Zillow Pro as a really good example of how we might do that. Again, you saw it today firsthand what an agent does and then what a team does who has 10 or 50 or 100 agents at that scale, is they are running a small business office with a whole bunch of customers with some amount of conversion rate of those customers in any given week, month, quarter. Our ability to help them figure out who to work with and how to work with them drives conversion gains, but it also drives volume in, because when they're doing a better job and they're doing a more efficient job, they're able to go grab and attract more customers. The tools we're building them that you saw from Cameron, those tools are the same tools they're gonna eventually use on their entire business. They're gonna be able to be more efficient at the customers that are already in their database or their referral customers, the same way that they're using Zillow customers. Our goal, as Jeremy said, is more and higher converting customers for agents and then more and higher converting agents to be able to work more effectively with those customers. Great. Thank you. Mike, I'm gonna add one thing onto At Unlock last year, I think we had an agent who made a really good example of this. They have 350,000 people in their database. They don't have a problem on number of customers. The prioritization of which customer to call is their bigger problem, and I think you could see from the stuff that Cameron was showing today some perfect examples of the prioritization that we're using real signals on integrated experiences that others simply cannot do. Great. Thanks, Brad. Thanks, Jeremy. Just as a follow-up, you know, great to see the new, you know, net income margin mid-cycle targets. You know, I think there is a little bit of a debate about what, you know, mid-cycle is at this stage. I would just ask, like, you know, do all these tools and this product roadmap give you the confidence around double-digit revenue growth, you know, even beyond 2026, you know, if the housing market kinda stays the same? Yeah, no forward guidance beyond 26, but I would say, Mike, we feel very confident in our ability to hit those mid-cycle targets. I understand there's a portion of that that's in our control and a portion that's not, but the portion that's in our control is a really attractive financial return before you even get any love from housing macro. We feel like we're well on track, and AI is gonna accelerate, if anything. Okay. Thank you. Yeah. Absolutely. Over here. Thanks. Jeff Meuler from Baird. Can you talk more about solution penetration rates within the rentals business? I think you did a really good job of providing a lot of penetration metrics for the various for-sale services. But within rentals, you obviously have outsized growth in multifamily, and I would think some of the larger multifamily managers are already using other vendors, accounting software providers or things like that for services you're trying to provide. Can you just help us with, I guess, end-to-end solution within rentals and where you're seeing penetration growth by solution? Yeah, I can start. I think you saw a little bit of it in Michael's, but that was less of his focus today. The way you wanna think about it is, you're right in your question, the more professional segments have more tools already at their disposal that they are using, and our strategy there is integrate in. The long tail, which is actually the majority of the inventory, does not have tools, and our strategy there is provide tools. You want those two strategies to meet, where they can, right? A good example of that is our applications product. The long tail uses our applications product 'cause they have nothing else. It saves them time and effort. It saves their consumers money. It's why it's preferred, and that portable application is on a growing set of inventory that's unique to Zillow, right? A good chunk of that long tail inventory comes to us uniquely because of tools like that. Well, the multifamily segment has an application product in many cases, but they want the benefit of Apply Now, and they want the benefit of tour scheduling, they want the benefit of the consumer experience, so we can partner with them to connect that in more places. The benefit to the consumer is they get a more consistent experience more broadly, and maybe we're pulling the long tail, and maybe we're pushing to multifamily to get to a more consistent experience. I think you'll see that strategy play out. There'll always be tools that partners have capabilities on, and it's what's the value we can add to plug into those tools or provide those tools. Over time, that tool set should provide more value across the marketplace because we have the biggest set of audience, and we have the biggest set of content to put it against. Can I add something? Please To a number of the questions. One thing that may not be clear that we've been invested in is a framework that allows us to leverage all the existing models that are out there, but also have the people and the expertise to build very specific smaller models and agents that help power these experiences. Where you might see some of our competitors or some others in our space, you know, just act as a proxy to the big players, when you start looking at the richness and the personalization and the depth of the types of products we're building and the interactions and the conversations that we're having, you're gonna see us separate ourselves quite a bit. That is because we are investing in these very specific models that help drive very specific use cases within real estate. We're just beginning on many of those, but it's accelerating, and you'll start to see us separate from the others from that. Yep. Dave, you had your hand up a minute ago. Thank you. Dae Lee from J.P. Morgan. Rich, when you talked about, you know, technology companies crossing the chasm, like, as we look at the AI chasm in front of you guys, like, what gives you the confidence that Zillow will be the one to move the fastest in crossing that chasm and be able to successfully cross that chasm? Where do you think we are in crossing that chasm right now? I mean, I think I would fall back on my, you know, on my checklist. You know, I talked for a while and fairly passionately about the checklist that I look for. You know, I can run through it maybe if I can pull it off the top of my head, but you're meeting some founders right here who are obviously very engaged in the company. You know, the complexity of the transaction is a real defense mechanism relative to new player disruption and incursion, especially by horizontal players. It's real. It's not permanent, okay? It is a real defense which gives the companies that have these complex transactions a lot more time, okay? Interestingly, wind back. The Netflix story that I told, the streaming thing didn't happen to Netflix until probably. You guys tell me. It must have been like 10 years. 2008. Okay? The graphical web showed up in 1996, okay? The easy transactions fell very quickly to internet disruption. You know, books, airline tickets, you know, some of the stuff that we did. Then the complicated transactions that were more byte heavy, which was the Netflix story, just a lot more byte heavy, okay, we knew it was gonna come, but it was a lot more byte heavy. That was delayed 10-12 years before that transition happened and drove the DVD into the ground and then eventually cable and broadcast into the ground. The complex transaction holds out a lot longer, or the big transaction holds out a lot longer. This kind of industry rails stuff that you're hearing us talk about and that you've seen, it's not permanent. You know, being the digital operating system for workflow in the industry is certainly not a permanent defense mechanism, but it is an incredible, it buys you a lot of time and puts you in a position. The risk there is that you get distracted on digitizing and automating the processes and people that are in place now, and you take your eye off the ball to the disruptor that's coming from the side doing something completely different, okay? Rest assured, we have red teams that we're funding all the time that are working really hard to figure out what that non-obvious disruption is in the whole thing, okay? We have done so since ChatGPT 3.5 launched. It was like a serious effort. It has been a serious effort for us to try to figure it out. We haven't seen it emerge from anyone else, but we watch very carefully the others as well. So anyway, I... You know, on the way there, you know, the first really interesting product from that red team that is working is actually this Zillow AI Mode that Nicholas was demonstrating. I mean, Nicholas and his team have been working on this for a long time, and they've drilled a lot of dry wells, okay? To my and Lloyd's and David's frustration, okay? I'm pretty blown away by what's happening with the small percentage of people that have been exposed to Zillow AI Mode right now. This is the aha. I think this is the aha moment for us. I'm trying to think what other... Oh, well, consumer I mean, I guess I would say the productivity gains that we're seeing are real from using all these AI tools internally. Yeah. Yeah. There's a lot of people that are using Replit, which is basically a non-developer vibe coding tool to express their ideas. All the developers are using Cursor and Claude Code or ChatGPT or Gemini. Those gains are real, and we're moving as fast as we can to become what we call an AI native company. Is it guaranteed that we're gonna be the one? No. But you know I really like our hand and I like our position and I like our mindset. Now, there's a lot of popular stuff happening on Twitter and whatever that says, "You gotta fire everybody or stop everything that you're doing for 30 days and figure out who's on board and who's not on board." You know, that's one way, and maybe if you're a lightweight, shallow SaaS company, you ought to do that. But we're doing our version of that and have been actually for the last three years, okay? You know, Hofmann able to hold fixed costs steady while growing at the same time is a manifestation of that strategy. It's working. You know, I'm pretty confident it's gonna continue to work. Yeah. Maybe just to wrap, 'cause it's so core to who these folks are. We have consistently been the innovator in the category for 20 years. We have the brand that we do. It is a hard category. Folks have figured that out the last few years. It is a hard category. We have consistently out-innovated. We are the leader, and we are highly paranoid people. Yeah. When you put all that together, we feel very confident in the future. From a technical perspective, I mean, when you're a leader in the space, and we have the data we have, and the user intent that we have, and you think about the models that we're building and training, who's gonna build the best personalization? Who's gonna build the best models to drive these experiences are the ones that have the expertise and the data to do that in the context. So I like our chances there. We're executing faster with the productivity gains that you're seeing, so we're getting deeper in our backlog faster. When you think about the richness and the complexity of the transaction, as we go deeper in the transaction, the kind of experiences that we're creating, others can't create. You should start seeing that when you wanna take a tour. We're giving real-time tour confirmations in-app, right? I mean, these are just the beginnings, but you go deeper, and it just gets more complex. The Zillow AI Mode that we're building can be deployed everywhere, right? I mean, it's not just. We're not just setting filters, right? I mean, you look at some of the others that are doing this, they're just setting filters. That's great. That's better than doing it manually, but that's just the beginning. I guess there's one more thing. We've always been threatened by the horizontal players, and there's always been a thesis. I mean, Wax talked about this a little bit in his presentation. The infrequency of this transaction makes it relatively unattractive to the ego-driven super geeks who are building horizontal platforms. Okay. I love these people. Like, we know these people. We grew up with these people at Microsoft, and I love them. We've purposely chosen categories that were less frequent and more complicated because we pretty much knew that it would be far down their priority list to get around to trying to subsume that. I would say, I don't know for sure, anything can happen, but it hasn't happened with the whole history of Google in real estate, and I would say it's unlikely to happen with the emerging LLM horizontals. You know, is going deep on real estate and focusing on this particular narrow vertical marketplace worth the effort relative to the other opportunities they have? I'm being a little candid here, but I just think that the logic has held, probably will hold. If it doesn't hold and it changes, we will figure out a way to, you know, based on everything David's saying and everything we presented today, we'll figure out a way to make sure we're still the winner and that we grow. But that should give you a little bit of comfort. It gives me comfort. We will be there when those, you know, we were the first app to launch within ChatGPT. Sure Right? For real estate. We're gonna be there to learn and figure out what's working. When you wanna have the richness of the experience that we, our customers want, there's gonna be a point where you need to jump into the full Zillow experience to have that. We're learning a lot, and we're gonna be there. These large LLM, horizontal LLMs are quite excited to partner with us, and we are gonna go meet our customers wherever they are. Yeah. If they're there, we're gonna be there. We're gonna be there. Yeah. I'll probably just wrap this section. Sorry. We're all piling on here. No, this is great, conversation. It gives all of you some insights. Like, we're having real debates. We've been having these debates for three years. Yeah. Like, this is real conversations about how we're preparing ourselves for this. The accelerating our strategy is not by accident that we picked that as a main point of the day, right? I think you can see the why. With that, I'll move on to the next, we got. Oh, um- Go ahead. Nicholas Jones, BNP Paribas. I guess, can you speak to data moats? I think there's a fear that as AI gets to play kind of everywhere, maybe on the legal side, a lot of the back-end information will start to kind of leak into the ecosystem. As we think about Zillow's business model, where are the really strong data moats, maybe on the supply side? I guess how does the super app strategy play into maybe building the moats? It sounds like, you know, there's a focus on this infrequent, maybe 14-year transaction, but you mentioned, I think 65% are rental viewers also look at for sale homes, which is more frequent. That seemingly would bridge maybe a timing gap between a 14-year transaction and a 4-year transaction or 3-year transaction. Does that Is that where the data moat is? It's kind of the consumer engagement data or on the supply side? Like, where are the data moats, and how should we be thinking about it? You wanna start? Yeah. Sure. David, you should jump in. Yeah. I mean, I hope you saw some of it today in some of the presentations and demos. I think the data moats are on both sides, right? On the consumer side, it's all of the content that's uniquely Zillow's. I mean, their rentals marketplace, a good chunk of that content is uniquely Zillow's. The proprietary rich media is uniquely Zillow's. All of the content we augment the commodity listings with is uniquely Zillow's. But the consumer content is probably the most interesting for the AI platform. All of the data post-listing content is in that five-month period when you're actually buying. All the examples Nicholas gave you when he was showcasing, that's all content beyond the listing description that the agent wrote. That's everything from the touring data to the insights about the neighborhood to what the Zestimate knows, all that data. Everything post kind of listing description, address, and price is what the buyer is using, and increasingly all of that is uniquely data for Zillow on the consumer side. On the agent side, it's all the productivity tools and data and workflows. It's everything from the showings and the tour data to the conversation insights about the buyers that agents and loan officers are using to the loan officer workflows. Like, all of that data is also unique to us. We think about it as, again, as I said this morning, it's taking the proprietary data that we all can see as users, but it's really the data below the surface, as Nick talked about, in the iceberg. That's all data that you don't see when you're using Zillow unless you're a buyer or seller, and that's all the data that's richer to actually help drive the transaction and now to help drive the platform to be a useful platform for buyers and sellers to use. David, I don't know if you wanna add. Yeah, no. Of course, that data's feeding into the models to make them better. You know, the 3D content just alone is an incredible source for making the Zestimate better and there are other cases. How we stitch that data into the experiences. I mean, Cameron and others showed how we're bringing that richness into those interactions. When you're an agent, you're seeing all this richness of the customer come to you in real time, right? I mean, that's really powerful. That's right. Yeah. It's all inside the fence. Yes. It's like it's not gonna be easy or possible or certainly easy for somebody to break inside the fence and steal all that data, and not that they would even know what to do with it. You know, we have an increasing amount of private conversational and click and behavioral data that is all proprietary to us. No one else is even really conceiving of this. Like, they're not even attempting to do this. This is a pretty audacious thing that we're building right now, and we're having success already building it. I've seen the test results of the AI Mode, right? I think, Mary Ellen, correct me if I'm wrong, we're working on getting beta test out to the audience, right? You build a test, some things I've seen, the 31 different quality tests that we run through there and the personalization that we can get versus other things we're testing against are very tangible and very real. You saw that, I think, in both Nicholas's experiences and well, Cameron's experiences today, right? You can see those are incredibly personal experiences, right? That's the differentiation of the data at the end of the day, is all the context and content and the integration all coming together. Yeah. Just to wrap there, maybe we can go to the next question. Yeah. Our ability to build that experience is uniquely ours. Yeah. A horizontal LLM, a player that's playing at the top of the funnel just doing lead generation, they can't create that agentic experience all the way through the transaction. All that data feeds all the way through. You just get confidence not only against the horizontal, but also those that are in our competitor space that are nowhere close with respect to assets and data differentiation. Right. Yeah. Thanks so much. Alec Brondolo from Wells Fargo. I think, you know, as I observed the AI Mode demo, I think the interesting thing about AI Mode is it goes from the customer finding information on Zillow to Zillow potentially making recommendations about how they should approach a transaction. How do you think about the potential channel conflict that introduces with the buyer's agent? Like, where, what should the buyer's agent be advising on? What should Zillow be advising on, and where's the line that you draw and say, "Hey, we don't wanna be, you know, part of that part of the transaction because we need to maintain value for the purchaser of the lead"? Yeah, it's a great question. There's no conflict. Yeah. I mean, that goes back to the philosophy. Yes. Go ask our Premier agents. Every time we invent a technology to hand them a higher quality lead, they go, "Thank you. Yes. When can I get more?" Right? From Real-Time Touring to the buyer's agency agreements to pre-approved connections, which Lloyd talked about, they first go, "Wait, why are you doing this?" Then they go, "Oh, that's a five times converting lead." It's the same thing. It's kind of the question earlier. Great agents wanna spend as little time as possible repeating themselves, doing the busy work, doing the automated follow-ups. They wanna spend time closing more deals, and if we can hand them higher quality customers, and AI Mode will do that. The more we can help prepare a customer. You saw in even the experience, like, that first customer asked so many more questions and then said, "Now I wanna book a tour." That was the meeting of the agent, and that customer is gonna be even higher converting for that agent. We see this as a conversion gain, and our agent teams will absolutely see it that way. That's how they've seen all of these software tools. Yeah. Maybe just to put some context around it, let's say we hand an agent 100 consumers today. We convert eight or 10 of those folks into transactions, just giving rough numbers. That's 2x-3x anybody else if you look at third-party data. There's still 90 of those folks. Yeah that are not converting, and our whole opportunity is make more and more of those folks higher and higher intent because it allows the agent on the other side to be more productive. They're spending less time with people that actually aren't going to transact. That's where we get more and more excited. We're not in conflict because our conversion rates can still go so much higher as we make people more and more thoughtful on what they're buying. If it was 8%-10% before AMOD, what is the conversion rate during AMOD? Uh- Two. Yeah. The question for the audience was if it was 8-10 before AMOD, like, what could it be with this? I don't think we're putting a target out there. The march we've done from wherever we started to quite a bit better than the rest of the category as a result of all the technology innovations. This will be yet another one. Yep. So... We're getting towards the end here, so I think we're gonna have one last question here, and then I'll have you guys wrap, and then I'll give you guys a little bit of details on next steps. Hi. Mizuho. Jack Han for Lloyd Frink. The Follow Up Boss demo looked really impressive. How much of this is effectively already in the business versus tools where there's more adoption or further product improvement ahead that can unlock better conversion rates for agents looking ahead? I mean, it's a. Jeremy ... it's a mix. No, Cam Yeah. Cam. Is Cam in the room? No. No, I can start. I mean, it was a mix, and I think he highlighted that, and you'll see some of that if you go outside and play with some of the tools too. Some of these capabilities are live. Jeremy talked about, you know, we ship features to initial users, and then we get 'em out to the rest of the users 'cause they claim 'em in a week. Some of these are capabilities we're piloting and testing. The through line on all that is they're all features to drive improvements in efficiency and conversion and effectiveness of the agent and to drive higher quality for the agent. That's kind of the overall grouping of them that you saw. This is all about, as Jeremy said, how do you know who to talk to and how to convert the folks you wanna talk to, both from Zillow customers and from the rest of your database? There's a long roadmap ahead for helping them do a better job of doing that. Yeah. Awesome. We're gonna wrap here. Thank you to all of you. This concludes our morning session. We're gonna have some live demonstration at kiosks that are outside and are gonna be open for about the next 90 minutes. Then we're going to start lunch in about 30 minutes up one floor, right, Mary Ellen? Correct. Up one more floor for lunch. There you're gonna have tables where our leaders are gonna be there for you to be able to continue this conversation. You'll find that some of you have assigned seats on the back of your tags, so you'll know which seat to sit at for some of the leaders we have. We also have some other leaders that are in the back of the room that lead a lot of the actual execution on a lot of this as well. There'll be an amazing group of leaders to be able to sit down and have lunch with, and welcome for you to continue that to be able to better understand our business. Thank you very much, everybody. Appreciate this morning. Thank you. Thank you.