Good morning,

This week’s Stratechery Interview is with new Zillow CEO Jeremy Wacksman. Wacksman is a long-time Zillow employee, having joined the company in 2009, three years after its founding. We also have a lot in common, including working at Microsoft after an MBA at Kellogg (we went in rather different directions since then!). Wacksman was previously COO of Zillow, and before that was CMO.

In this interview we take advantage of Wacksman’s long tenure to do a deep dive on Zillow over time, including how it became a weak aggregator (and what that means), what happened with the company’s foray into iBuying, and why the company’s new Super App strategy was born out of empathy for both consumers and agents. We also discuss the recent settlement with the National Association of Realtors, which has the potential to fundamentally reshape the U.S. homebuying market, and what the implications of that decision could mean for Zillow.

As a reminder, all Stratechery content, including interviews, is available as a podcast; click the link at the top of this email to add Stratechery to your podcast player.

On to the Interview:

An Interview with Zillow CEO Jeremy Wacksman About Evolving Strategy

This interview is lightly edited for clarity.

Background

Jeremy Wacksman, welcome to Stratechery.

JW: Thanks for having me. Excited to be here.

I always like to start these interviews learning more about my subjects and their backgrounds. You’ve been at Zillow for a long time now but where did you start? By start I don’t mean just in your career, where’d you grow up? Did you move house all the time? Did you dream of real estate as a kid? What was the path to ending up where you did?

JW: How far back do we want to go?

Let’s go all the way back.

JW: I’m a Midwesterner, born and raised. I was born in Cleveland and grew up in Cincinnati, Ohio. I started my professional career in Austin, Texas, I graduated with a computer engineering degree from Purdue. Thought I wanted to be an engineer.

Midwest and Big Ten, we’re off to a great start here.

JW: Exactly. Yeah, we can talk Big Ten sports as long as you want.

I realized I wasn’t going to be a great engineer, so I got the degree but quickly moved into product management and more on the business side. I worked for a startup in the Dot-com 1.0 days called Trilogy Software down in Austin, which gave me my exposure to the Internet, software, and digitizing industries selling enterprise software and the Internet basically to a bunch of different vertical industries. I was there for about five years, I got my MBA, which the folks at Zillow don’t hold against me, and focused on marketing.

I moved out to Seattle, actually, to work from Microsoft for a couple of years. I worked for Xbox in their consumer group and I worked on mobile really briefly, and then I met [Zillow Co-Founder and Co-Executive Chairman] Rich [Barton] and [Zillow Co-Founder and Co-Executive Chairman] Lloyd Frink, our co-founder, in 2009. Zillow was a couple years old, they had just come out of the Great Financial Crisis in housing and hadn’t found a business model yet but they were starting to grow and they wanted someone to work on product and marketing.

I just fell in love with the team, and the opportunity, and the category. I was not a real estate junkie growing up but I, obviously, have fallen in love with it. It’s a space that’s easy to fall in love with, it’s really a fun space that we all dream about and live in every day, it’s a passion that I’ve cultivated and grown over the last fifteen years here now.

It’s a good point. Everyone has takes on real estate, especially once one has to go through a transaction, that’s for sure. And hey, I just missed you at Microsoft. Are you surprised at, as you look back at Microsoft, I was there from 2011 to 2013, so I left six months before Satya Nadella took over, are you surprised to look back at the company compared to when you were there?

JW: I am. I’ve read a lot of what you have written about Microsoft since we’re both alums. Xbox was a really fun thing to work on but it never quite fit in strategically at the time, and I think you’ve actually written about this a bunch in terms of what Microsoft’s ultimate end game should have been with the device versus just playing in the console gaming space. I am really surprised and impressed at how they have found their way since you and I were there and how they’ve become really great at understanding who they are, and what they offer, and getting aligned on strategy which I think wasn’t quite as strong back in the day.

Just don’t calculate how much the options you left behind might be worth today!

Aggregating Customers

You start at Zillow, you’re there for a long time so 2011, it’s 2024 today. You are just named CEO, when was the first time when you were at Zillow and you had the, “If I were CEO”, thought as far as how the company was operating?

JW: 2009 was when I started and the iPhone had just been unveiled to the world.

That’s right. Sorry, I got the dates wrong but yes.

JW: No, that’s okay. That was the first time I went through a, “scratch the record, we need to pivot where the company’s going” moment, and we’ve done a bunch of those, and I think all great companies are willing to do that, and great founders and leaders are willing to do that. But that was the first time I realized you can get the company focused on any new mission and goal if you set a big — Rich Barton likes to talk about big, hairy, audacious goals — but big mountain climbing objectives and get the team rallied.

That was actually one of my first times where I realized, as a product leader and as someone who’s driving strategy for part of the business, you can have a huge impact and many of our product leaders have huge impacts. They’re like mini CEOs in making the decisions on where we should go as a business, so it was a formative experience for me to get to Zillow where we had no mobile experience, and within the span of six months we had to figure out, not just, “How do we launch an app?”, and the App Store hasn’t even launched yet, but this is probably going to become the default platform, especially for a category like ours where, “On the go”, and, “What’s available near me?”, and, “I want to walk down the street”, is just clearly a better use case than desktop search which is what we were before that.

Zillow has, obviously, been hugely successful aggregating customers. You mentioned when you showed up they were still trying to figure out the business model. Was that an explicit where, “Look, we need to get a big customer base”? What was the focus in those early days? How did you think about the flywheel that would drive your business? Was it just all about customer acquisition then and then you were scrambling to get a business model? What was the path dependency here as far as where you ended up?

JW: It really has always been about the customer, and the business model will figure itself out, and I think that’s been a guiding principle of the culture of the company. The mission of the company when I started was, “Power to the people”. It’s evolved since then but it still has this same ethos of empower the customer, empower the consumer, the buyer, the seller, the renter with great information, great tools and the rest will figure itself out. That’s as starkly true as the founding mission as it was the founding business model.

You’re right, we set out to aggregate demand and to really create useful products before we really figured out how we were going to monetize it. We didn’t have listings on the site for the first few years. We just had the Zestimate, automated home valuations on half the homes in the country, then more homes in the country, then increasing accuracy, increasing products. Then after a few years we figured out how to bring listings alongside that and started to create a local ads business model around it. The culture here is and has always been figure out what the customer needs or wants, figure out a way to help them get it, and then the rest will take care of itself.

And I think it’s fair to characterize a lot of your 2010s as a bit of casting around figuring out what direction, what worked, what was defensible, what wasn’t. You end up with this model of you can contact an agent, and that agent is generally on the buying side. Who is the customer in that transaction then? Is it the customer or the viewer or the buyer or is it the agents that are just looking for someone that’s looking for a house?

JW: It’s a great question because I think any two-sided marketplace has to know who their primary persona is and for us it’s the consumer. The customer is a consumer, and we’re explicit about that in our language internally. We have alliterative names for all of our personas, and we have personas on the demand side: the buyer, the seller, the renter, the homeowner. But we also have personas on the supplier side, on the professional side. We have agents, team leads, loan officers.

But for us, the sorting of the stack is really clear, we build for the consumer customer. The marketplace works when we can create win-win experiences and that was the great thing about our initial ads business model was a high-intent customer, buyer, or seller, raising their hand was what an advertiser really wanted, and the more we could do to help empower, educate that consumer, that buyer the higher intent that lead was for the advertiser who was buying them.

Was there a real challenge though in gauging and grading that intent? Because, of course, you want to make it easy for someone to quickly contact an agent, but then agents get flooded with all these sorts of contacts, then agents don’t respond. How did you balance and tune that aspect over time so that you could maintain this was actually a valuable interaction as opposed to just more noise, more email on the Internet?

JW: And having to do that in a category where most of the customers are first-time customers, even repeat customers, it’s been a decade so the Internet wasn’t even around the last time you did it, so you’re educating and creating a category and teaching a consumer what to do, or a buyer that yes, you can — not just ask a question of an agent you can actually go book a tour the way you might book a restaurant on OpenTable. Then how do you help that person understand what’s going to happen?

So yeah, we have to always balance the supply-demand interaction, and we’ve invested a bunch in trying to help ensure consumers know what they’re getting, for lack of a better word. We talk to most of our consumers, when they start an interaction with us, and go, “Okay, you want to book a tour? Do you have a relationship with an agent yet? Okay, we’ll connect you with one if you don’t”, and that lead qualification, for lack of a better term, really helps educate the customer. “Oh, I didn’t mean to do that I actually just had a question about the house”, or, “Oh yes, I am looking to buy, I am just getting started, I do want to talk to somebody”.

We’re now doing that on the financing side too. Most buyers have no idea, when they get started, how to figure out what they can afford, and they end up in a mortgage funnel having to get a pre-qualification or a pre-approval as a way to understand well, “Where should I even shop? What price range?”, and so we’re trying to help them ask those questions. You have people who are like “I need to get a loan tomorrow”, and you have people who are just learning, and they don’t even know what these terms mean, and trying to help them sift through those different paths on their way to buying a house.

How did that work on the agent side? I think there was a transition you went through in the 2010s where instead of having a wide open funnel for agents you started to really qualify agents and say, “Who’s actually going to be responsive? Who’s going to actually answer the calls?”, what was the thinking that went into that? How important does that remain today to have your agents that you depend on as opposed to agents as a whole?

JW: It starts again from the customer. We looked at the customer experience and said, “Better than half of these folks are not getting an answer”, they’re asking a question, they’re contacting an agent, they’re wanting to go see a house, and we have an advertiser who’s paying for that opportunity and they’re not responsive for many good reasons and for some bad reasons.

So just to make sure, the way it works is a customer, you have agents who bid on a geographic area and then you distribute the customer requests based on who bid the most and disperse it based on those lines. Is that a fair sort of summary of how it works at a high level?

JW: Yes. That’s a good summary of the ads business model which we are moving away from, which we can talk about as part of the Super App strategy.

We will for sure.

JW: But the historical ads-based or market-based advertising model is a set of advertisers, who they make their living in a set of neighborhoods or zip codes, and so they target their advertising, their visibility to buyers and sellers in those zip codes, and when buyers want to take a tour or want to ask a question about a listing they get connected with one of those advertisers who’s paying for that opportunity and wants to make that marriage between that buyer and that agent happen, and ultimately help that person buy a house, that’s how that works, you’re right.

What we found was there is a smaller set of professional agents, usually teams of agents, who had built technology and operations capabilities to actually handle Internet consumers who want an answer quickly. Or, who might want an answer over text message rather than over the phone, or have staff to help with questions in off hours, and all those things that we all — when we press buttons on our phone we expect instant response, but that was new to the real estate industry a decade ago.

So helping those teams build those capabilities, or find the teams that had done the building of those capabilities and can satisfy a modern Internet consumer was a big part of our focus. And you’re right, that led us to really honing in on a set of agents and agent partners that can provide good customer service to customers who ask.

One of my theses is that a lot of these Internet advertising businesses actually end up creating their customers. Facebook is an obvious example here where you have all these entities that only exist because of Facebook advertising. Did you sort of do the same thing when it comes to real estate agents? How many of your old traditional agents actually figure this out? Or was there a new generation of real estate agents that started out with the assumption that, “I’m going to get my leads from Zillow”, and then built their operations around that?

JW: I think some of both to be fair, I don’t know that we can take credit for inventing the concept of a real estate team. An entrepreneur that is running their own business that’s basically an independent contractor at a brokerage, I think that was happening before us. I think we for sure helped accelerate that and gave demand fuel, right, like kindling to the ability for them to grow their businesses.

What makes a great real estate agent in my neighborhood is very different than what makes them in your neighborhood and these are entrepreneurial, in many cases, second career startup motivated folks who have their own way of selling, and advising, and guiding, and it doesn’t look the same neighborhood to neighborhood. It naturally attracted people who wanted growth from technology. Some of that is the team concept and maybe there is some parallel there to your Facebook example, but I think some of that is also just change agent entrepreneurs who saw an opportunity to really rewire how they did their business.

iBuying

I want to sort of bring this up to the current day bit-by-bit. Okay, you started out, you aggregated demand, you had the Zestimates, then you add in the MLS listings so you actually have content there. At some point you have to make money, you get into this lead generation business. Was there a sense internally that okay, we have aggregated demand but we’re pretty far removed from the transaction, and that fundamentally limits our monetization capability? Can you really effectively monetize as an Aggregator if you’re not deep into the transaction and are more pointing people in the general direction? Was there a sense of, “We need to figure out how to get deeper into this to really harvest the value we’re creating”?

JW: Most of our thinking came from the customer side more than the business model side so it’s maybe a similar analogy but on the demand side.

That’s what I’m wondering though, is there a bit where you got deep on the customer side, and then you’re like, “Oh, the business model is kind of over there, we’ve got to figure out how to reel this in”.

JW: I think it was more the transition we went through as a company five or so years ago, which you could broadly say as moving from advertising to the transaction.

We had started by just trying to align our efforts with our advertisers. We said, “Hey, what if we have a shared success model where we don’t get paid until they get paid, and what if we invent all this software to help make sure the leads are better qualified and the people get a good response and follow up and we can help them do that?”, and we kept moving into the transaction and we kept talking to buyers and sellers and you still got, “The transaction’s still broken”. We did a survey at one point that was really seminal in my mind that was like half of all buyers cry while buying a house, like half.

(laughing) That’s not ideal.

JW: Half of the people who go through the process cry. And we all laugh, but we all laugh because we kind of can see that having gone through the transaction. This months long, you’ve got to coordinate with 12 different companies, something can go wrong at the last minute. You got to sign a stack of papers you don’t understand, your agent’s trying to help you but you still have to coordinate with the other side. There’s a negotiation, is it going to work out? It’s your whole bank account, the whole thing.

So we looked around at the business model, the company, the opportunity, and we said, okay, we’ve turned on the lights for the data, like you said, before Zillow and companies like ours, you couldn’t get access to all the listings, you couldn’t get access to what homes are worth. Now the customer’s empowered. A buyer can get a whole lot more off their phone before they even have to go set foot in a house.

However, once they click the buy button, it’s all still broken and so that was our way into, “Wow, we have an opportunity to fix that”. Of course the TAM on fixing that is even bigger, but it started from the customer side and the, “Okay, well what problem hasn’t yet been solved and why aren’t customers getting what they want? Why is it so hard to move? Why is it so expensive to move? Why does it break so often? Why do so many people cry?”, that’s really what led us down this path.

Well, the initial foray into this was obviously this iBuying thing. Rich Barton has claimed that he was inspired to return to the company and dive into that by the idea of, I think I wrote about Opendoor and this idea of just fundamentally transforming the business. I don’t know if he was just kissing my rear end, but he said his response to reading that was, “Hey, we got into this business to change it and we’ve just sort of solidified the way that it works”, which is this buyer/seller agent taking 6%. And “Yeah, we’ve empowered the consumer to a degree, but my sense was we’re not into the transaction.” Meanwhile, you had Facebook nipping on your heels as far as advertising and stuff goes, lead generation.

Why go all the way into doing it all, when you think back to that decision? And I want your perspective on the unwind of that approach. But was it, “Okay, the real value’s in the transaction”, and the most audacious way to fix that is to take the transaction itself? Was it feeling a real threat from Opendoor? What was the sense of motivation when you made that big switch?

JW: Yeah, I think it was two things kind of in parallel. It was that we recognized we wanted to go solve the transaction problem. And I think Rich, when Rich came back, it was really to, like I said, for lack of a better framework, move from an advertising business to a transaction business because that’s where we could really affect the most change if you looked at what hasn’t yet gotten better. Why aren’t the transaction economics better for the customer and for the industry? And then at that same time, we saw maybe a short circuit to fixing the transaction.

Just do the transaction yourself.

JW: That’s right. We fell in love with the idea that market-making could just get us there quickly. There’s all these problems with the buyer, buyer’s agent, seller, seller’s agent, all the things behind the scenes, this traditional arm’s length real estate transaction. Inserting a market-maker means we could fix it and delight the customer a lot sooner. So we ran it at the same time we were moving strategically down into the transaction.

So what went wrong?

JW: The fundamental assumption that you could market-make is what went wrong, and that was a painful lesson that we learned and that’s the price of innovation. We did a big swing on the way in and I think as you wrote about, we did a big swing on the way out once we realized our assumptions were wrong. That the assumption for iBuying was you could make market rate offers because you could market-make, and that would delight most customers and they’d prefer that service and it would be equivalent to a traditional arm’s length transaction.

It turns out you can’t do that. You end up being a counterparty to your customer, which we didn’t love. We didn’t love having to negotiate with a buyer/seller or tell everyone that their home wasn’t beautiful, and there’s a real value in paying a fee for certainty or speed, but you have to give up a little. It’s not price maximizing, and that’s what we learned in trying to run that business too, and that’s okay.

If you remember Rich, when he came back, talked about we thought this could be a really big way that people buy and sell homes because it could be market-making, that’s what got us interested and we learned for ourselves that it’s not market-making. It’s a trade-in service, which is a really valuable one and we offer that now with a partner in Opendoor, but it’s not the primary way you’re going to be able to buy and sell a house, and so we moved on from that and took a lot of learnings from that and that gave rise to this Super App strategy, which is actually trying to digitize the traditional arm’s length transaction, which is a lot maybe harder.

Just one more question on the iBuying before we get the Super App strategy. I do think it’s interesting, but I think the context of how you ended up here is fascinating, plus I haven’t talked to anyone from Zillow since this all went down. But where do you think about the balance though? As you’re looking back, you’re the new CEO, you can throw Rich under the bus, he’s not going to listen to this, it’ll be fine. Between Zillow’s making mistakes or Zillow maybe overestimating its data capabilities versus just the iBuying concept being fundamentally flawed. Like you said, you’re still offering that now in conjunction with Opendoor, which was viewed as a threat. So where do you think that balance lies?

JW: I mean, I think at the time when we decided to wind down, we were having our own operational challenges, and everyone else hadn’t yet seen the challenge with market-making, price forecasting in all scenarios and being able to offer market rate pricing. So we learned that lesson and I think at the time, maybe we were the first ones to learn about it. I think since, everyone has learned that lesson and what looked like iBuying as market-making back in 2021 now looks like more like a trade-in service.

Is there also a bit where you had a bigger business and to add in this super high risk element made it, it was sort of a conglomeration in some respects that didn’t really make sense. Whereas you have Opendoor, that’s all they do, so investors know what they’re getting into. Was that a real consideration?

JW: I think you wrote at the time when we wound down that we could afford to make that decision more than others could, and I think that’s still true. We said really clearly we thought the business model would work differently than it did, and we had the fortunate position of being able to learn that lesson as a painful lesson and move on and move towards a better way to work on the transaction while having a great aggregation business. Other companies that were single-purpose built had to figure out a way to operate in iBuying even in a world where iBuying is no longer market-making. So that’s right, our path is different than theirs for sure.

The Super App Strategy

You mentioned that you were shifting into this, “We need to get into transactions”, before iBuying came along. So my thoughts about this was you got into iBuying and then you realize that’s not a great business, but there’s all these pieces around it that you can get into. So do I have the order of operations wrong? You wanted to get there, went all the way, and then this is still the strategy that drove you to iBuying in the first place? I’m referring to the Super App strategy. You’re getting into mortgages, you’re getting more into listings, all these sorts of things. But really I would frame it as getting into the transactions.

JW: Yeah, the order of operations internally as a shorthand was moving from ads to transaction. As we started to do that, saw iBuying. Said, well, that’s maybe the short circuit to build the transaction rewiring. Learned that lesson, and now are building, trying to build the software and operations to digitize the transaction more traditionally, which requires Zillow, Zillow customers, agents, loan officers, software for people that use Zillow and don’t use Zillow and the whole thing, which we can get into. So it was definitely an evolution.

Well, give the Super App pitch. What is the overarching strategy, the shift away being an ads business to what you’re doing today?

JW: Yeah, the Super App pitch is that if I could give you a remote control on your phone so that the same way you can order an Uber or book an Airbnb, you could do all of your real estate transaction inside of the Zillow app. You could talk to your loan officer and your agent 24/7 via group chat. You could understand via a pizza tracker where your loan is and are you missing any documents and it could all be in one place.

It would be easier, it would be simpler, and most people would pick that and most people would want that versus the analog process, which is wet signatures and tons of back and forth and tons of risk and tons of repeat data collection and tons of cost and not the ability to communicate and be in control via the remote control via your phone, so that’s what we’re trying to build.

Obviously that requires a lot and it requires a lot both for the customer and for the supply-side of the marketplace. But that’s what’s led us into mortgage, to your point. That’s what’s led us into trying to build virtual and on-demand touring, when you go book a tour to see a house or virtually tour a house with our technology. That’s what’s going to lead us into title and escrow, which is something we’ll build as part of the transaction. That’s what’s led us into agent software.

So we’ve bought a few companies and are integrating companies so that the agents on the other side of that Zillow glass that you’re using are able to converse with you and work with you on Zillow while they’re using their CRM. So that’s what we’re trying to pull together to just give you a better buying and selling experience.

Data and Real Estate

Do you think there’s a bit where in of the long run up to this point where you under-invested in getting your own data? Whereas you had the MLS listings that got a lot of data for free and that just plugged really nicely into your aggregation model. It was almost Google-esque where they just provide the interface and everyone sort of plugs into them. What made you realize, no, we have to actually go out and source data? Not that you’re taking pictures per se, but we have to build tools so that our agents actually upload stuff and put that in there and then you get these listings that are special and unique.

JW: It’s interesting to have worked on a marketplace where the core data set is commoditized, to your point. I think maybe most of your listeners or readers don’t know that in the US real estate market, which is very unique from the rest of the world, there’s cooperation between all the brokerages and agents so that everyone can see everyone else’s listings and there’s a huge benefit to that. You and I can just log into Zillow or any real estate app and see most of everything that’s for sale for free. In most other countries, you don’t get that. You have to go call someone or duck behind a velvet rope to see what’s for sale or see prices or any of that, so there’s this huge consumer good that comes from that.

What that meant, exactly as you said, was we didn’t have a unique data advantage to build our marketplace, the listings were commoditized. So it really sharpened our focus on customer product, everything from the Zestimate and data like derivative data around the core listings that drove the marketplace, everything from great tools and recommendations and AI capabilities that we built to try and make the content more engaging for you. We did a bunch of stuff not in the shopping experience to create the brand as well, which we can talk about and now we’re using a lot of software and AI capabilities to build, as you alluded to, what we call Super Listings, like a more immersive listing than is traditionally shared in the industry.

Which are amazing, by the way. My favorite is where you have the floor plans and you can just click on a button and see it there. It drives me up the wall any real estate listing that does not have a floor plan is just, it’s like, what are we doing here? I’ve been tried to figure out this house and I have no idea where I am. So a big thumbs up to that.

JW: Exactly. Well, great, and we can come back and talk about that. That’s a technology innovation that we’ve been working on for five plus years. That started with pre-generative AI craze. We would call it machine learning and computer vision, and we hired bunch of folks-

Yeah. What’s entailed in creating that? I mean, how much extra work is that for the agent?

JW: Yeah, so that is called Listing Showcase for those that are listening and it is, you could think of it like a Super Listing.

So a traditional listing, which a photographer goes out and shoots 25 to 50 photos of each room, and then you write a description and you put a beds, bath, and price and you upload it to MLS and that’s okay circa 2006. But as you just said, we all want to really virtually tour our house as much as possible before we can, you have to capture a little bit more data to do that.

There was a bunch of technology coming that was super high-end, dedicated hardware, set it up in every room and it would kind of do a whole bunch of really high-res panos, and you could do this kind of virtual tour thing. The problem with that is that in real estate, no one’s going to spend the money on the capture, you have to make the capture free fast and easy, you have to get to the level of like a hundred dollars 360 camera or even the new iPhone camera. That’s what we’ve spent a bunch of time working on, is you can now create the showcase listings with just an iPhone and with just grabbing a few panos in each room and then we do a bunch of photo location.

We do a bunch of AI human-in-the-loop assisted annotation for the floor plan and you end up with what you’re describing is an interactive floor plan where all the photos are on the floor plan, and you can see on each photo which way it’s pointing in the room, and you can walk through the house and generative AI stitches all that together to make it feel like it’s a 3D walkthrough. So we had to solve this problem of how can you get that technology off of the fidelity of media that a traditional real estate photographer will already go get.

But the tension I see from a business model perspective is what you need is for the seller agent to do it — you’re trying to make it just as little work as possible, but they need to do more work. But you’re coming from a background where your traditional business model is in service of the buying agent. So how did you think about resolving that it’s worth it for the selling agents to make your listings better?

JW: Well, one thing that everyone I think misses about the industry is that most seller’s agents work with buyers and most buyer’s agents will have listings. So they’re constantly, if you sell an agent on the capability, they’re going to then want to use it for their entire client base, and so that helps.

But also all sellers are buyers, or most sellers are buyers and so when a seller, like you just said — you just said, well, it drives me crazy that listings don’t have this — well, when you go to sell your house, ideally you ask your listing agent, are you going to buy or use a showcase and then when buyers see it, those buyers are selling houses. So you want the flywheel to come from the buy-sell interaction.

Now it’s early, we’re on 1% of listings or 1% of new listings nationwide. We just launched this product less than a year ago, but we are seeing that sort of flywheel go. In the markets where we have a little bit more penetration, we’re seeing anecdotally buyers requesting it. We’re seeing sellers request it from seller’s agents. We’re seeing seller’s agents who use it on their first listing see it attract and get the home sold faster and then be converted and then go start pitching it to the rest of their sellers. So again, it’s early days, but it is starting. But all the way back to your question, all of this investment in technology and product was on top of basically a data set that is commoditized.

Right.

JW: So there isn’t a unique differentiation on top of the data set. I think that actually sharpens your focus on ensuring you’re solving great customer problems because you can’t just rely on unique listings, unique content to fracture the marketplace.

Well, speaking of unique listings and content, you’re facing a threat from CoStar and homes.com, and they’re like the data kings, that’s sort of the core of their businesses, is data on real estate, mostly on the commercial side, but they had big success with apartments.com. Was there a bit where because the market was so commoditized as far as data goes, that there was maybe a weak point in terms of having unique data, or do you see that as sort of being — you just clearly said, “I don’t think that’s a big deal,” and yet you’re investing in unique data. So where is the balance there?

JW: I hate to sound like a broken record, but we always work back from the buyer and the seller. Is there data missing to help the buyer get more educated and maybe be more transaction-ready, make a better decision, or the seller help market their home better? So that’s kind of what led to Showcase, was, “Okay, technology can actually help the seller sell their home better”. It’s not just about advertising or trying to have content to drive audience, it’s actually about Showcase listings sell faster and sell for more money and it’s not just because only a few of them, it’s because they actually get people to spend more time with them, and they get a better sense of the house.

So we thought about data and content that way. It’s more about how do you make the actual listing better? That’s true with things like the Zestimate too. It’s like the Zestimate is the best automated valuation model there is, it’s still only as accurate as automated evaluation model, it’s wrong all the time, but it’s a really good starting point to figure out, “What should I offer on this house alongside the listing price?”.

For us, it’s about content around the listing and around what the decision is in the shopping period, and then we can talk about all the things you need to do in the actual buying period. That’s where we’re focused a lot right now, is once you want to kind of virtually click the buy button, you have to figure out financing, you have to actually acquire a mortgage, you have to actually figure out what to write an offer on, you have to actually go through the transaction, and that’s where a lot of the data and the help for the customers still is not online and not digitized.

The NAR Settlement

There was this major court case, the National Association of Realtors settlement that ruled that the forced tying between selling and buying agents was illegal. However, the injunction against that collusion only went into effect in August. What’s your high level take on the decision, and the way you are thinking about it and the opportunities or risks it presents to Zillow?

JW: Great for consumers and will be great for good agents is my high level take, there’s a lot that’s been written about it. And you’re right, we’re finally having all the changes go into effect, and so you’ll kind of see these play out in practice. For folks who are listening that don’t follow the real estate industry that closely—

Give us the overview, you’re the expert.

JW: The two high level changes are both about buyer and seller choice and education.

So on the buy side, historically a buyer didn’t have to sign an agreement with a buyer’s agent. A seller signs a listing agreement to sell their house, but many buyers did not sign buyer’s agreements. There’s many states where they do and there’s many great agents that are already doing it, but it wasn’t really default, it wasn’t required. Now, every buyer has to sign an agreement at some point with their agents, that I’m hiring you, you get paid in the transaction, I understand that, here’s what you get paid. That’s great, because most buyers, if you ask them, didn’t understand how everyone got paid, how it all worked, and that’s not great.

And then on the sell side, there used to be a mandatory sharing of compensation between the seller’s agent and the buyer’s agent, the seller didn’t get any choice in that matter. And so now, that’s optional. That can’t be displayed on the MLS, but it’s also a choice that the seller has. One would argue the seller always had that choice, but the mandatory MLS rule made that choice really hard to actually have a choice. So now, the can offer to share commission can say, “Hey, write me an offer”, and I know there’ll be a concession in there for the buyer, and I’ll just negotiate it when I get an offer, they can choose how they want to do it.

So in both cases, on the seller side, on the buyer side, it’s a more empowered customer, and it’s a more educated customer and we can come back to what we’re doing, we’re leaning into that education. On the buy side, we are innovating a touring agreement, like a dating form because we want to help educate customers before they get married to an agent. Well, what are you going to get into? What should you expect? What should a contract look like? And help them be informed about their choice before they make it, so that’s what changing.

Is there a risk though? Generally aggregator type business really works well when there’s no friction, and I think the question I have is a lot of these discussions is all about, “Yes, we want to make the experience better, but that comes with introducing friction”, that could be friction as far as the seller agent, and taking better pictures, and using this new software. In this case, if you have a buyer, if they want to actually see a house, they have to sign an agreement. Just as a anecdotal, using Zillow, I’m always wary of clicking that “Contact Real Estate Agent” because I don’t only get 47 phone calls over the next 24 hours, but I have the opposite problem, I want less contact and not more in prompt contact. But now, also if I have to sign a contract, I don’t know about that. Is there a concern that this is grist in the gears of a system that’s worked very well for you?

JW: Well, first of all, you don’t get 47 home calls if you book a tour with us, I can promise you that. You will get an outreach from Zillow, making sure you wanted to book the tour at time and you didn’t do it accidentally, and then we’ll connect you with an agent.

Sorry, I was being hyperbolic. Yes, good call out and correction, duly noted!

JW: No, no, you’re good. And the reason I kind of point that out is because for a long time, we’ve been focused on quality over quantity in the funnel, and we’ve done a lot to add friction.

Yeah, as we referred to earlier, and that was a shift that was like a decade ago, so that would be a little unfair on my part.

JW: Yeah, what we found in that though is that when you trade quality and making sure the customer gets what they wanted, and you weed out the ones who did it accidentally, you get much higher conversion. So the lead volume or the connection volume might go down slightly, but the overall conversion rate goes way up, which also, back to your supply/demand challenge, if we can only hand touring agents who are purpose-trained to do tours, customers who really want to take a tour, they’re going to be really good at satisfying that tour and win those customers over, and you’re not wasting their time with people who don’t want to do that.

So the touring agreement, friction, we look at as really helpful friction because it’s not — none of these things dramatically decrease the volume. There are slight speed bumps, but those slight speed bumps, make sure you’re weeding out the people who didn’t really want to do what they didn’t want to do.

How do you solve the problem of signing an agreement with a buying agent when you don’t even know yet what the selling agent, because you don’t even know the house yet, what their terms are going to be?

JW: So that’s the big reason we leaned into innovating here, was the framework we use is we think every buyer should sign a dating form before they get married, and we were concerned that some others in the industry might try and jump right to the marriage form, to an exclusive agreement, right? “Hey Ben, I just met you, but you’ve got to sign this form before I open the door, and it says that you’re going to work with me no matter what for the next two years” — that’s not good, right?

So we pioneered, effectively, a touring form, more of a dating form. It says, “Hey, you can book tours. I can take you on tours and there’s not a fee for that service. If you choose to write an offer with me on a house, then this is what my fees will be, and you can sign another exclusive agreement later”. It’s more of an education disclosure form to make sure that the customer has a choice and most customers should talk to more than one agent before they hire one.

So for us, we look at it as really good friction because the alternative, just jumping to the marriage form, would not be good for a buyer or seller and I would think most reputable agents wouldn’t want to do that. But not saying that some wouldn’t try to use the agreement to kind of convert business too early.

I know you talked about shifting from the ad model into getting deeper to the transaction, but the ad business is still an important business. Does this decision threaten that to an extent, because to the extent you’re getting paid by the buying agent and the buying agent got paid behind the scenes by the seller agent, now if that connection is severed, your monetization engine is on the wrong side of the ledger?

JW: We don’t see it working that way, and we haven’t seen it work that way in practice. The change that we see happening is maybe just more dispersion of fees across agents. When there was a unilateral offer, everyone just got paid whatever the market rate was for that listing sharing, and now, different agents can negotiate different service levels.

What typically happens is good agents are really good and they’re worth their money, and maybe folks who are not as good, and now have to have a conversation about what services cost and what their value is. They may lose out, or they may do less deals and that’s what we’re — it’s still early in the data, but that’s what we’re seeing in the data. We’re seeing, if anything, just dispersion, and we expect really good agents to — it’s possible they’ll grab more share here because before the conversation with a buyer was, don’t worry about it’s free. Now the conversation with the buyer is like, “This is what I do for you, the reason why it hire me is I know the neighborhood, I will help you write the right offer, I will help you understand the specifics of the physical challenges in this neighborhood or in this, what goes right and wrong with these types of homes, I will help you win”. That’s a valuable service, and the people who offer that service get paid for it and the people who are just like, “I don’t know, I’ll open some doors to you. Don’t worry, it’s free,” they’re going to probably lose out more.

Agent Empathy

I think just zooming out, and this sort of ties a lot of the stuff we’ve talked about together, if you think about a real estate agent, my perception is the person that is — maybe this is the wrong assumption — that is more “desperate” and desperate is not the right word, but motivated to some extent, in the market is the seller because they have one house to sell, they need to sell it. Whereas a buyer, they have a lot of houses they could buy, and I think one of the things that made iBuying attractive is, at least in theory, it was a very solid proposition focused on the seller. Like, “Look, you are going to unload your house. You need to move country, move cities or whatever it might be, or get a bigger house, there’s a guarantee and a finality that this will be taken care of, we have to figure out the price, et cetera”, and yeah, we talked about why that maybe didn’t quite work out.

However, as this market bifurcates, I guess Zillow ended up more on the buyer side, particularly with these sorts of things. Is there a real push to change this? This gets into you want better listings, you want these sort of things, that is the seller agent. Do you need to really get into that transaction where listing agents go to Zillow first and that is the problem you need to solve and that needs to be your customer. You say you focus on the consumer, but then your business model was sort of the buying agent and maybe the seller and selling agent, maybe that was the correct motivation for iBuying. Maybe that wasn’t the right strategy, but there does need to be a shift in focus on that side of the transaction?

JW: I think we’ll always focus on the seller and the buyer first, and the helping the agent solve the seller’s problem and the buyer’s problem kind of comes naturally from that. And we, for sure, when we got into iBuying, it was about — we found this great tool for sellers, to your point, it maybe solves a lot of problems, it maybe was too good to be true.

Right. They solve our problems, we solve them and take on the risk.

JW: But there are still a lot of needs for the seller. And again, the big part on this is most sellers are buyers. So yes, you have to get the listing marketing for your house, and you’re going to hire an agent to sell your house, but you also are then going to go buy another house and you have to actually go be a buyer and go figure out what you can afford and which mortgage you’re going to qualify for, and you have to actually go spend your Saturdays booking tours and taking tours, and you have to figure out what to offer and hire an agent. So you still have to go do all those things on both sides.

So we are for sure working — now that we’re doing this more broad Super App strategy, which we think appeals to the traditional transaction, we find ourselves investing more not just in the seller and the buyer, and you asked the question specifically about seller’s agents, we actually just find ourselves investing more in the agent. I think that’s a big shift over the last five years, has been the recognition that we can’t actually get the transaction better for a buyer or a seller if we don’t actually help with software and tools, not just for the broker or the team lead that’s running the shop, but actually for the individual agent and we’ve spent, I think north of a billion dollars buying and now integrating a bunch of software to help both listing agents and more touring and buy-side agents do their jobs better.

Maybe that’s a bit where, yeah, you were super focused on the customer and you aggregated a lot of demand, but at the end of the day, the customer, maybe the focus needs to be a little bit more on the agent, and that does feel like the fundamental shift that’s happened.

JW: I think that’s right. I think a better way to think about it is less the business model shift, it’s more if you want to build this Super App, you have the buyer on one side using the Zillow app, and if there’s no one behind the screen, able to connect with them and answer the questions and do things for them, it doesn’t really work, and it was this recognition that, okay, well, to build a great buyer experience, to build a great listing experience, we have to go and work with our agent partners and go, “Well, why aren’t you able to do this more, better, faster?”, and it’s, “Oh, well you, you’re doing it with pen and paper, or you’re doing it doing drive time in the car, or you’re talking to the wrong customers, or you’re not being reminded who to follow up with, or all these things that CRMs can solve, and workforce management can solve, and schedule management can solve”. So yes, we got a lot of empathy for the agent side of the marketplace.

Right, that’s my question. What drove that empathy? Because I feel like you’ve spent a very long time absent that empathy to a certain extent. Maybe it’s like the MLS data delivery made things so easy and you had the Zestimate delivering consumers, and it feels like the early business model was easy, and what you’re doing now, why did it take until the 2020s to do that? Was it the pain from the iBuying experience that you opened your eyes to all these issues? Was it potential competitive threats? What you’re doing makes total sense to me, this shift makes sense. What I can’t figure out is, why didn’t this happen a decade ago? Why now? Why are you now going in that direction?

JW: I think it was just a maturation of the company phase, right? The first ten plus years were bringing the Internet to real estate, and we were excited. If you think about — go back to what kind of BHAG [big, hairy, audacious goal] did we call, internally, what was the big rallying cry? It was get all the listings, and then it was raise the awareness with customers. We still only had X customers, how do we get to Y customers? How do we get more people? How do we teach the category? When we started advertising, no one knew anyone in the category, they couldn’t name a single real estate brand, us or anyone else because they didn’t realize you could do this.

Was there a bit where as you did it well and you won, that it was easy to not see the next step? You see it now, but maybe it took a little bit longer than it should have.

JW: We were definitely squarely focused on empowering people with information. If you go back and read what we probably said publicly back then, it was information marketplace and it was more tools for customers to shop and to dream and to start their decisions, that’s what we said. Our verbs were dream, shop, triage, personalize, all those type of things, and now our verbs are buy, finance, and sell.

So yes, it was. We got to that point in the technology comes to the category and we said, “Well, this is great, now you can do all this on your phone and you have the app and all that”, and then like I said earlier, we started doing research on the customer and what problems they had and we still found that okay, but once you click the button, there’s a record scratch and it’s too hard and most people can’t do it. And they cry and they fail and we’re like, “Woof, okay”, well, that’s a way bigger problem to solve.

Actually the way to meet our original goal of making the consumer problem better is we have to go in through the backside.

JW: That’s right.

MLS Scenarios

With this MLS thing, you mentioned earlier in passing, it wasn’t technically mandated that buyers and sellers cooperate on the commissions, but it was a price of admission to MLS and MLS being the fabric that tied this all together, everyone had access to the same listings, a huge asset to Zillow previously. Is there any concern that MLS starts to fall apart and there isn’t this widely commoditized hose of data that everything gets, and what is your contingency planning and thinking about that in the long run? Or is that what is motivating this push into getting your own and better data? It’s not just that it’s better, it’s that you need data sources.

JW: Gosh, we hope not. We are lucky as consumers in this country that we have this kind of cooperative data sharing. It would be a huge step backwards for buyers and sellers if it went away.

Is there a risk though if you’re setting agents really against each other, they have to win business?

JW: I think one of the big reasons it’s not a risk is because most every agent, like I said, works both sides and most every broker is brokering a bunch of agents that work both sides, and so they all get the benefits of give-to-get on listings. There’s just a huge amount of good that comes from that. So even with the changes that remove compensation from that part of the marketplace and start to create more customer negotiating directly with their side of the transaction, those agents still we hope see all the benefits of the cooperation.

Again, we are strong advocates for it even though it forces commoditization on our marketplace and others, because there is such a good consumer good, and in a world where that went away, you’d end up with pay-to-play marketplaces and you’d end up with, you couldn’t go one place for all the listings, and people would end up having to spend more, you’d have to spend more marketing your listings on all these sites.

You can go look at other countries. The way Australia and these other real estate marketplaces work is you actually end up paying more per listing to have go to market and the customer is not nearly as in control, so we are very hopeful that cooperative element stays because it is such a good — not just for the industry, it’s actually really good for buyers and sellers. Sites like Zillow and others can create all this free availability of content here that you can’t do anywhere else because of it.

Is there a possibility that this is one of those classic Chesterton’s fence sort of things where this “collusion” is torn down and all these advantages that came with it going, to your point, the other countries, you pay lower real estate fees, that is a payoff, but to your point, does that actually drive up the price as a whole because there’s less information available? It’s a hard counterfactual to really tease out given how different real estate markets can be for very obvious reasons, but is this a theoretical risk or is it something that you really have an eye on or concerned about?

JW: I think we think it’s more theoretical. Again, back to the changes that ended up happening are much more evolutionary than revolutionary because they are really about let’s make sure the customer has choice and let’s make sure there’s not something mandated. There were versions of this that could have been an outright banning on cooperating at all on any content and I think that it would’ve been harder to have free, fair listings access for all in a world where there was a lot more prohibition.

If your goal is there’s still cooperation in sharing listings, that basically there’s still going to be cooperation in fees as well and we end up with actually very few changes at all.

JW: The hardest thing we had to navigate during all this period of uncertainty was people doing the pattern match to other countries for those listening that look at real estate in other countries and going, “Well, but Zillow wins more in those markets, so why wouldn’t you advocate for the market to go backwards because your profit margin will look better?”.

Exactly.

JW: Our argument is again, we’re here for the customer and there is some real good in this marketplace and you can build a lot of good and we can grow our company really big in a world where there is listing sharing and there are all these benefits and all these things we are doing to grow our company to — you’d move back to a classified business model which isn’t really good for the buyer and the seller. So that’s the big part of why we’re like — but it was this weird argument where we’re talking to investors about like, “Well, yeah, the margin profile of that business might look better, Zillow would probably win more, but we don’t actually think that’s good”, it’s not really good for anyone. It’s definitely not even good for the suppliers who would in theory, enact it.

Rentals

Why is the rentals business doing well? This is the opposite question to the CoStar question earlier. You’re making inroads on their bread and butter as well. What’s working there?

JW: Our rentals business is maybe a cleaner version of your Aggregation Theory than even the for sale one, because as we had to talk about, there’s this tortured data construct inside it. There is no national database of all rental listings, and our strategy for rentals, which is pretty unique, has been to organize as much of the rental inventory as possible into a version of a national database. We’ve been working on this for maybe ten years and for those listening, the reason that’s a hard problem is yeah, there’s 50 or a 100,000 big buildings out there that are professionally managed, there’s a lot of data and content advertisers, but there’s four plus million. We call them long tails.

It’s a long tail, yeah.

JW: Right, yeah. Single family duplexes owned by your uncle, a dentist, yourself, you may have one or two, and those landlords, it’s not their day job and they only turn over once every three to four years and a decade ago they used to use Craigslist.

And they don’t want to pay a subscription to run ads.

JW: And they don’t need to spend money on that either, what they need is quality tools. So we spent a long time, years organizing that supply, and we now have the most listings for sure, we have a good chunk — the majority I think — of the long tail listings and we have an increasing share of the multifamily listings, and we did that again for the customer.

What the renter wants is as much of the inventory as possible. We don’t want to have to open ten apps, you still have to open multiple apps, it’s impossible to get it all, but more and more renters realize, “Oh, Zillow has the most, and I should spend most of my time on Zillow looking for rentals”, and we paired that with tools.

So for a bunch of those folks, we offer them a portable application and you pay once for one credit pull, and you can use that application at a bunch of properties over a 30-day period. We offer landlords the ability to create custom leases for their state. We even take payments, so a landlord can ask Rachel — sorry, the renter, we call the renter, Rachel, internally.

You gave away one of your personas. Yes, we now know what it is.

JW: I did, yeah. All of our personas have alliteration. The buyer is Beth, the seller is Sebastian or Susan and the renter is Rachel.

That makes sense.

JW: But anyway, that platform is what attracted all the supply and that’s what then attracted the audience. So the more supply we got, the more audience we got.

So it’s interesting, it turned out it’s basically been the opposite of your home business where you started out with the demand, I guess because you got supply for free to a certain extent via MLS, but this you had to go in the opposite direction. Does that feel like, in some respects, it’s not as big of a business but a more cohesive business because you started with supply?

JW: I think the word cohesive is a good one, you’re right. On the for-sale side, we still started with supply. It was just supply was things like the Zestimate and other content to attract the audience.

On the rental side, yes, we sought out to organize as much of the listing content with listing tools and homeowner outreach, and then that’s what led to demand, and I think that’s a tighter reinforcing loop for sure. We are seeing the flywheel of every year, folks who come to Zillow to post their listing goes up, and then folks who come as renters goes up and our audience lead is increasing while our listings lead is increasing and so now you’re seeing the monetization come.

The biggest miss on the rental side in that flywheel was the professional segment. They had a bunch of places they could just go get bespoke leads for revenue management, and as we’ve gotten big, we now are a pretty large source for them, and they’re really interested in bringing their content to Zillow. So I think it is a more tightly integrated flywheel for sure. And there’s an opportunity, I think we put out it’s close to a $500 million run rate business and we see that’s a billion plus dollar business. So it’s not small, it’s now a fifth of our company revenue and it’s still growing really fast.

Well, I find this market so fascinating for all the different reasons, particularly this combination of aggregation and the question of which side of the equation you are. We’ve actually run out of time, we didn’t get into, you’re doing these AI listings, you have a Vision Pro app, which by the way seems such an incredible use case for this, but talk about a problem and getting a flywheel started as far as supply and demand goes. But I appreciate that it exists, so thank you for building that.

JW: Absolutely. We jumped on almost all new platforms when they come, we jumped on the mobile phone early, we were lucky enough to get asked by Apple to jump on the iPad early. You knew the Vision Pro would be a small market initially just given the price point, but exactly to your point, it’s such a fantastic reference use case demo of Showcase.

So for everyone out there has a Vision Pro, you can go get the Zillow app and actually try it and it’s that interactive listing that Ben talked about, and you can then just actually see it 360 and it’s kind of crazy. So yes, someday there’ll be more of that platform out there in the hands of more customers, and that’ll be a really fun way to browse real estate.

Jeremy Wacksman, it was great to talk to you. It’d be great to touch base in a little bit and see how all this plays out. I think it’s going to be very interesting.

JW: Absolutely. I’d love to. Really enjoyed it, Ben. Thanks.

---

This Daily Update Interview is also available as a podcast. To receive it in your podcast player, visit Stratechery.

The Daily Update is intended for a single recipient, but occasional forwarding is totally fine! If you would like to order multiple subscriptions for your team with a group discount (minimum 5), please contact me directly.

Thanks for being a supporter, and have a great day!