An Interview with Opendoor CEO Eric Wu About Building a Marketplace in a Real Estate Slowdown

Good morning,

Today’s Stratechery Interview is with Eric Wu, the founder and CEO of Opendoor. Opendoor is very much in the middle of testing the thesis that homebuying can survive a market downturn. The company had huge losses in the third quarter as it sold inventory at steep discounts and discounted the value of what it still had in stock; the company also laid off 18% of its workforce.

At the same time, Opendoor is also seeking to leverage its first-party buying product into a new marketplace called Opendoor Exclusives; from the company’s blog post:

We’re aiming to make the sales experience as easy as selling to Opendoor. Homeowners can request an offer from our network of buyers, on top of an Opendoor offer, and sell in minutes without the need for repairs, extensive home prep, or months of open houses or listings. The experience end-to-end is designed to put home sellers in control, with no upfront commitment, flexibility to close early, and certainty of close…

With Opendoor Exclusives, each of our homes comes with transparent, buy-it-now pricing and an Appraisal Price Match Guarantee, meaning if the home appraises for less, we’ll bridge the gap for buyers up to $50,000. And our buyers save up to $10,000 on every home. These features dramatically simplify the process, give buyers more confidence around paying the right price, and make purchasing more affordable.

Our belief is that by building a better, more efficient, and more transparent system, we can improve the outcomes for both buyers and sellers. Marketplaces have the unique ability to layer in more trust, simplify the process, and eliminate much of the complexity, opacity, and cost. We intend to do just that for real estate.

I had a chance to talk to Wu about both his and Opendoor’s origin story, competing and now partnering with Zillow, why the speed of the current slowdown caught Opendoor by surprise, acquiring customers for an infrequent transaction, and how the company will get a marketplace off of the ground.

To listen to this interview as a podcast — which I recommend! — click the link at the top of this email to add Stratechery to your podcast player.

On to the interview:

An Interview with Opendoor CEO Eric Wu About Building a Marketplace in a Real Estate Slowdown

This interview is lightly edited for clarity.

Wu’s Real Estate Background

Eric, it’s good to have you on Stratechery. Opendoor is a public company now, so I do want to get into more specifics about the business than we might with a private company, but like I often do in those conversations, I want to talk about you first. Tell me more about yourself. Where’d you grow up? How’d you get into technology? Give me the Eric Wu origin story.

Eric Wu: Ben, excited to be here. It’s good to make the connection and conversation. It’s been a while since we chatted live. I grew up in Phoenix, Arizona, and what you may not know is that one in three jobs are real estate related in Phoenix. Maybe there’s something in the water. But when I got to college, I started getting into just hacking and building some websites. I built an apartment rental website, a college job board website. At the same time, I ended up buying my first home when I was a sophomore and also became fascinated by the ins and outs of the real estate transaction. So, by the time I graduated, I purchased about 20 homes, plus or minus.

Did you own all them or you were buying them and flipping them?

EW: I bought them and then rented them out.

Got it.

EW: But I did the hard work of renovating the homes myself. I often joked that back then I was living at Home Depot trying to learn how to lay tile and fix plumbing and this and that. It was a pretty fun time. And so, yeah, was actually fortunate enough, as a result of doing some of the website building stuff and the real estate stuff, to not have to get a job after college, which was really the impetus for getting into entrepreneurship and starting companies. And so, I decided to marry the two passions, which is tech and real estate, and start a few companies. The first one was Rent Advisor, which was a review site for locations, apartments, and landlords. And the next one was Movity, which we aggregated location data like crime and commute times and provided insights. But the backstory is that one of my first angel investors was Keith Rabois, and so he spent about a year trying to convince me to start a Priceline for homes idea or join Square.

Pick one. A or B.

EW: Yeah. You can do one of these two things, Eric, but you can’t continue on this current path. I actually didn’t do either — not sure if that was fortunate or unfortunate. Instead we sold to Trulia in 2010, and it was really there, actually, that I was able to get really deep into understanding the consumer experience around buying, selling real estate and the problems and solutions.

This is pre-Opendoor, and I do want to get into some of the lessons that you’ve learned over the last few years. But as you were dealing with this as an amateur home buyer, renter, and then working in this market and trying to build these review sites, what was your prevailing view of the market, why it worked the way it did? And I guess if you can remember, what was it then and what is it now? And what lessons have you learned along the way?

EW: Yeah. It’s a good question. I would say at the time, Trulia, Redfin, Zillow, Realtor, and there a handful of others, were exceptional at bringing MLS data online and building user-friendly interfaces. But my view of the world was slightly different in that…I don’t know if you play any video games at all.

I pretend like I play and buy games and then never actually play them.

EW: I had this view that it should be a first-person video game where you can do everything within the same world. You can explore it, you can jump, you can pick up things, and you have to move systems, if that makes sense. And at the time, again, these companies were exceptional at bringing data online and building experiences around it, but in my head, the experience was like if Uber and Lyft showed you a car and a map and you had to call three brokers to book the ride, and then it’d be unclear about timing, and you had to use them for future rides and you had to pay them with cash. Right? And so, that’s what the user experience mostly resembles today.

And so, the end state in my mind for the user experience was clear, which was it would look way more like hailing a ride or buying something online or booking a flight. We worked really hard at Trulia to improve the experience of search and discovery, and that was very valuable, but it never really completed the journey for me. Right? You couldn’t actually tour a home, you couldn’t buy a home, you couldn’t transact and close and move in. And that was part of the impetus of leaving and starting Opendoor, was that I had this view of the world that it should be possible to buy and sell a home in a few taps, in a few clicks, with just your mobile device.

Do you think that your perspective, though, was shaped by your experience? On one hand, you had far more visibility into all the pain points around this because, even in college, like you just said, you had bought 20 homes. Right? And I don’t know if you kept doing it or not, but that’s a lot of houses. However, is that very representative, because most people aren’t buying 20 houses, they’re buying one house. What do you view as the trade off between having iterated on it a lot and got a lot of experience, versus yes, there is this pain point, but for you the pain felt super high because you were encountering it so often, versus the way that normal people might measure that pain?

EW: I’m not sure if it’s frequency that is the determinant of whether there should be pain or not. I do think that the frequency of the transaction, whether it’s a real estate transaction or ordering food or buying something off of Amazon, increases or decreases the innovation velocity. But if you talk to anyone who’s been through a real estate transaction, I would venture to guess that 95% of them are dissatisfied with the experience. And so, they don’t have the incentive to go fix it themselves because they’re not dealing with it day by day, and that’s why someone else has to come in and actually fix the system.

But it’s a fair point that I had experienced it so many times over and over again that I felt compelled to fix it. And then, after joining Trulia, I had spent a huge portion of time talking to consumers every week, and also felt compelled to fix the transactional layer. But I think the point is fair, which is each individual, because they only do it once every five years, once every seven years, there’s not a lot of incentive to actually search for alternatives.

How did your view of those challenges shift, then, once you were actually in the space? Or has your estimation remained pretty consistent from even back in your Trulia days or even before then when you were doing the buying?

EW: Yeah. Maybe reframing the question. What has been counterintuitive about the space now that I’ve spent, I’m going to call it, 20 years in the category, I’ve been surprised by the dependencies on brokers, and part of that is actually structural in that if you think about how the system is built, it’s a two-sided marketplace. It’s more like a two-sided classified ad system, but a two sided marketplace in which in order to complete the next step you have to be licensed. And so, there’s regulatory blockers to allowing consumers to complete steps. A good example is you can’t use certain types of lock boxes unless you’re an agent, which means you can’t access the home. Or you can’t submit an offer with the contract that everyone uses unless you’re licensed. And so, these regulatory blockers actually create lock in for the system itself, and as a result of no governing body on top of MLS, there’s no notion of a managed marketplace.

These are things that as you dig into, it’s very obvious that, if you phrase it this way, would you rather have all the tools in your pocket and be able to take the next step with your mobile device, the consumer would say yes. But actually, there are decent reasons why you can’t do it. That was a bit counterintuitive to me. And the other thing I’d say about brokers is that they actually do serve a purpose. People are already voting with their feet saying, “Okay, I can either use a discount broker and save money or use a broker I know and pay retail price.” And what’s counterintuitive with that is that by using someone you know you’re basically paying for trust and mistake avoidance, and it’s a pretty high price. But the fact that you didn’t actually choose a discount broker to complete the transaction means placing some value on that trust even if it costs you an extra $20-30,000.

The Opendoor Story

So, talk about Opendoor specifically and the origin story. You mentioned Keith Rabois before. Obviously he’s been very vocal about the space. But let’s get the Eric Wu version. You talked about him recruiting you. I’m curious on both the personal level how you ended up doing Opendoor specifically, but then also what perspective were you bringing to the market and the specific problem you wanted to solve both in the short-term and what your long-term vision was.

EW: In 2013 after I had left Trulia, I had the very basic view that it should be possible to buy and sell a home with just your mobile device, as I mentioned. I had reconnected with Keith in 2013 with a view that you can actually start with the demand side and make the purchase of the home really seamless by being the bridge-financing piece. Or if you found a home you loved, you’d click buy, some entity would come and actually purchase the home on the spot for you, and then you can complete the transaction 30, 45 days later. There’s a bunch of companies doing that now, oddly enough, nine years later.

But we had ideated it a bit and we made a relatively important strategic decision, which was to start with sellers, and that was maybe a bit contrarian, because most people think of marketplaces, you have to start with demand. But we said to ourselves, one, there are a bunch of problems on the seller side and they’re more open to paying a liquidity premium for convenience.

If they know they can sell it right away, that’s worth a lot more to them. Because they only have one house. They have one house and they need to sell it, whereas a buyer, there’s a whole bunch of houses they could buy.

EW: Right. And it’s more of a transaction for a seller, whereas for a buyer, it’s very much an experience, and that wasn’t the driving force. I would say that the strategic decision was that if you can start to win supply and attract sellers or build a better engine to attract sellers, you can over time leverage that inventory and seller funnel to actually connect the dots with demand. So we said, “Okay, what would it look like if we were building an experience that made it incredibly simple to sell a house?” And really, the thing that we said was that the product positioning must break through the clutter and must be able to be communicated in 140 characters of less. This is one of the things that actually we still press on in the company, which is we have to have a position that actually breaks through the clutter.

So we said, “Okay, let’s make it possible to sell your home in 30 seconds or less.” And what stuck out about that was we phrased it in two different ways, which is sell your home in 30 seconds or less or sell your home in a few clicks. But that felt interesting. It felt differentiated. And it made clear to the consumer through the dimensions of speed, less hassle, less complexity, and less process, that this was a different and better solution than listing on the market and dealing with all of the hassle. And so, this is in late 2013. I told Keith that I’d be excited to start the company and he led the first round.

What was the key thing you had to do to make it possible to buy it in 30 seconds? Is this where the investment in the pricing algorithm is really the key factor?

EW: There are certain types of businesses that require two to three miracles or maybe they’re effectively running two to three different startups, and so there wasn’t just one thing. Now, the pricing engine was one of the most critical pieces, which is how do we build a pricing system that enables us to be able to price a home and purchase it or at least give the customer a quote in minutes. We didn’t want the delay. And so, we had to invest very heavily in that pricing system. One of my co-founders is a data scientist, one of the first data scientists from Square. And so, we’ve always prioritized pricing as a core competency.

The other pieces that were critical to build was, one, we needed the capital. At the time, this was in 2013 and early 2014, there weren’t companies raising hundreds of millions of dollars at billion dollar valuations without a ton of proof points and traction and scale. And so, we had to raise a meaningful amount of debt and equity capital to get the business off the ground, which again, was more unique at the time.

The third piece is that we had to build operations. We took a very vertically integrated approach to the problem and we prioritized the consumer experience above all else which is, again, how do we make this as simple and delightful as possible? We vertically integrated operations from sales to support which served as the customer’s agent in many ways to renovations and repairs and home maintenance.

And then I would say the last piece that became actually easier over time was just the consumer experience itself. The positioning, the branding. But I would say the critical piece in the beginning was how do we actually build the pricing piece and then actually do the transaction with capital operations?

So, I wrote about Opendoor when it was first announced, and my main takeaway was like I love the fact it existed just because of the sheer audacity of it, because of all the points that you listed. It’s not just one thing to build a front end for discovery, which is a very obvious Internet play to be at the head of the funnel and then monetizing basically via advertising, which is effectively the Zillow model. But to your point, you are going to have to take on a huge amount of debt facilities. You’re going to have to carry all this inventory to make it work. You’re going to have to depend on this algorithm of being right. I had no idea if it would work. I’ll be honest, I’m still not completely sure if it’s going to work. But I loved the audacity around it.

Competing With Zillow

What was kind of surprising, and I definitely did not see coming, is that Rich Barton agreed with me. He headed back to Zillow to do the same thing. Were you surprised at how quickly Zillow came in and wanted to compete with you in this market? Did you assume, in part because it was so audacious, you would have the space to yourself for a little bit longer?

EW: Well, I guess we can thank you, Ben, for that.

According to Rich! Who knows if it’s true.

EW: After getting to know Rich a bit more, I don’t think it’s surprising. He’s got a lot of boldness to him. He’s audacious. He’s ambitious. And so, while I think that was surprising at the time, in hindsight, it may not be as surprising.

That said, I’d always assume that because of the capital intensity of the business and again having to build multiple miracles, especially because you’re taking a more vertically integrated approach to problem…

Right. And much, much, much lower margins than what Zillow was enjoying.

EW: Much lower margins. So I thought that our runway, our lead time would be years if not a decade in the same way I think that Amazon invested pretty heavily in logistics and fulfillment and had a multi-year, multi-decade head start over competitors. And so that was always the base case. What I think we failed to estimate was that there was just a market bull run…

There was so much capital around that. Actually anyone could get the capital to try the business.

EW: Exactly. So we had competition coming from not only the industry leaders like Zillow and Redfin, but also a whole host of clones that were startups and variations of the idea and the business model. Back to your original point around Zillow, they came in pretty quickly after hearing the idea or at least maybe even reading your blog post. I think that required again some boldness, and Rich certainly took the swing.

Why do you think Zillow failed? I mean, Redfin just exited I think just a couple weeks ago. What did they get wrong about this space that you think that Opendoor has gotten right?

EW: Well, it’s hard to speculate how those organizations are organized and run in the focus areas. I think you can always say that it’s hard to do multiple things exceptionally well and maintain a lead. But I would say that it is a hard business, and that becomes a mode over time in the same way that you could say that DoorDash or Amazon, Apple, and Tesla are more challenging businesses than pure software businesses because they have logistics. And to some extent they’re more vertically integrated. And so again, I can’t speculate as to why their priorities shifted or why they were unsuccessful in this endeavor. I would say that for us it’s very important to continue to chip away at each inefficiency in that kind of value chain where there’s a bunch of transactional complexity at every step, and by being a principal in the transaction, we have to build the platform from scratch. And we have a lot of incentive to actually eek out gains at every single step for us and then ultimately for consumers.

Yeah. I mean, I think there’s a bit like you have no choice but to figure it out I think is pretty valuable in this space. Whereas Zillow had a good business, and there was some aspect there where that was probably why it never made sense to take on the risk in the first place. But I think just to go back to Keith, he’s talked a lot about Zillow not being very good at evaluating the price of a house, and that Opendoor was much better. And he’s talked about that being a real advantage. I mean, you don’t have to speculate about Zillow specifically, but number one, do you think that’s right as far as that being advantage? And I guess number two, that gets into some of your recent results. You had a whole bunch of houses, particularly in quarter two, that were basically bought at a price that was too high. I mean, did your model fail you? Is this just a further down the road manifestation of the mistakes that Zillow made? What went wrong there?

EW: Let me make one more comment about doing two things. I think you might have even written about this, that it’s hard to think of many examples where companies transition from being a high margin media company to a vertically integrated low margin platform. I think Yelp and DoorDash are examples, Google and Amazon. I’m sure you can think of more examples. So that’s part of the reason why starting with the bottom of the funnel and being very focused on the transaction layer and eeking out gains there gives us the opportunity to move upstream in a way that’s difficult to clone.

The Real Estate Slowdown

To the point around the importance of price, I would say that we’ve gotten very, very good at understanding the current value of a home and the drivers of that value from features such as the slope of the yard to views to the type of roof to the interior condition. There’s a lot of really, really great data collection and great modeling on top of that to able to price the asset in real time.

Given the shift in the market, we certainly made a forecasting prediction that has not been accurate about the slope of the change in the past three months. And to give you some context, the market moved from positive 7% appreciation per quarter to negative 1% in three months, and that is actually steeper than what happened during the [Great Financial Crisis] (GFC). And so we’ve seen a once-in-a-40-year move in home prices on top of a move in velocity that we’ve actually never seen in housing. And so there’s a difference between the transition period, which is okay, can you have predicted the steepness of that slope? And we did have models that said that’s possible, but we didn’t say it was likely. So there’s a transition period where you’ve purchased assets, the market move, there’s an outlier event, and you have to take losses on those assets even though at that time you may have priced them accurately.

This is the biggest questions I have. I was kind of surprised on your recent earnings call that the shift in rates — it was suggested that that was a surprise to you. And I guess I have questions in a couple respects. Number one, I was surprised to hear that because I didn’t feel surprised by the market rates, given the inflation that was going on. It seemed likely that the Fed would respond. If the Fed responded, it seemed likely that mortgage rates would move. So number one, why would that be a surprise?

But then number two, this gets into the questions around an algorithm and a model. No matter how good it is, definitionally an algorithm is using data that already exists. So it’s fundamentally backwards looking. And is this a bit here where, yes, that can be very good in doing relative pricing like what this house should be relative to the other one, but it actually is relatively worthless in doing macro prediction, and maybe that was under-appreciated before last quarter?

EW: I think what I’d say is that we were pricing in the interest rate movements but not to the effect that they shifted the home price appreciation from positive 7% to negative 1% at that speed.

So basically the second order derivative of that increase was not what you expected.

EW: Exactly. And so we had already started to widen spreads in anticipation for rates to rise. Now, they rose slightly faster than we anticipated, and then the home prices moved dramatically because demand just paused. There was uncertainty about where the rates would end up and buyers just sat out. That was a shift that we hadn’t seen before in any of the back testing and any of the data that we analyzed from previous home price depreciation, even the GFC. And so I would say that from a macro standpoint, by definition outlier events are nearly impossible to predict. And what has to happen for us, and I talked a bit about this on the earnings call, is that we have to build to widen spreads or decrease spreads based on our view of volatility and risk. But the platform itself has to be able to service customers in any situation.

And so you have to go from having just a first party business that can demonstrate positive yield economics after an event like what just happened. We are going to demonstrate that because we’ve widened spreads. There’s still conversion. There’ll be positive contribution margin. There’ll be a transition period where we mispriced the assets, and we’ll actually sell those with discipline and being very thoughtful about that.

I think the second piece that has to exist is that the platform has to build to service all customers. You have to be able to come to Opendoor saying, “I love what you’re offering,” which is a better way to sell my home. And instead of us buying all the homes, we are one of many buyers, and the marketplace itself is actually a better way to sell vis-a-vis MLS and brokers.

I want to get to the marketplace part. It is obviously very interesting. Marketplaces are always catnip for me, but the worry I have is that Opendoor is sort of at the end of a bull whip here where when stuff’s on the way up, maybe some buyers are certainly motivated by convenience, and they want to have a quick transaction. They’re willing to pay a premium for that. But there is a bit where to the extent that Opendoor needs inventory, it’s going to perhaps have to offer a bit more than you would like, and your margin is going to be a bit less so that you can acquire assets. Then on the flip side, when it’s sort of going in the opposite direction, when people really need someone to buy their house because, look, all the other buyers out there, they have super high mortgage rates. They can’t afford it. Can you please buy my house? Then Opendoor’s like, “Well, we can’t really afford to buy your house right now because, so our premium is going to be so large, it’s going to end up not being worth it.”

How can you shift from being on the back end of the bull whip? Do you think you can move to being more of a market maker or a driver instead of being subject to these changes and downstream from them? You can be further up, and I think this is probably going to be critical for you whenever this market turns. That’s probably where Opendoor can make the most money because you’ll have bought a bunch of assets at low prices. You will be sensitive to when people come in the market looking for stuff. You’ll have a lot of inventory for them. But how do you actually predict that going forward? You just said macro events are hard to predict, but macro events seem to be controlling your business. And again, relative pricing is great, but how do you handle that fundamental issue?

EW: I would say that what we’ve demonstrated is that when the market is moving fast on the positive side, and it’s a seller’s market per se, and it’s actually easy to sell, we still see incredibly high conversion because we’re not just offering certainty of execution. We’re offering maximum convenience.

You’re selling convenience, right. Makes total sense.

EW: And so I think we still saw very, very strong conversion and growth when the market was very healthy. It was very easy to sell. It was a seller’s market. On the converse, which is what we’re experiencing now, it’s a buyer’s market. It’s incredibly hard to sell, and as our value proposition has increased, what I’ve been surprised by is that the conversion is still higher than we anticipated. We’re converting somewhere between 10 and 15% of true sellers that come to us with very high spreads, which again speaks to the fact that the value proposition increases as there’s uncertainty, and it’s more of a hassle to sell.

And so I get the point that, “Hey, there are these macro events that actually will impact our spreads,” of which we have to increase spreads when there’s more volatility risk, and then we can decrease spreads when there’s less. And then the ultimate impact on the consumer experience is that if you want certainty and convenience, you have to pay a bit more depending on the environment. I think what we’re aiming to build is a system where you can come to Opendoor, and we can deliver more convenience, less hassle, and more certainty than listing on the market in any situation whether you sell to us or you sell to one of our buyers.


EW: And that’s the end state. We’re certainly excited about our first party business because it has elements that enable us to control the consumer experience end-to-end and really delight consumers. What we can do now is we can take those elements and that platform as a way to launch the marketplace, which is to make the experience of selling to one of our buyers almost as delightful, almost as convenient, and almost as certain.

Cohort Forecasting

So I want to get to the marketplace because obviously that solves a lot of the problems for Opendoor. In particular, carrying inventory. With a marketplace you’re sort of surfing the market as opposed to being at the tail end of it. I did have one other question though. When you said you sold around a third of your Q2 cohort last quarter in part by lowering prices, you also lowered the measurement of your inventory, which was a big part of your losses, like 500 some million dollars. And you said you will sell up to 65% of it by the end of this quarter. You also talked about the houses you acquired in Q3 where you paid more for them, you had a larger spread, are selling pretty well and you’re maintaining margins on them.

I guess the question I worry about in this regard is I remember when there were a lot of these delivery businesses going to market. I think there was one that I wrote a big piece about, I think it was Blue Apron or something. My critique of them was that when they went to the market they were building LTV models, lifetime value models, based on their initial cohorts, and the problem is the quality of the cohorts degrade over time. So their customer acquisition costs would spiral out of control and meant that it just wasn’t a scalable business.

Is there an extent to which that applies to houses as well? How do you understand how well a cohort’s going to sell? Because isn’t it the case that the best houses are going to sell first and the further down you get, the longer they’re on the market, you’re just going to have to discount more and more? And is that a problem you’ve seen previously? Is it worse now or how do you just think about measuring and forecasting when it comes to inventory in hand, knowing that some stuff flips super quickly and some stuff just takes a while?

EW: Well, that goes into the forecast, and so we look at the business on an acquisition cohort basis. And some percent of the cohort will sell quickly for very, very healthy margins. Some will sell less quickly for lower margins, and we’re actually baking that into the spreads. And so if we want to target say a four to 6% contribution margin range, some percent of the cohort on an acquisition basis will sell above that and some will sell below that. And so I get the point on miscalculating LTV and CAC based on consumer personas, and maybe there are early adopters and so on and so forth. And as you get into the different persona groups, there’s higher churn. I think the same thing exists for probably on the supply side for Uber and Lyft, which is that drivers kind of get worse as you find the marginal driver.

Yep, for sure. Yeah.

EW: And so for us, we haven’t seen that be the case, outside of making sure that our pricing by boxes is actually very disciplined around what we can predict. That’s why we do have a buy box in the first place. We’re not going to go price every asset. Because these assets, actually, as they become more bespoke, the outcomes have higher variance. And so we want to make sure that we’re actually pricing a certain type of asset, that we can forecast the outcomes within a certain error band, if that makes sense.

Right. So basically, this is the inverse of the issue before, where precision in pricing doesn’t fix the macro problem, but you’re saying you feel pretty confident in your relative pricing ability, in sort of a micro view.

EW: Yes.

Acquiring Customers

So we’re getting to the marketplace bit, but I do want to understand, starting with your first party business, is how does your acquisition engine work? How do you actually get people to sell houses to Opendoor? And I guess question number two is, how has that changed over time, if it has?

EW: I would say this is where part of the thesis has proven to be true, which is, if you’re able to provide an online offer in minutes, homeowners and, importantly, home sellers, will start their journey with us. The way we framed it internally is that imagine you had a envelope full of cash for your home, and it took a minute to open it up. You would attract lots of customers saying, “Let me go see how much I can get.”

This was always the Zillow argument, which is they had this built-in seller acquisition market, because everyone went to Zillow to see how much their house was worth. And so their relative expenditure on marketing would presumably be less, obviously if it’s not pricing correctly, that’s a problem if you’re actually making purchases. But is that what you’ve seen happen, are people just going to Opendoor because they want to know how much it’s worth? Or do you have to spend a lot on paid acquisition, to make people even aware that Opendoor is a better alternative than the Zestimate?

EW: In the beginning we had to spend to raise our brand awareness. I would say over time, actually, again 90% of our sellers come to us without an agent. 90% of sellers are selling, or visiting us, without having listed their house. And some of our mature markets, we’ve registered 30% of all sellers before they actually have selected an agent. And so the behavior around starting with the Opendoor offer is starting to play out. But I would say there are three big channels for us. Obviously there’s paid.

And where is paid? Is that on social media? Is that on billboards? What’s your most productive channels?

EW: Direct mail has been surprisingly productive, because we’re able to target individual sellers with individual homes. And you can predict how long they’ve lived there, and whether they’re more likely to be moving. And you can actually include a price, because we know a lot of about the home already. And so that’s been very effective for us.

I would say the second big channel for us has been in partnerships. We’re able to partner with some of the online real estate companies. We partner with a lot of home builders, who have buyers who are sellers. And the complication with buying a new build — and actually the same for buying a home on the market as well — is that you have to sell your home to qualify for the next house. And so it actually is a blocker to accomplishing a life goal, which is like when you have a new family or a new job. And so we’re able to unblock that. And so home builders have a lot of incentive to actually get the seller, who is also a buyer, to actually get an Opendoor quote, because we can guarantee the close.

That partnership angle is super interesting, makes a ton of sense. Do you have to pay? Is there any money involved in that? Or is your value to the builder so significant that it’s sort of a free partnership, in that regard?

EW: It’s a free partnership. It’s unpaid. we love these situations where there’s a win-win-win. The builder wins because there’s no contingencies on the sale. They can continue the build, complete the build knowing that we’ll be there. The customer wins because they don’t have to go figure out how to sell their home before putting a deposit down on the new build, and moving twice. And then we win because we get supply for free, with no CAC. And so we love that channel. Lennar’s been a big supporter of the company since early days, and that that’s grown over time.

The last channel is word of mouth. People just coming to the site, hearing about it from a friend, our MPS is north of 75. And that’s driven a lot of growth as well.

So this is really interesting to me. One of the questions I’ve always had about Opendoor, sort of looking at your financials, is when you look at cost of revenue, really all of it is the direct costs that are associated with buying a home, renovating it, and some of your inventory costs. Whereas sales and marketing is all below the line in operational costs. And the reason why I’ve always wondered about this goes to the question I asked you earlier, about being a frequent home buyer, but most people don’t buy homes that often. And so my question has always been, to what extent should customer acquisition costs actually be a cost of goods sold? Or at least thought of in that way, because you have to acquire that seller, and then that seller sells that home.

This is different than a market where there are frequent transactions, and the idea is that you acquire a customer — go back to Blue Apron for an example, even though it was a bad business. But the idea is, once you get a customer, then they order again and again and again. So you only need to do that expenditure once, and in return you get a multitude of transactions. What’s your view on that? How much leverage do you feel like you do get over time, on those paid channels, for example? Or is there an aspect where you do have to earn new supply on an ongoing basis, which means there is a marginal, variable aspect to your marketing costs, as opposed to being fixed and leverageable over time?

EW: We certainly look at our economics internally, fully loaded. Which includes sales and support and marketing. And what I say is that, your assumption is right, which is, unlike Blue Apron, which if you acquire customer-

Again, a bad business, we should use a different business, say DoorDash or something like that.

EW: Yeah, DoorDash.

Yeah, continue.

EW: We have to acquire supply. So someone has to come to us, they have to get a quote, they have to hear about us somewhere. And so the reason why it’s so important to have something that cuts trough the clutter, from a positioning standpoint, is we need that cost to be very low, or free. And so these partnerships matter a lot. We’re excited about these partnerships with, again, Zillow, and Redfin, and like. And then organic matters a lot.

We need brand awareness to lift all of the channels. But insofar as we can make the value proposition so clear to consumers, that when they hear about it, they’re like, why wouldn’t I do that? Then our cost to acquire a very, very valuable part of the ecosystem, which a high intense seller, can remain low. And again, going back to the thing we’re going to get to is that, we have to find ways, and we’ll find ways to convert that seller, right? Because, again, that is, in our view, one of the most important parts of the marketplace, which is supply.

Opendoor Exclusives

This is exactly where I’m going. You’ve kind of intuited of the direction of my questions, which is, you want to move to this marketplace model, and that’s been the vision for a long time. There was that slide in your seed round about this being the long term vision: starting with first party was a stepping stone to being a marketplace. And, again, as I sort mentioned earlier, this solves so many problems for you. If you can be very good at the micro level, the key thing is not being at the tail end of that bull whip, being much more towards the front end. If you’re just facilitating transactions, it doesn’t really matter to you what level those transactions are happening at. If the third quarter’s way down from second quarter, so be it, you’re still taking your 5%. But you have to have the demand.

All your value proposition for the supply has made a lot of sense. But to date, you’ve relied on the demand that’s there; you put your stuff in MLS, you have brokers that actually sell your houses. How do you actually get people, at scale, to actually make purchases on the supply side? And this is where my questions about acquisition costs certainly come to the forefront, right? Because you’re trying to acquire people who are less motivated, perhaps, or who have a lot of options. And you are going up against a Zillow and a Redfin, with a different business model, to be sure. But how are you thinking about building that out, which is going to be essential for a marketplace to work?

EW: Let me answer each one of these components, because it is actually an interesting discussion to have. You have to have a differentiated experience to acquire buyers as well. So what we talk about a lot is how do we cut through the clutter on the supply side? I think we’ve done an exceptional job with that. And then the question is, okay, if we’re building a two-sided marketplace, we need supply and demand. How do we acquire all of the demand? And there’s two things I would say is that, one, the buyer, in this case, which is a benefit to us, prefers complete selection. It’s a considered purchase, and on average, buyers visit 10, 12 homes before they make an offer.

Which makes the dynamics very different from the seller, who has one house that they have to get rid of.

EW: Exactly. And so, just by having inventory, that means that the customer on the demand side is going to likely want to see that inventory.

And so you’re going to be listing all the MLS listings, plus your listings, and so you’ll have a super set of all the available houses.

EW: Well I think that goes back to this notion of original content, or exclusive content, where, if we have homes that you can’t find anywhere else, that’s certainly a very strong value proposition that cuts through the clutter, that would then encourage you to say, “Okay, I should probably search on this site as well.” And there’s not necessarily a moment where you have to choose Redfin or us. We’re actually very happy with people multi-homing, using both sites, as long as we’re part of the consideration set. And if we have a home in our marketplace that fits the customer needs, then we would love to complete that transaction in a very seamless way. And so I’d say, point number one is that buyers want complete selection. And what we’ve seen is that, just by having signage in front of our homes, and people seeing a home for sale on MLS, and walking in person, and then seeing a for sale sign down the street, they’ll actually download the app and visit that house.

So if you visit Opendoor, do you see MLS listings? Or are you seeing only Opendoor listings?

EW: On our exclusive marketplace, you only see Opendoor listings.

Got it.

EW: And there’s an argument to be made that, over time, you can weave in MLS, which we have access to, and have both sets of listings, but we want the app to feel different. We want it to feel like these are homes you can’t find anywhere else. We want that engagement to be that way. We want the notifications to be that way. And over time, we can backfill with inventory. I think Airbnb does some of this with hotels. If they have sparse listings, they’ll kind of backfill with hotels. And so there’s certainly that opportunity, but the experience itself should feel like there’s a unique set of selection here, and a differentiated way to purchase that selection.

And so I would say, okay, we have the benefit of a complete search for most buyers. We can use signage to acquire lots of customers because they’re actually driving the street, to a home down the street they may buy, and they want to visit the houses in the neighborhood before making that purchase.

The second benefit is that we actually can use the current system to cross sell our homes. So we have a bunch of homes in inventory, a bunch of homes that sellers have. We can actually post to MLS, consumers will walk into those homes, and we can actually notify them of other homes down the street that aren’t on that system. And so it enables us actually acquire customers from the existing infrastructure, if that makes sense.

And the way you do it is, houses are on Opendoor exclusives for a period of time before you then put them on MLS, when they become broadly available. Is that right?

EW: Yeah, we talk about as windowing the content. So for 14 days, it is exclusive to us, and then we’ll syndicate. And through syndication you actually acquire more customers, which then you can actually cross sell back into the exclusive content.

Let me make one very important point. When we launched exclusives, we actually thought about this as the crux, which is, how do we acquire all this demand? We’re at 20% of our inventory, Opendoor’s inventory, in our three markets that we’re live in. We sell 20% of our inventory in that 14 day window. And we map that to MLS. MLS sells roughly 30% of its inventory in 14 days. So we’re not too far off, actually, with almost more than one 10th of the distribution. So MLS has full distribution, and is selling 30% within 14 days. We don’t have full distribution, we’re selling 20% within 14 days. And so there’s actually good evidence that we are able to match MLS’s velocity with less distribution.

On one hand can, the reason you have exclusive content is because you are acquiring the houses. So, by definition, you control it, you decide how it’s there. But how do you get exclusive content that is third party? What’s the virtuous cycle there? If you want more and more content to be just on Opendoor, but you are not going to necessarily buy it, how do you get to that next step?

EW: Well I think it starts with why it’s so important to have the first party business, and the driver of the supply. So we have all these high intense sellers, coming to us every day, raising their hand saying, “Opendoor, help me sell my house,” effectively. And so what we found is that it’s actually very high converting to say to a customer, “We can make you an offer, and we can simplify the transaction for you, ahead of listing. And on top of that we can go see if we have a buyer, that we work with now, that can make you a higher offer.” And so the seller can opt into that system. And we found that that is very high converting.

So basically you come in and say, “We’ll offer you $250,000, there’s a chance you might be able to get $260 or $270,000. So put it on the market.” I mean, if they put it on the market, trying to get that buyer, do they have that $250,000 offer in pocket? Because, presumably, the market could change. So how long do you want that to be out there as an option for them?

EW: So the window is 14 days. And the way it’s framed to a seller today is that we’re going to make you an offer, and we’re going to look for other offers that are going to be higher than ours. At the end of this window, we’re going to actually present you all the offers. And again, in one of our first customer cases, we had an offer from us, we had an offer from an investor, an offer from a retail buyer. We presented all three offers to the customer, and they selected the one that fit their needs.

How important are the institutional buyers to making this work? Are they sort of the guaranteed second buyer, or market makers, in some respects, where it’s not just Opendoor making an offer, but there’s someone else that’s interested as well?

EW: Certainly it’s easier for them to conform to our customer experience standards, because they look similar to us, more so than is similar to a consumer. I would say that, again, going back to the fact that we’re selling 20% of our inventory to buyers on exclusives, that means that we can likely sell 20% of our inventory of sellers to our buyers, who are willing to accept higher offers. We haven’t found any real pushback on the pitch of getting sellers to open up the window to accept higher offers, from other buyers. They’re not marketing the home. They’re saying to us, “If there’s a higher bid, go find it for me.”

REITs have been historically 10 to 15% of our transactions, and in many cases they’re able to place a bid on a home that we have under contract but not yet closed on. And so that means that in a steady state world, 10 to 15% of our transactions will be a marketplace transaction to an investor because they’re able to bid on the home before we’ve even closed on it and close on our timeline. And so while we believe that the category will be robust for many years to come, the REITs have raised lots of capital. There’s a lot of logic why some of these homes should be rentals. It’s not dependent on that because again, we have a growing consumer base that’s shopping with us, buying with us, and we have a growing seller base that’s willing to actually give us some time to see if there’s a higher bid.

Are you going to have to do paid marketing though to get more buyers? Or do you anticipate all of the growth being organic via the signs or other channels?

EW: I think we’re open to paid marketing. Again, the goal is to actually build a better marketplace that actually tracks both sides.

That’s why marketplaces are hard though, right? Once they’re going, it’s great. It’s really, really hard to get them going.

EW: Yeah, I think paid marketing can be a component to it, and we’ve seen really strong traction and growth with not a single dollar of paid marketing. Again, we’ve done no marketing on the demand side to date. And so, it’ll only be amplifier of the success of the marketplace. And the last thing I’d say is that our relationships with the REITs have been built over the past five, six years and we’re actually vertically integrating into their systems to make this bidding process easier. And so I think there’s going to be a richness of consumer demand. There’s going to be a richness of investor demand and that presents more options for the seller to either accept an Opendoor offer, consumer offer or an investor offer. And if they don’t like any of the offers, they can still elect to list a home on the market. And so we’re really just a pre-market, pre-MLS system for them in which they can still list on the market if they choose to.

The big question I have big picture is I can really see the value proposition for the first-party business when it comes to a seller. I think it’s pretty straightforward. It’s obvious. The convenience, the guaranteed transaction, all those things. That then goes down the road and means you have exclusive content. Okay, I get that. That makes sense. I’m still a little unsure as to why someone would want to sell on your marketplacr. I guess theoretically they could get a higher price. I guess the core differentiation makes sense, but the further down it goes down the chain, I do wonder about what weight it has. But in an optimistic perspective, how does that work out? How does this path all pay off? I mean, let’s say 100X in value. How does Opendoor get there, and I guess first off is surviving this market now, but how does this pay off in the long run?

EW: Yeah, there are two questions there, which is why would a seller use exclusives to get additional offers? And then the second is how does this all play out for Opendoor to 100X? I would say the thing to keep in mind is that the alternative is actually a pretty poor experience. The seller is saying in their mind, “I could either say yes to Opendoor looking for more offers for me on my behalf at no cost for 14 days. Or, I can go find a realtor and do repairs and list on the market for 180 days.” There’s a sizable cost to actually going with the alternative.

Right, but there’s a selection problem for you, right? If something will sell in 14 days, you probably could have just bought it yourself and then made money off it.

EW: Well, what we’ve seen is that actually the pie is much bigger. The amount of people saying, “Okay, I’ll give you the 14 days to go put together a collection of offers” is actually much larger than us just presenting our offer itself and them saying yes or no.

Because they’re worried that you’re undercutting them, so they want an alternative.

EW: Right. They may say, “Okay, well maybe Opendoor’s offer will be high or low, so why don’t I get more options on the table in this time window.”

But the other thing I’d say is one, the transaction’s too complex. We’re making that very simple. We’re going to standardize the transaction around a simple process. It’s also too lengthy. So, sellers don’t know how many steps they need to take, how many open houses. We’re going to going to confine that to a certain timeframe. There’s way too much uncertainty. And so with our system, all of our offers are guaranteed to close. So, that gives them certainty that there isn’t renegotiation around repairs at the point of sale.

And then, the last piece is that there’s a lack of fairness in the system. We’ve launched what we call an appraisal price match guarantee, which is for a buyer and a seller, if the appraisal comes higher or low, we actually match that. So, we’re able to do things as a marketplace that each individual seller with a broker can not do. This is the reason why sellers are saying yes to us and buyers are saying yes to us.

Upside and Downside

So what’s the 100X payoff then? Opendoor has been flirting with a $1.00 per share. How does it become $100 per share?

EW: Well, I’m waiting for you to write the next blog post and then we’ll respond!

Like I said, the vision is so audacious. The question about real estate in general is everyone looks at real estate like, “Look, this market makes no sense.” You’re actually paying the highest commission you pay for anything for the largest ticket price item you buy, and it’s just a crazy amount of money for the value you’re getting, and it’s a big pain in the neck. You have to clean up your house. You have to do showings. And, if you want to buy a house, you don’t know if you overpay. It’s a total mess of a market.

And so, the narrative of “Look, we’re going to fix this” is very, very attractive, but I think broadly speaking, markets that are super weird and messed up are often emergent. They emerge from certain conditions that made sense in the time and place. In one respect, that means they might be obsolete. On the other hand, there may be factors like the human body. It’s so complex and unknowable. Then you add on this weird bit about people don’t transact that often so there’s less pressure to change things, and there’s a high trust component that you talked about before. In many respects, because real estate looks so messed up, it almost makes me feel like there are that many more factors inside the real estate market that we don’t even realize yet. You don’t even realize it because you haven’t been there for long enough. And maybe it’s that way for a reason, but then maybe it’s just stupid, and Opendoor is going to fix the whole thing and make it super transactional and it’s going to be 100X.

Why should there be faith that you’re going to figure out all the problems, all the reasons it is why it is, or actually it is in fact stupid, and Opendoor is going to be better and you’re going to be the ones to figure it out.

EW: Well, I’ll answer the question how do we get to 100X, which is if you’re thinking about fundamentally where the consumer experience is broken, it’s the fact that it’s MLS, which is an essential database and brokers on both sides. It’s buyer beware, and so there’s massive inefficiency and it’s zero sum. Whereas a marketplace that is managed can actually build features that are better for both sides in the system itself.
Now, if that attracts the customers at scale, you can imagine a world where this goes from in some markets, we’re 10% market share, 5-10% market share, it goes to 20 to 30 to 40 to 50 to 100% over time.

Right, and then your customer acquisition costs basically are nonexistent because everyone knows you just go to Opendoor to do this.

EW: Right, and even in the short term, it’s actually okay for us for customers to say, “Okay, I’m going to use Opendoor and MLS to shop,” which is a win for us. And it’s okay for sellers to say, “I’m going to start my journey with Opendoor. I’m going to start to see if they give me a quote and an offer from one of their buyers or themselves that fits my needs so I can skip the line.” If that behavior starts to really take hold on both sides, then it’s very obvious that we can go from where we are today, which is again 10% in some markets, to 20, 30, 40, 50%.

Where does it go wrong? What’s the opposite case? How does it end up not working out?

EW: I mean we just talked about some of the complexities-

Yeah. It’s more of the same, right? Big macro shifts are hard to predict and pricing models are not particularly useful for it. That’s basically the story.

EW: Well, ideally, we’re post that transition and moving forward, there’s not going to be another once in a 40 year shift. I would say that we talked a bit about focus, and I would say that the best companies have a second act. They’re able to take their first product and turn it into a platform. This is probably the case for Amazon and lots of other companies. So, that’s where the real challenge is. It’s how do we take this first act, which is this first-party business that attracts hundreds of thousands of sellers every month that are high intent saying, “Help me solve my problem,” and then we have inventory that tracks hundreds of thousands of buyers that are coming to us saying, “I’d love to buy a house.” And how do we build a marketplace on top of that behavior? That’s not an easy lift, and that’s a second act for us and that’s what we’re betting on.

At the end of the day, should you and Zillow just really just hang out together? Because they have the customer acquisition on both sides, but particularly on the buyer side. And that feels like that would solve a lot of your problems.

EW: Well, it’s a good prompt. As a former competitor and now a partner of Zillow, we have a lot of admiration for them.

I mean, you announced a partnership on the seller side, I think that’s going to launch next year where they will show an Opendoor estimate as part of the process.

EW: Yes, we’re excited about the integration on the seller side, and again they have all this rich audience that is looking to Zillow for pricing, of which an Opendoor offer could be part of that feature set. And so we’re excited about the partnership and experience there, and I think there’s a lot of opportunity on the demand side. Now imagine a world where we’re getting lots of real sellers coming into our collective funnel saying, “Help me sell.” Some end up listing on MLS and using agents, some sell to us, and then we have a whole rich experience for buyers both on Opendoor and Zillow to shop and buy direct. We think that could be a world that actually is a win-win for both companies.

Well, after all these years, are you still as passionate about buying 20 homes and flipping them? I guess you’re doing the same thing you were doing all along.

EW: Well, I guess I’m doing that times a thousand. I think that as you spend more time in a category, there’s actually two directions you can take. One is that you become less energized because the incumbents are so strong, whether it’s for regulatory reasons or the transactions are infrequent, but the thing that continues to be where I get a tremendous amount of energy and passion is that the consumer problem is fundamentally broken, and I wake up every day trying to fix that. And so, it’s become my life mission, and actually the passion’s increased with time because there’s still so many problems to solve.

Does the passion increase with doubters?

EW: I’ve always been an underdog in my life, and I think it’s easy in moments like this to say, “Okay, well maybe the opportunity is more narrow,” which is “Okay, why didn’t spreads make money on some folks that actually need liquidity?” But again, if we’re successful in building this platform and we have this conversation in five years, I think the tone will be very different.

I think almost by definition if we have the conversation in five years, that means the tone’s going to be different, which is gratifying. You’re all in and you’re going to make it or you’re not and it’s going to be very interesting to watch.

EW: The last thing I’ll say is that we often talk about how only with the fullness of time is our story told. We’re very, very focused and persistent on building this the right way for consumers and not taking the incremental approach of selling leads to realtors. So, in the end, we will have a far superior experience that’s vertically integrated as a result and then the business results will come from that.

One quick question on that. One thing is that today, because you are just buying houses and then realtors get to sell it, and I would imagine from a realtor perspective, dealing with your houses is easy. You have people that are excited. They’re looking for new things as opposed to working through all these problems. So, Opendoor is a great partner for realtors because you’re like, “Hey, we did the hard work of getting this house. Can you please go sell it for us and make your commission?” It’s a win-win. However, as you move into the demand side, is there a worry that you now have channel conflict where you’re trying to sell the house yourself, and maybe realtors are going to want to start freezing you out or driving people away from Opendoor properties?

EW: It’s a good assumption. I would say that one, we actually sell homes ourselves. We’re the owner of the home and we sell homes, and buyers come in with representation representing them.


EW: Looking for their best interests of whether this is the right home or not. So, the question is does steering exist if there’s channel conflict? And, what I would say is that we’re not out to displace to buyer and his or her agent because the buyer’s already selected to use that person, and for the right reasons. Whether they trust the person, they’ve known the person, the person’s great at giving advice and they’re helping them shop for homes. We just want to be in addition to that. If they find a home on our marketplace that they love, they can buy it from us directly and actually save money, or they can use their agent. Actually both work for us.

Well, I look forward to continuing to watch and see how you go. I think this is the Opendoor thing. I wrote about Narrative this week, right? Opendoor has just an incredible narrative. And the question is does that narrative become reality and how does it work out? And well, you’ve made it this far. I’d love to come back in five years. I hope we do.

EW: Okay, we’ll get that scheduled in five years.

Sounds good. We’ll put it on the calendar.

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!