People

How Opendoor is innovating its way back to selling homes safely

Ever tried valuing a home when you can't send round an inspector?

How Opendoor is innovating its way back to selling homes safely

The same, but different.

Photo: Avi Waxman

Opendoor's business model works when people move houses. But while we've all been sheltering in place, buying a new home hasn't been high on many to-do lists.

The 6-year-old SoftBank-backed real estate technology company — which hopes to sideline the traditional marketplace by offering convenience, speed and instant offers to sellers based on data and algorithms — has had to reckon with a huge slowdown of the housing market brought about by the pandemic. In April, the Commerce Department reported that new home sales in March were down 15.4% compared to the same time frame a year before, and Opendoor laid off 35% of its workforce.

But in recent weeks, Opendoor has seen users return to the platform as parts of the country have reopened. Data showing June's residential sales — and potentially that trend — is set to be released on Friday. For co-founder and CTO Ian Wong, moving forward has meant refocusing the efforts on making the largely analog real estate process as seamless and contactless as possible for customers. "What we've seen is that customers not only crave convenience now," he told Protocol in a recent interview. "They also really want assurance that [the process] is safe."

In an interview with Protocol, Wong described how the company is valuing homes using less information than in the past, how you train an algorithm to weigh a nice garden against mixed appliances, and what legacy real estate practices still need to be disrupted.

This interview has been condensed and edited for clarity.

In late June, Opendoor said people were returning to the platform and to the real estate market. Since then, we've seen spikes of coronavirus cases across the country. How has that affected that trend?

What we've seen with COVID has really evolved. In March, we saw COVID drastically impact many industries, and real estate was one of them. And what we've seen over the last few months is that life still goes on. People still need to move, people are still having babies, getting married, sizing up, sizing down, relocating for that next opportunity. So we've seen people continue to move and actually participate in housing transactions, and despite all the things that are going on, we continue to see that.

So as there's a surge of cases, I think what might be happening is the counties are more prepared. And also people are more cognizant of the risks and [are] navigating around those risks as they are trying to also navigate their lives. So we've seen actually the market itself be surprisingly resilient in these times.

What's the message you're trying to get across to the folks who are still making the decision to move during the pandemic?

I think what COVID has done across many industries — real estate being one of them — is accelerate a lot of consumer trends. You and I are on Zoom call. People are probably more comfortable ordering from Instacart. And in many ways, the consumer trend is that customers want things that are on-demand, self-service and purely digital. And that hasn't been the case in real estate for a very long time until [now].

Actually, what we've done this last few months is really fast forward a lot of the product innovations that have been on our backlog to give customers more choice and also fully digitalize the experience so that it's completely safe and contactless for our customers to actually buy or sell homes.

One of the differentiating factors for Opendoor has been convenience. Has leaning into that been helpful as people have been forced to stay at home?

What we've seen is that customers not only crave convenience now; they also really want assurance that it is safe. Previously, if you wanted to sell your home to Opendoor, we actually had to send an inspector to your home to assess the condition. And we ultimately had to send vendors into your home to repair the home to get ready for listing. Obviously that's a no-go while trying to comply to the CDC guidelines when it comes to social distancing. Now, you can actually have a virtual assessment. The other part that was surprisingly analog was closing. We still had notaries where you have to physically sign a piece of paper in order to transact, and we are using e-notaries now to really make the closing process purely online.

Then on the buyer side, previously, a lot of folks were touring homes in person. Again, that's not really feasible given the safety guidelines, so we've introduced virtual tours where we can actually have an agent who will be able to comply with safety guidelines to help you. Alternatively, if you want to visit one of our homes, we've increased the frequency of cleaning. And with all of our investment and technology, we can ensure that you're the only customer in a home at any single point in time, so that when you visit an open house, it's going to be a safe and private experience for you and your party.

How has switching from data generated by in-person inspections and assessments to data generated virtually affected the pricing algorithm Opendoor uses?

We had to obviously change a little bit in terms of how we transform that data into a form that the algorithm would take in. But by and large, we were able to rely on a lot of things that we've built in-house.

Just to take a step back, our pricing algorithms underpin the entire business because the whole purpose of Opendoor is to become this housing marketplace where we can give sellers fair market offers and also make it accessible for buyers as well. So pricing is super important, and that's how we can make sure that we can give accurate valuations at scale that are consistent.

But when it comes to actually valuing a home, there are all sorts of things that are very qualitative: there's curb appeal, there's the moment you walk into the home, how does that home feel to you? There are things around external obsolescence: Is there a power line cutting across, or is there that beautiful tree that everyone wants to have in the backyard? So we've actually devoted a lot of energy and investment into how we turn these qualitative sensory signals into things that the algorithm could understand and price off of.

We've already done all of those types of investments. And, in fact, when we send an inspector, we actually have a custom-built application for our in-house inspectors to use in order to capture and translate those qualitative factors into the quantitative factors. So while we've had to do some adjustments at the tooling end, we are able to leverage a lot of infrastructure that we've already built.

In terms of turning that qualitative data into inputs the algorithm can work with, is there a particular one that's been unique or challenging?

I think there are a lot of obvious ones like the condition of a home or the type of countertop or the type of flooring, but where it gets really interesting is how these qualities intersect with one another. The quality of the layout is something that actually is quite difficult to quantify. We've had to invest a lot of energy into understanding that. Curb appeal and the appeal of the backyard is something that is really hard for a computer to say, "oh yeah this looks good or not." And it takes a lot of human input. This interplay between the algorithm and a human expert to actually tease apart the curb appeal or the backyard appeal of the home.

And on top of that, not only do you have these qualities that are hard to quantify, but they also interact in different ways. So what if you have a beautiful curb appeal but you have mixed appliances. How do you take these two factors into account? What's been really interesting is learning from that data set in terms of being able to feed that into our pricing algorithm and combining it in different ways to come up with an accurate prediction.

What other kinds of factors are feeding the algorithm?

We look at every single transaction that's been conducted in any of our markets since at least when transactions are recorded, which normally goes back 10 years. So we have at least the last 10 years of data of all the transactions that have happened along with every single parcel that exists in these markets because we have to understand how the neighborhoods look. One thing that I would add that's interesting about the complexity of the problem is that we have to do a bit of time travel.

This asset class is heterogeneous, meaning it doesn't look the same both in space because obviously your neighbor's home may not look the same as yours, but also in time. The same home might be priced differently a year ago and a year from now. We have to spend a lot of energy investing in understanding how our algorithms would have performed 10 years ago. And also [understanding] where prices would be a month from now, two months from now, three months from now. Because we want to really understand if our model is both accurate historically in time and will continue to be accurate as we forecast into the future.

To borrow terminology you've used, what's still clumsy in real estate that you're trying to innovate around?

One that I would highlight — because I can go on and on about this — is just all the back and forth and all the paperwork that takes place in a real estate transaction. It's like buying this 500-pack of printer paper, except it's filled with texts that no one can read. It's very opaque.

And what we want to do is make the housing transaction much more accessible. So we've done a lot of work around making these purchase agreements much more lightweight. We've done a lot of work around digitizing, helping consumers understand exactly what they're involved with and ultimately helping them sign this paperwork and having a record of it with a few clicks of a button. Centralizing that all into one seamless experience is one thing that we spent quite a bit of energy trying to solve and still have some ways to go.

As you're planning ahead, what are you thinking will be your next big focus before the year's end?

We've relaunched many of our markets. Before COVID, we were in 21 markets, and we're going to be relaunching the rest of the markets by the end of the year. That's one immediate thing that we're very focused on. And the other thing is we've introduced a whole new set of products. We have the "Sell Direct," we have "List with Opendoor," which is a more traditional way of listing your home. We have "Home Reserve," where we'll buy your next home for you, and you can actually move in while we list your vacant home.

Our vision is to be a one-stop shop for all things real estate. So what we're going to be focused on for the rest of the year and for next year is tying all these pieces together so that no matter if you're buying or selling and trading in your home, it can all be done with a few clicks.

Fintech

Judge Zia Faruqui is trying to teach you crypto, one ‘SNL’ reference at a time

His decisions on major cryptocurrency cases have quoted "The Big Lebowski," "SNL," and "Dr. Strangelove." That’s because he wants you — yes, you — to read them.

The ways Zia Faruqui (right) has weighed on cases that have come before him can give lawyers clues as to what legal frameworks will pass muster.

Photo: Carolyn Van Houten/The Washington Post via Getty Images

“Cryptocurrency and related software analytics tools are ‘The wave of the future, Dude. One hundred percent electronic.’”

That’s not a quote from "The Big Lebowski" — at least, not directly. It’s a quote from a Washington, D.C., district court memorandum opinion on the role cryptocurrency analytics tools can play in government investigations. The author is Magistrate Judge Zia Faruqui.

Keep Reading Show less
Veronica Irwin

Veronica Irwin (@vronirwin) is a San Francisco-based reporter at Protocol covering fintech. Previously she was at the San Francisco Examiner, covering tech from a hyper-local angle. Before that, her byline was featured in SF Weekly, The Nation, Techworker, Ms. Magazine and The Frisc.

The financial technology transformation is driving competition, creating consumer choice, and shaping the future of finance. Hear from seven fintech leaders who are reshaping the future of finance, and join the inaugural Financial Technology Association Fintech Summit to learn more.

Keep Reading Show less
FTA
The Financial Technology Association (FTA) represents industry leaders shaping the future of finance. We champion the power of technology-centered financial services and advocate for the modernization of financial regulation to support inclusion and responsible innovation.
Enterprise

AWS CEO: The cloud isn’t just about technology

As AWS preps for its annual re:Invent conference, Adam Selipsky talks product strategy, support for hybrid environments, and the value of the cloud in uncertain economic times.

Photo: Noah Berger/Getty Images for Amazon Web Services

AWS is gearing up for re:Invent, its annual cloud computing conference where announcements this year are expected to focus on its end-to-end data strategy and delivering new industry-specific services.

It will be the second re:Invent with CEO Adam Selipsky as leader of the industry’s largest cloud provider after his return last year to AWS from data visualization company Tableau Software.

Keep Reading Show less
Donna Goodison

Donna Goodison (@dgoodison) is Protocol's senior reporter focusing on enterprise infrastructure technology, from the 'Big 3' cloud computing providers to data centers. She previously covered the public cloud at CRN after 15 years as a business reporter for the Boston Herald. Based in Massachusetts, she also has worked as a Boston Globe freelancer, business reporter at the Boston Business Journal and real estate reporter at Banker & Tradesman after toiling at weekly newspapers.

Image: Protocol

We launched Protocol in February 2020 to cover the evolving power center of tech. It is with deep sadness that just under three years later, we are winding down the publication.

As of today, we will not publish any more stories. All of our newsletters, apart from our flagship, Source Code, will no longer be sent. Source Code will be published and sent for the next few weeks, but it will also close down in December.

Keep Reading Show less
Bennett Richardson

Bennett Richardson ( @bennettrich) is the president of Protocol. Prior to joining Protocol in 2019, Bennett was executive director of global strategic partnerships at POLITICO, where he led strategic growth efforts including POLITICO's European expansion in Brussels and POLITICO's creative agency POLITICO Focus during his six years with the company. Prior to POLITICO, Bennett was co-founder and CMO of Hinge, the mobile dating company recently acquired by Match Group. Bennett began his career in digital and social brand marketing working with major brands across tech, energy, and health care at leading marketing and communications agencies including Edelman and GMMB. Bennett is originally from Portland, Maine, and received his bachelor's degree from Colgate University.

Enterprise

Why large enterprises struggle to find suitable platforms for MLops

As companies expand their use of AI beyond running just a few machine learning models, and as larger enterprises go from deploying hundreds of models to thousands and even millions of models, ML practitioners say that they have yet to find what they need from prepackaged MLops systems.

As companies expand their use of AI beyond running just a few machine learning models, ML practitioners say that they have yet to find what they need from prepackaged MLops systems.

Photo: artpartner-images via Getty Images

On any given day, Lily AI runs hundreds of machine learning models using computer vision and natural language processing that are customized for its retail and ecommerce clients to make website product recommendations, forecast demand, and plan merchandising. But this spring when the company was in the market for a machine learning operations platform to manage its expanding model roster, it wasn’t easy to find a suitable off-the-shelf system that could handle such a large number of models in deployment while also meeting other criteria.

Some MLops platforms are not well-suited for maintaining even more than 10 machine learning models when it comes to keeping track of data, navigating their user interfaces, or reporting capabilities, Matthew Nokleby, machine learning manager for Lily AI’s product intelligence team, told Protocol earlier this year. “The duct tape starts to show,” he said.

Keep Reading Show less
Kate Kaye

Kate Kaye is an award-winning multimedia reporter digging deep and telling print, digital and audio stories. She covers AI and data for Protocol. Her reporting on AI and tech ethics issues has been published in OneZero, Fast Company, MIT Technology Review, CityLab, Ad Age and Digiday and heard on NPR. Kate is the creator of RedTailMedia.org and is the author of "Campaign '08: A Turning Point for Digital Media," a book about how the 2008 presidential campaigns used digital media and data.

Latest Stories
Bulletins