How ‘clever’ tech brought down a US property giant

For some American homeowners, one number is the most important indicator of wealth in their life: the Zestimate. 

An approximation of their house’s value from the property website Zillow, based on public records and a computer algorithm, it is a real-time assessment of people’s most valuable asset – and checked religiously by potential sellers. 

When they do not like what they see, things can turn ugly. “F— Zillow and their ‘zestimates’. This shouldn’t even be legal,” one online commenter wrote after their home’s estimated value dropped by $70,000 (£52,400). 

The Zestimate is so influential that some have even taken the Seattle-based company to court, claiming the calculations damage the sale value of their home.

But three years ago, New York-listed Zillow decided to stake its own future on its algorithmic magic. Previously an online search engine for property, founder Spencer Rascoff said the company would go one further, buying up and selling homes based on its own estimates of their value.

Zillow is a controversial home valuation site in the US

Credit: Tiffany Hagler-Geard/Bloomberg

“We think that we have the resources, the scale, the analytics, the consumer insights, the housing expertise, that will make us a really effective bidder in our marketplace,” Rascoff told investors.

The proposition was simple. Homeowners could log on and receive an algorithmically-generated estimate of how much Zillow was willing to pay – typically a slight discount to what they might expect from the gruelling sale experience. 

The deal would be completed in as little as a week, at which point the company would apply a lick of paint to the property and sell it on at a small profit. Zillow outlined bold plans to buy 5,000 homes a month, and quadrupled its borrowing capacity to pay for it.

But Rascoff’s faith in the company’s own genius was misplaced. Earlier this month, his co-founder Richard Barton announced an embarrassing U-turn that Zillow would shut down its home-buying operation after three years and lay off thousands of employees. 

Shares fell 25pc the next day and it now faces the prospect of offloading thousands of homes, losing hundreds of millions of dollars in the process.

“It’s tempting to draw a parallel to WeWork,” says Mike DelPrete, a property tech adviser, referring to the office space company that spectacularly imploded when its venture capital-fuelled hype was not matched by profits. 

“The question you’ve got to ask: is this a technology company, or a real estate company? You’re sitting on thousands of houses and, at the end of the day, this is real estate – there’s cracks in the foundation and painting that needs to be done. 

“Technology can only take you so far and as the Zillow example shows, if technology fails, it can fail spectacularly. This is the kind of business that would keep me up at night if I was the CEO or an investor.”

Outlining the decision to analysts, Barton admitted Zillow’s prediction engine had not proved as intelligent as hoped. “We have been unable to predict future pricing of homes to a level of accuracy that makes this a safe business to be in,” he said, adding there was a concern that “the model cranks out a business that has a high likelihood, at some point, of putting the whole company at risk”.

WeWork shares initially rocketed on listing, but the party didn't last

Credit: Michael Nagle/Bloomberg

Zillow’s rise and fall, based on a belief that Silicon Valley data expertise could crack the housing market, was the most visible failure in the world of “iBuying”, a growing trend of capital-rich start-ups using technology to shake up how houses are bought and sold.

While proponents of the idea say it takes the stress out of selling – turning a confusing and laborious process into a seamless one – sceptics say it represents a highly-leveraged and hubristic belief in the power of technology.

One property tech entrepreneur says the housing market lacks enough data for pricing algorithms to beat local knowledge. “I just don’t think that data science can price a property with that accuracy,” they say. 

“If you’ve got one thing wrong, the condition of the kitchen, or if the garden is overlooked, that can drive a difference of 10pc, which is the difference between making money and not making money.”

Despite Zillow’s exit, iBuying – the “i” refers to instant – is a growing phenomenon. America’s biggest player, San Francisco company Opendoor, said last week it had bought 15,181 homes in the last quarter, up 79pc year-on-year. 

The Zillow estimate can have a big impact on house prices

Credit: Luke Sharrett/Bloomberg

In comments to analysts, Opendoor’s chief financial officer Carrie Wheeler suggested that Zillow’s problems were its own: “We’re very good at this. This is core to what we do. We have built… very robust pricing systems”.

Another rival, OfferPad, also stuck to its guns, announcing a 258pc increase in the homes it had bought. Both companies, however, remain loss-making, reliant on investors’ and lenders’ faith for the cash to finance house purchases.

Zillow’s rivals have insisted that iBuying is not necessarily a bad business. Trillions of dollars are spent on new houses each year, meaning even a fraction of that presents a huge opportunity. 

Opendoor says its margin on a property is between 4pc and 6pc – largely from service fees and extra services such as mortgage referrals. Hardly what Silicon Valley is used to, but potentially sustainable.

That is, so far. Zillow’s failure happened in the middle of a red hot housing market – what if it took a turn for the worse?

“You’ve got no idea what’s going to happen,” says DelPrete. You’re sitting on literally billions of dollars of inventory. Their value is not going to go to zero overnight. But you just lose a per cent here, a per cent there. And you’re underwater. That would give me stomach ulcers.” 

Zillow’s Barton said this was his main concern when his company shut down iBuying, telling investors it would be “naive” to think that its model could withstand a downturn.

For now, iBuying has remained largely a US phenomenon. While the idea has had some traction in Britain, high transaction costs such as stamp duty and a longer sales process are a deterrent. 

London-based Nested, a start-up that aimed to imitate Opendoor across the pond, has moved away from the model and towards an online estate agency. 

Last year, a property industry body launched OfferHive, which buys houses in cases where stamp duty does not apply. And Alex Chesterman, founder of UK property site Zoopla, said the company steered clear from the idea as it was too capital intensive and risky.

But Opendoor’s boss Eric Wu insists the move to iBuying is inevitable. “We’re in a generational shift in housing from offline to online,” he said last week. 

What starts in Silicon Valley has a tendency to spread. But if both property and tech are prone to bubbles, the combination has the potential to be calamitous.

Leave a Reply

Your email address will not be published. Required fields are marked *