The Rise and Fall of Zillow iBuying
Imagine having access to industry-leading real estate data and using it to buy and flip real estate in one of the strongest markets in history. You could make a killing, right?
In November 2021, Zillow shocked Wall Street by announcing that it was closing down its iBuying venture.
Zillow’s Foray Into Blind Buying of Real Estate
Launched in December 2019, Zillow Offers bought homes sight unseen for cash. Promising higher prices and faster processing times, Zillow Offers expanded quickly. The program utilized Zillow’s proprietary algorithm, the Zestimate®, to provide an indication of value. Zillow’s program would extend offers to sellers based on this data.
Initially, the future looked bright. What went wrong?
Discovering the Shortfalls of AI
Artificial intelligence (“AI”) is truly an amazing development of our time. But clearly, AI has blind spots, too. Zillow learned the hard way. After losing a billion dollars, Zillow management admitted that its artificial intelligence-based estimate of a property’s value was of limited value.
According to a recent NPR article, another issue that plagued Zillow’s iBuying was that the homeowners who took them up on their offers had far more information than Zillow. While Zillow’s AI could quickly assess neighborhood and market details, the homeowners knew more about potentially significant problems such as deferred maintenance and past damage that could escape the algorithm. So, during a market where bidding wars were common, these homeowners determined they could get more from an AI buying service than by prettying up their property and showing it in person.
These weren’t the only issues, however. Zillow’s iBuyer segment grew fast…obviously too fast. Mike DelPrete, a real estate technology specialist, said the online company’s high-priced purchases accelerated even as market conditions changed. Zillow purchased more homes in the third quarter of 2021 than it had bought in the previous three quarters combined. [ii]
Zillow’s own shareholder letter notes that “higher-than-anticipated conversion rates” and “unintentionally purchasing homes at higher prices” were reasons for its failure. [iii]
Both of these reasons, interestingly, appear to be specific to Zillow.
At the same time, competitors were still handling their internet home buying profitably. They appeared to do this by adjusting to market conditions faster than Zillow. For example, competitors Opendoor and Offerpad, beginning in July 2021, realized that the market was cooling off. They began offering less for homes. Meanwhile, Zillow continued to pay high prices.
A Million Here and a Billion There
At the same time, Zillow’s move was controversial since it effectively cannibalized opportunities from its main customer base—homebuyers and real estate agents. Even worse, now many believe Zillow intends to sell its inventory primarily to corporations. Those are the same ones that renovate and offer the homes back to the community as rentals, making tight residential markets even tighter. According to a Bloomberg article, Zillow let these firms have the first shot at these homes before sales were opened up to individuals. [v]
Comparison to BPO
This entire situation questions the effectiveness of Zillow’s Zestimate® tool. Nevertheless, the industry continues to use this as a data point to help assess a home’s value. DelPrete notes that “it’s a good tool for what it is” but that “it’s a mistake to think this tool can accurately predict house prices now or in the future.”
As always, it’s better to put more weight into tools that have proven their value over the years, like the BPO (Broker Price Opinion). Yes, there’s a moral to this story. Until AI is farther along, it’s safest to rely on one other ingredient….the human touch.