Zillow Withdraws from iBuying Market in What’s Being Called an “Algorithmic Failure”
While the ZillowNow AI algorithm did not have the accuracy needed to predict home prices, causing a loss on every sale in its last quarter, iBuying seems to have a future.
The retreat of Zillow from the market and flipping houses - buying them using real estate technology in a practice known as “iBuying,” fixing them up a little and selling them quickly - is being called “an algorithmic failure” in some circles, a failure of AI to deliver on the business plan.
In a quarterly earnings call on Nov. 2, 2021, the company said it would write down more than $540 million as a result of its exit from the iBuying business, according to an account from CNN. The company also announced it was cutting 2,000 jobs, about 25 percent of the staff.
Zillow executives stated that the company bought homes in the previous quarter for prices higher than it believes it could sell them. “We’ve determined the unpredictability in forecasting home prices far exceeds what we anticipated and continuing to scale Zillow Offers would result in too much volatility in earnings and on the balance sheet,” stated Rich Barton, Zillow’s cofounder and CEO.
Zillow launched the service in 2018 in Phoenix and Las Vegas. It was operating in 25 cities most recently. The company ended the quarter with 9,790 homes in inventory and 8,172 under contract that it will still purchase and sell over the next six months. In the latest quarter, Zillow reported buying 9.680 homes and selling 3,032, losing $80,771 on each sale, according to the CNN report.
Pundits Assess
AI pundits weighed in with their assessments. AI angel investor Ron Erickson stated in an account in GeekWire that, “Everything in our world cannot be managed with algorithms parsing big data! They failed to think through the implications of buying all those homes.”
Technology entrepreneur Galen Ward stated via Twitter, “It would be fascinating to find out where in the stack Zillow's failure lives. Was it incorrect use of ML? Too much trust in ML? Aggressive management that wouldn't take "we aren't ready" for an answer? Wrong KPIs?”
York Baur, CEO of MoxiWorks of Seattle, which sells cloud-based software tools to residential real estate brokers, said Zillow put too much faith in machines to do what humans can do better.
“All the AI and machine learning in the world isn’t yet up to the task of the complexity of valuing a home in a rapidly changing market, and this move by Zillow is proof,” Baur stated to GeekWire. “They invented computer home valuation with the Zestimate 15 years ago, and it’s still not accurate after 15 years of refinement and billions of dollars invested.”
Baur added, “What this says to me is that we need to stop over-applying technology in an effort to replace humans, and instead focus on applying technology to make humans better.”
Tomasz Piskorski, a professor at Columbia Business School who researches iBuying, stated Zillow appears to have reached beyond the capability of the algorithm. “There was no problem with the algorithm as long as they stay within the boundaries of the business model and buy cookie-cutter homes that are easier to sell,” Piskorski stated in an account in Wired. “There are a lot of things that affect the valuation of homes that even very sophisticated algorithms cannot catch.”
Despite the setback, iBuyers bought 19 homes every day in Phoenix in the second quarter of 2021, about one in every 20 bought there, and 165 a day across the US, one in every 100, according to Zillow’s survey of the market. Redfin, Opendoor and Offerpad are continuing in the iBuying market.
“In principle, it’s a viable business model, but its viability changes with market positions,” stated Piskorski. “The gold rush has only just begun.”
The iBuying Market is Young; Opendoor Was First in 2014
The iBuying trend is young and not well-penetrated. The first iBuying company, Opendoor, started out in 2014. In September, iBuying accounted for about one percent of all home sales, although up to six percent in some metro areas, according to a recent account in The New York Times.
On average, an iBuyer offers .22 percent less than fair market value for a home, and charges slightly higher fees, about 1.3 percent more than a conventional listing agent, according to the Times account. The advantage for the seller is the process takes days instead of weeks, and can involve few or no open houses and showings to potential buyers.
The business model of Redfin, a full service real estate brokerage founded in 2004, is to undercut the competition based on sellers paying Redfin a discounted fee of 1.0 or 1.5 percent to list the seller’s home. The firm now offers the RedfinNow iBuying service, which represent less than one percent of the company’’s sales. “It’s a new way to liquidate a house,” stated Glenn Kelman, CEO of Redfin, “and a tiny fraction of customers are going to buy or sell a house that way.”
AI is also being employed by iLenders, fully digital mortgage brokers, such as Better.com. Buyers can gain an advantage by not having to tour homes and make multiple offers. If a buyer is approved for a mortgage, which can happen any hour of the day or night, it could provide the entire purchase price up front, allowing the buyer to pay entirely with cash.
Other fintech lenders, including Accept.inc., can make cash offers on a buyer’s behalf, with no added fees for sellers or buyers. iLenders front the cash and make money off mortgage and title costs.
While the experience of Zillow’s iBuying unit may have put a damper on the trend, Opendoor and others are moving to expand their iBuying efforts, and more conventional brokers are looking to add an iBuying service.
One professional sees a bright future for all-cash home purchases. “The mortgage itself hasn’t evolved since, like, the 1900s,” stated Paul Tyger, a purchase manager at Better.com. “But we think that in the future, every transaction will be all cash.”
For the buyer’s sake, the algorithm writers hope for accurate predictions on where a home’s price is going.
Read the source articles from CNN, in GeekWire, in Wired and in The New York Times.