Zillow Taps AI to Improve Its Home Value Estimates
Stories of people getting cash offers for their homes at tens of thousands of dollars above the asking price has become normal. This year, stocks in the US real estate market hit an all-time high while home prices hit an all-time high. Redfin CEO Glenn Kelman recently highlighted the insanity with a Tweeter telling the story of a home buyer who offered to name his firstborn after the seller’s name and was refused.
As the hot US real estate market began to overheat, in February, Zillow began making initial cash offers to buy homes based on his price estimate. Now Zillow has updated its algorithm behind these estimates in a way the company says will make them more accurate and allow Zillow to offer to buy more homes.
Homeowners of approximately 900,000 properties were initially eligible to receive automatic cash offers for the purchase of a home. Zillow’s chief analytics officer Stan Humphries said the change to his artificial intelligence would increase that pool by 30%. A company spokesperson said Zillow Offers could close sales in as little as a week.
Zillow previously determined home values using nearly 1,000 variations of derivative algorithms for local markets. From now on, all prices nationwide will be decided by a single neural network. Zillow says the new algorithm will reduce its pricing errors by 11.5% for off-market homes in nearly 30 regions across the United States. When compared to the previous version of the algorithm, errors decreased the most in Phoenix, followed by San Antonio, Tampa and Houston.
Using the new algorithm, Zillow will update its estimates of the value of 104 million properties in 23 US markets more frequently. When the company was founded in 2005, the ratings were updated monthly. More recently, they have been updated several times a week; now some estimates may be updated daily.
Zillow home ratings are popular topics of local conversation, especially in hot housing markets. A recent Saturday Night Live skit compared Zillow surfing to phone sex and joked that for people in their late 30s, “the pleasure you once got from sex now comes from looking at other people’s homes.”
Zillow has been developing its home valuation algorithm for 15 years. A move towards a single neural network began in 2019 following a public competition to improve its estimates, which drew more than 3,000 teams vying for a million dollar prize. Two of the three finalists in this competition exploited approaches based on neural networks, a form of deep learning that Zillow says is more able to recognize the relationship between the data used to estimate a home’s valuation. .
For example, Humphries said the new approach is better able to capture the value of waterfront properties or how valuations are influenced by the size of nearby homes.
“The old approach would be difficult for this county to understand the value of waterfront homes,” he said. “Neural networks that are being trained across the country could take information from other parts of the country about the value of the waterfront and apply it in a local geography, even if there wasn’t a lot of houses like that in this geography. “
Home value estimates are based on dozens of factors about a property, including square footage and location. Some listings also include data from tax assessments and sales records. Since 2016, the company has also been using computer vision systems that draw conclusions from photos of real estate listings.
By expanding its home buying program, the company hopes to bring e-commerce one-click sensitivity to the real estate market, CEO Rich Barton said. “Ultimately, we want to have most of the homes in the country inside that buy box of direct offers,” he said.
The company is also hoping to capitalize on what it calls “the big shake-up,” as more people seek to work from home more often and Millennials look to buy homes.
Billions of people check Zillow’s lists every year, many, Barton said, to dream or indulge in a bit of voyeurism. For them, Zillow is trying to ‘engage or entertain’, but for people who buy and sell homes, Zillow wants to sell more services. Beyond buying homes, Zillow is also increasing services such as home loans, home insurance, and assistance through the escrow process.
More great WIRED stories