In announcing the Zillow Prize yesterday, the Seattle-based real estate media company challenged “data scientists, engineers and visionaries” to tweak what Z believes is “one of the highest-profile, most accurate and sophisticated examples of machine learning,” according to a company press release.
Zillow provides information about homes, real estate listings, and mortgages throughout the United States via their website and mobile applications.
The Zestimate home valuation calculates what is believed to be the eventual sale price of a home. The U.S. median absolute percent error currently stands at 5 percent, improved from 14 percent in 2006, according to Zillow. Its debut in 2006 presented the first opportunity for homeowners to gauge their home’s estimated value, data previously only available to real estate agents, property appraisers and lenders.
“We still spend enormous resources on improving the Zestimate, and are proud that with advancements in machine learning and cloud computing, we’ve brought the error rate down to 5 percent nationwide,” said Stan Humphries, creator of the Zestimate home valuation and Zillow Group chief analytics officer. “While that error rate is incredibly low, we know the next round of innovation will come from imaginative solutions involving everything from deep learning to hyperlocal data sets — the type of work perfect for crowdsourcing within a competitive environment.”
Zillow Group’s estimating tool is not without controversy. Home owners and home buyers argued over its accuracy, as each side of a transaction sought get a leg up. Last year, Zillow co-founder and executive chairman Rich Barton described Zestimate as “very provocative and personal and a little voyeuristic.”
Earlier this month, Zillow reported revenue in the first quarter increased 32% to $245.8 million from $186 million a year earlier. Analysts had forecast $236.03 million. Earnings were 11 cents per share, beating the consensus analyst estimates of 5 cents per share.
Use of Zillow’s mobile apps and websites reached an all-time high of nearly 180 million unique users in March, the company said in announcing the results. More than 166 million monthly unique users on average visited the sites during the first quarter, an increase of 7% year-over-year. Z brands include Zillow, Trulia and, since March, StreetEasy.
The contest is being administered by Kaggle, a platform designed to connect data scientists with complex machine learning problems. It will be held in two rounds, the public qualifying round, now open and concludes Jan. 17, 2018, and a private final round that kicks off Feb. 1, 2018 and ends Jan. 15, 2019.
Z stock has gained 27% in the past three months, leading up to Memorial Day, the traditional start of summer home buying season.