Zillow® launched the most sophisticated and accurate Zestimate® home valuation in history on Thursday, using real-time data and computer vision to identify and value improvements to individual homes.
The Zestimate was launched in 2006 using facts from public records to estimate a home's value. With today's update, it can now, in a sense, "see" in photographs features that humans would instantly understand, such as curb appeal and natural light. The new Zestimate uses neural networks and computer vision to distinguish between high and low-end finishes and to incorporate the value of features like updated bathroom fixtures, fireplaces, and remodeled kitchens. The Zestimate also now uses real-time data from for-sale homes, including list price and how long a home has been on the market.
Today's launch incorporates winning ideas from Zillow Prize®, a two-year, $1 million data science competition. For homes listed for sale, the error rate is now less than 2%, meaning half of all Zestimates fall within 2% of the home's eventual sale price. Sellers, listing agents, and homeowners provide photos for the Zestimate to evaluate. Homeowners can claim their home on Zillow to edit, add, or remove photos at any time in just a few simple clicks.
While the Zestimate can't interpret home features by name, the way a human would, the image recognition model can classify patterns in the pixels of photographs and correlate them to home value. For example, while the human eye sees tile or granite countertops, the Zestimate identifies two different pixel patterns.
"The new Zestimate was inspired by the way the human brain interprets scenes, objects, and images," said Stan Humphries, Chief Analytics Officer and creator of the Zestimate. "It's a big leap forward, because it means the Zestimate can now understand not just a home's facts and figures, but its quality and curb appeal. By training neural networks using the millions of home photos on Zillow, the Zestimate now values the features people are most proud of – like new granite countertops, a remodeled bathroom, or a meticulously landscaped backyard."
The Zestimate is a starting point in determining a home's value and is not an official appraisal. Zillow calculates Zestimates for more than 100 million homes nationwide. The company's 13 years of experience estimating home values is an important aspect of Zillow Offers®, Zillow's home-buying and selling program, and fulfilling its mission to change the way people buy and sell homes. In eligible Zillow Offers markets, Zillow buys a house, prepares it for showings, and then quickly lists it for sale – all on the consumer's timeline. The Zestimate is not impacted by a homeowner asking for, accepting, or declining an offer from Zillow on their home, and receiving an offer will not change the Zestimate.
Recently more than 3,800 data science teams from 91 countries competed in the Zillow Prize competition to improve the accuracy of the Zestimate. The ideas from the winning solution include sophisticated machine learning techniques to identify and take into account special home features and location details, such as number of bathroom fixtures, commute times, and even a home's proximity to a park or freeway.
To calculate the Zestimate, Zillow uses data from county and tax assessor records, and direct feeds from hundreds of multiple listing services and brokerages. Additionally, homeowners have the ability to update facts about their home to ensure their Zestimate has the most accurate, up-to-date information. More than 70 million homes on Zillow have been updated by the community of users. The Zestimate uses artificial intelligence and other advanced technologies such as computer vision and a deep-learning-based neural network to incorporate data from photographs into the Zestimate.