Why Would Zillow Pay $1 Million to Improve a Real Estate Price Estimate Algorithm?

Zillow's Zestimate has been a cornerstone for homeowners, appraisers, and real estate enthusiasts for over a decade. This popular feature is a testament to Zillow's dedication to homeowners seeking accurate property value estimates. The complexity of the Zestimate algorithm, which relies on a significant amount of public records data as well as MLS multiple listing services, is not only a significant marketing tool but also a project with the potential to revolutionize the real estate market.

How Does an Automated Valuation Model Work?

An automated valuation model (AVM) like Zestimate must consider numerous variables that cannot be easily accounted for by a raw mathematical equation. These variables include:

Mature Plantings and Unique Views: How do you account for a mature cactus garden or a breathtaking view of a mountain range? Distressed Sellers and Obnoxious Neighbors: How do you quantify a negative influence on the property? Buyer's Gut Feeling: What about a prospective buyer's impression of the property, which may affect their decision? Curb Appeal: A visually appealing facade can significantly impact a property's value. Non-Arms Length Sales: How do you evaluate transactions where the buyer and seller have a pre-existing relationship?

The Role of Fuzzy Logic

To address these complexities, Zillow's algorithm uses fuzzy logic. This method involves scanning the MLS narrative for keywords that trigger specific value changes or quantity descriptions. For example, if the MLS mentions something like 'sunny southern exposure' or 'view of the mountains', the algorithm can infer an increase in value. Similarly, phrases like 'overlooking the freeway' or 'next to a noisy nightclub' can trigger a decrease in value.

Investments in Continuous Improvement

The $1 million investment in this project from Zillow is seen as a significant allocation, reflecting the company's commitment to improving their chief marketing tool. As Stan Humphries, Zillow's chief economist, explains, the accuracy of the Zestimate is crucial for maintaining trust among users. Each year, as the algorithm learns from real estate booms and busts, it becomes more sophisticated. The accuracy of the Zestimate has been gradually improving, with some credit given to events like the Great Recession for providing valuable data for the algorithm to learn from.

The Impact of Perfecting an Algorithm

If Zillow can prove that their algorithm is scientifically accurate, the implications could be far-reaching. For instance, financial institutions like banks and lenders might start relying on the Zestimate as a cost-saving measure for property valuations. This could eventually challenge established real estate interests and transform the entire model of buying and selling real estate.

Zillow is not yet there, but every year the accuracy of the Zestimate improves. Innovations in technology and the increasing number of bright minds entering the workforce contribute to new ideas and theories being tested and examined. While this quest poses a challenge for real estate professionals, it is exciting to see a company invest such resources in developing this technology.

As Stan Humphries states, while the project may one day challenge traditional professions, it is a promising venture for the future of real estate.