Episode Summary: Big data is often a buzz word, but if you’re trying to quantify data around homes in the U.S. and pair that with hard to quantify information – like images – you’re likely running into the frontiers of machine learning technology. This is something Zillow deals with daily. In this episode, Stan Humphries, chief analytics officer and economist for Zillow, speaks about where they’re leveraging machine learning and artificial intelligence (hint: almost everywhere), and what he believes are the keys for deriving real ROI opportunities using this technology. Humphries also offers insights for how other companies can model the successful decision-making processes and implementation strategies used by Zillow.
Expertise: Corporate strategy, data science and engineering, marketing and business analytics
Brief Recognition: Stan was one of Zillow’s earliest pre-launch employees and is the creator of the Zestimate and its first algorithm. Prior to joining Zillow, Stan spent five years at Expedia, where he ran the advanced analytics team. Before Expedia, Stan served as a researcher and faculty member at the University of Virginia, and was previously a Presidential Management Fellow where he served at NASA, the Office of Science and Technology Policy in the Executive Office of the President, and the Technology Administration within the Department of Commerce. Stan has also served in the U.S. Peace Corps, where he taught high school physics and chemistry in the West African country of Benin. Stan has a BA from Davidson College, a MS in foreign service from Georgetown University and a Ph.D. in government from the University of Virginia.
Current Affiliations: Chief Analytics Officer & Chief Economist of Zillow Group; Member of Fannie Mae’s Affordable Housing Advisory Council and the Commerce Department’s Data Advisory Council; Member of the Visiting Committee of the Department of Economics at the University of Washington
(1:13) I don’t think many people are all that familiar with how AI or machine learning is part of what you guys do – where does that play a role at Zillow?
(3:55) When you mention estimate for home values in certain areas that are more granular…is that something that is informed by a ML or AI process?
(9:13) The other thing you had brought up has to do with home improvement through Zillow, Diggs…where people can say, what would it cost, if I wanted to get a new roof…you have some degree of ML being applied to those projects as well?
(14:45) What’s some of your advice for really teasing out proper ROI from AI in industry?
(18:47) One thing you’re talking about as a key behavior…is that you have analytic people at high-level business meetings helping guide big objectives…it sounds like in being data-driven, that’s a behavior that people who even want to think about machine learning should already be engaging…
Related Emerj Interviews:
How Business Event Data and Predictive Analytics Help Deliver Better ROI (Nicholas Clark, CTO of DoubleDutch)
Predictive Analytics Offers Customized Solutions to Complex Problems (Swami Chandrasekaran, Chief Technologist & Executive Architect for IBM Watson, Industry Applications & Solutions)
How Machine Learning Shapes Your eBay Experience (Zoher Karu, VP, Chief Data Officer at eBay)