The Future of Prediction: How Google Searches Foreshadow Housing Prices and Sales
We demonstrate how data from search engines such as Google provide an accurate but simple way to predict future business activities. Applying our methodology to predict housing market trends, we find that a housing search index is strongly predictive of future housing market sales and prices. For state-level predictions in the United States, the use of search data produces out-of-sample predictions with a smaller mean absolute error than the baseline model that uses conventional data but lacks search data. Furthermore, we find that our simple model of using search frequencies beat the predictions published by experts from the National Association of Realtors by 23.6% for future US home sales. We also demonstrate how these data can be used in other markets, such as home appliance sales. This type of "nanoeconomic" data can transform prediction in numerous markets, thereby improving business and consumer decision making.
We thank Karl Case, Avi Goldfarb, Andrea Meyer, Dana Meyer, Shachar Reichman, Lu Han, and Hal Varian as well as seminar participants at the NBER, MIT, the Workshop on Information Systems and Economics, and the International Conference on Information Systems for valuable comments on this research. The MIT Center for Digital Business provided generous funding.