Last Updated Apr 9, 2010 5:24 PM EDT
Housing is another areas where digital data can add a lot to a forecast. The online phone directory Yellowbook says an upturn in searches related to remodeling tipped them off to a housing rebound months before major retailers like Home Depot (HD) and Lowe's (LOW) released their numbers. Researchers at MIT used a similar strategy with Google Trends to predict a recent surge in home sales. Not only did they spot it two months before the National Associaton of Realtors, but their predictions were far more accurate. At the more local level economists are using daily data streams from urban networks as indicators. San Francisco's chief economist, Ted Egan, didn't want to wait months for the state to release sales tax data. So he started following the amount of traffic going through subway stations near the cities largest shopping district instead. This data says Egan, "Is more timely and more focused on our city than the more well-established official statistics."
Egan says he would like to build a program that tracks Cragislist, giving him a comprehensive sense of how the local market is shifting for just about everything. Of course, context is still key. "Back in 2008 when financial markets were crashing, we saw an uptick in searches on 'Luxury Goods'"," says Google's chief economist, Hal Varian. "We dug into this a little bit and discovered that this was due to an uptick on searches for "gold"...not too surprising given what was happening at the time."
Image from Flickr user bobster855