Data mining search engines to create innovative products

In the Wall Street Journal (sorry, no link available), Kevin J. Delaney explains how companies of all sizes are mining search engine data to come up with innovative new product and service offerings. In one example, Austin-based National Instruments found that engineers were searching for hardware products with the search term “USB” (Universal Serial Bus), so the company decided to start selling new versions of its products with USB interfaces instead of requiring users to install new circuit boards inside their desktop PCs. Very smart. In another example, Siemens Medical Solutions decided to base the name of its new personal health-card product based largely on search-related data gleaned from Yahoo and Google.

Search engine companies like Google and Yahoo actually seem to be encouraging users to explore all their search data, confident that they can convince users to pay even more for highly-targeted advertising solutions. A senior manager for Google Analytics, for example, compares the search data during a certain time period to “online votes” for product features and service offerings. The more votes (i.e. search queries), the greater the likelihood of success. In one example, Google thinks it can predict box office revenue based on a comparative study of search queries for certain movies. (Maybe this is over-simplifying things, but people searching for “big,” “fat,” “Greek” and “wedding” would be interested in a film like My Big Fat Greek Wedding?)

[image: Pickers Getting Coal From Waste Dump]