Every Wednesday on AltSearchEngines, we go visit the offices of an alternative search engine to hear from the CEO.  Today, we’ve asked Bob Chandra, founder of grayboxx to tell us about their unique project.

Bob Chandra:  Founded in 2005, grayboxx is a local online search site that provides extensive rankings and recommendations for local businesses and professional services in the United States using our PreferenceScore technology. We operate out of a converted restaurant overlooking a little pocket-canyon in the town of Saratoga, near Silicon Valley’s southwest corner. We’re different from the other local search engines because our results and quality measurements are based on a vast collection of anonymous data about people’s everyday interactions with local businesses. Other local search engines rely completely on user-reviews, which are important and useful but ultimately lack in scope and geographic coverage.

Scope of User Reviews: In considering scope, there are more than 7,000 different types of businesses in the United States. User review Web sites give feedback on slightly more than 200 categories. Few of us are highly motivated to write a user review about our experiences with a plumber, a proctologist, or a bail bondsman. Our method of capturing interactions with businesses means that users don’t have to explicitly volunteer information; instead we have popularity scores on ALL the kinds of businesses they patronize. Grayboxx provides feedback on more than 6,000 types of businesses.

Geographic Coverage of User Reviews: In terms of geographic coverage, almost all user reviews are written for the largest 15 or so cities in the country, but more than 70 percent of the Internet population in the United States live outside these metro areas. This means that user review-based local searching doesn’t give reliable results for most of us.

Comparisons:  Grayboxx works equally as well in small cities and towns as it does in large cities, but the difference is most striking outside of the major metro areas. For example:

Using Yelp to search for a reliable “building contractor” in Lewiston, Idaho yields a single result with no user review attached. An identical search on grayboxx returns a list of 157 contractors in Lewiston, each ranked in popularity according to active expressions of community interest.Also, a search on grayboxx for a “notary public” in Pensacola, Florida provides 20 results, with most ranked by popularity according to neighbor-recommendations. An identical query on Citysearch yields a mere six results, all unranked in terms of popularity and all lacking user reviews.Yahoo! Local provides 24 results when queried for “tire repair” in Grand Junction, Colorado, but only two of these businesses were reviewed, and each received just one review. Grayboxx returns a more statistically significant sample of more than 100 results, with 45 of these ranked according to active expressions of community interest.So How Does It Work?  PreferenceScore accurately ranks the popularity and quality of businesses by paying attention to the anonymous actions of a community. Some of the numerous ways that people demonstrate preference for a business are through their purchasing (or returning) choices, restaurant reservations, an entry in an online address book or calendar, a mail order, a visit to a Web site … and the list goes on and on.

To further elaborate on grayboxx’s unique scoring methodology, when all private data tied to a community’s actions has been removed, PreferenceScore refines the information into a simple numerical ranking that shows a business’ popularity and actual human foot traffic. For example, a search on grayboxx for “bicycle shop” in Abilene, Texas yields Bike Town as the number one result as measured by grayboxx’s neighbor-recommendations system. In 2007, Bike Town was recognized as Abilene’s best local bike shop in the Reader’s Choice Awards sponsored by the Abilene Reporter-News, illustrating how grayboxx’s automated recommendation system maps well to the businesses that Abilene residents choose as best.

Conclusion:  User reviews are important, but really, they’re just the icing on the cake. In the few instances when they are available, grayboxx does provide links to third-party user reviews. Ultimately, through our automated ranking and recommendations technology we’d like PreferenceScore to do for local search what PageRank did for the Web.

This entry was posted on Wednesday, October 17th, 2007 at 9:26 pm and is filed under CEO Views. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Sphere: Related Content