Google dips hands into messy local search

December 21, 2009

Google’s mobile ambitions make this a natural pairing. Yet Yelp’s untidy human element may fit awkwardly with Google’s clean algorithmic bent.

The biggest growth opportunities in search are local and mobile. The increasing prevalence of smart phones, and the phenomenal increase in iPhone data traffic in particular, shows that people want to find things while on the go. It’s a rich seam for marketers too – the closer a customer is to a business’s doors, the more likely they are to walk through them. If Google can capture this localized search traffic, advertising revenue should follow.

Yelp fits the bill in several ways. Its customer reviews are another valuable set of data that Google can serve up hand in glove with its other data and applications to users in potentially new and useful ways. It would, for example, encourage small advertisers to go to Google rather than directories or local papers for their advertising. And, of course, a deal prevents rival Facebook from getting a leg up.

Yet swallowing a Yelp could cause indigestion. A $500 million deal would come in at more than 15 times 2009 sales. Moreover, Yelp relies on users to write reviews. About a third of all reviews were five stars, or the highest possible. Only 7 percent were one star.

This may have something to do with the site’s dominance in California, a state known for its puppyish optimism. It could be writing negative posts makes people uncomfortable. Or perhaps business people simply write good reviews for their own establishments. Either way, the distribution of results looks too skewed towards favorable reviews.

Google, of course, has plenty of experience with people trying to game its system of search. But schemes are easier to find in large sets of data because their distorted wakes are more visible. It’s harder to figure out if the one glowing review of a corner pizza shop is real or a plant. Google’s search dominance comes from users’ trust that they are being served accurate, mathematically sound, results. The success of local search will depend on making sure what’s found is believable.

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see