Reports that Google nearly paid about $500 million to acquire local review site Yelp are drawing even more attention to the already-hot local business search/review space and fueling debate over the merits of various local search models. While some would say that real-time, location-aware social search will ultimately trump sites like Yelp, I think it will instead enhance Yelp in a way consistent with my general thesis that social networking will add a layer of value to traditional search rather than really displacing it.
My colleagues, Om and Liz, used the convincing example of restaurant recommendations to demonstrate the competitive threat to Yelp and its ilk posed by certain kinds of social media. But move beyond restaurants to other search topics, like landscapers or divorce lawyers, and you soon confront the more fundamental limitations of social search.
As I’ve said before, the unique value of social recommendations comes mainly in two ways:
- You know more about the people making the recommendations. If five friends tell me where to eat, I’m taking the advice of the guy I know first-hand is a great cook and discarding the advice of the guy whose wife owns the restaurant he’s recommending.
- The people giving you advice have a stake in the outcome, since they care about you or, at least, know they’ll run into you again.
(Liz hints at a third concept, that social recommendations are better because your friends know you and your tastes. That’s important to include but a relatively minor factor, I think.)
That said, social recommendations are limited in two directions: If your social group is too small, the suggestions are less helpful. (For example, you need a divorce lawyer but none of your friends has been divorced.) If the group is too large, your connections with recommenders could be too loose to yield the aforementioned three benefits, which Liz pointed out come from “intimate” groups. In that sense, in fact, the value of social networks may be waning more every day, as we keep opening our circles to folks we barely spoke to even in high school. As we further dilute the meaning of “friend,” Yelp gains the upper hand in this debate.
The ideal local recommendation engine would include detailed information like you find in Yelp but with a layer of social media on top – a mashup of Yelp with Facebook or Twitter or Foursquare, say, or some combination thereof, with social recommendations (real-time and not, location-based and not) given prominence when available. It would also include relevant hard data like location, office hours, etc., and a mix of reviews from strangers – both designated critics like restaurant reviewers (and perhaps, as a premium service, ConsumerReports.com and similar authorities) and first-hand customer accounts from average Joes.
Netflix’s recommendation system is not a bad model to emulate here, in part. Though Netflix can’t really fully leverage the one element – location – that will be most important in local search, it also includes an element that local search has yet to truly utilize: recommendations from anonymous people “like you” – people with matching tastes or perhaps relevant traits in common, like income levels. Both Yelp and its rivals in the social networking world could take a stab at addressing that functionality. Exactly how they do that is one of many challenges to tackle next.