Recommendation site Hunch has been going through a reboot lately. Back in June, it stopped showing results to people who are not signed in, and earlier this month it redesigned its home page to offer personalized taste recommendations across a wide variety of categories such as dog breeds, U.S. national parks, camcorders, soft drinks, luggage, and film directors.
Now it is testing out local recommendations on a map with a sidebar showing restaurants, nightlife, hotels, spas, clothing stores, and more. Hunch local tries to figure out which spots your friends on different services might like (you can sign in with your Twitter or Facebook account) and offers them up at the top of its local search results. Each spot has a corresponding pin on the map. You can filter by different types of venues, and there is also a slider which lets you select more personalized “unique” results or more “popular” ones.
The restaurant recommendations it gave me are pretty decent for an early alpha. In New York City, it suggested Katz’s Delicatessen (a classic), Artichoke Pizza (trendy), Momofuku Noodle Bar (if only I could get in), and Hundred Acres (my wife went there last night! no joke). Each spot contains links back to profiles on Foursquare, Yelp, Hunch, or other places, just like a local search engine.
“It starts out looking at what your Facebook, Twitter, Foursquare friends like, and then gets smarter over time as people give feedback,” says Hunch founder Chris Dixon. He notes that the feature just launched in alpha and is far from fully baked. His plan is to partner with Foursquare, Yelp and others to get their direct data feed of all of their places, which should improve the data.
Hunch took a hit when it started requiring that all visitors sign in. According to comScore, the site went from about 750,000 unique visitors in May, to 250,000 in June, but it already started rebounding in July to 350,000. These numbers undercount Hunch’s actual visitors by at least half, but the trend is right. By focusing on its core “taste graph” and giving people actionable recommendations every time they log in, Hunch is making the right moves to get back on track. Before, Hunch was interesting, but vague. I wasn’t really sure why I needed to go there. Now there are more and more specific reasons, and Hunch Local is something I will definitely go back to try out when I need to find a new place for lunch.
Hunch is a consumer web application that is building the “taste graph” of the internet, mapping every person on the internet to every entity on the internet and their affinity for that entity. An entity could be a web site, a cookbook, a hotel room, a celebrity, a restaurant, etc. Hunch creates a taste profile by asking them a series of questions which range from serious to profound and subsequently can make recommendations personalized to that user, which live in...