Steve Cheney wrote an excellent analysis of Facebook’s new search tool in Graph Search’s Dirty Promise and the Con of the Facebook “Like”. The problem? Most “Likes” on Facebook are bought by ad agencies, not earned organically. The result:

One direct effect of all this passive liking is an ugly messy data set with a bunch of implicit signals… that are wrong. What happens when your girlfriend types in “restaurants in San Francisco” into graph search and P.F. Chang’s gets spit out because it’s the most-liked restaurant.  Was a bad Chinese chain the kind of serendipity you were looking for on your date? Didn’t think so. 

I also like Ariel Seidman’s take on the challenges Graph Search will need to overcome in Can we make that search box bigger?:

Consumers think in terms of I got a job to do. What product will I hire to do this job? For restaurant searches I hire Yelp. I need a flight to Chicago I hire kayak. I need to start looking for jobs I’ll hit LinkedIn or Indeed. Each of these have their own experience, community, privacy expectations, and detailed data. As BranchOut has shown people do not want Facebook to be the place to manage their professional life, they have hired a different product for that job.

I know I tend to be too skeptical about this kind of stuff, but Graph Search just strikes me as another solution in search of a problem that doesn’t exist.