One Small Step for Facebook, One Giant Leap…

Perhaps those of you outside the world of local, social search did not experience the ground shake when Facebook announced their new “Keyword Search” yesterday. But make no mistake, whether you’re in the industry, or you’re a regular user of social media, this development augurs a sea change in the way people will find information in the future. Indeed, a thoughtful article by @joshconstine on TechCrunch spells it out in the title: “Facebook Post Search: Strong Recommendations, Yelp Should Worry.”


To briefly recap, Facebook announced they have enabled (on US, mobile iOS right now, Android and the rest of the world later) a keyword search capability. That means you can now search back among old posts you wrote or saw on your newsfeed. The examples they highlighted were city names, as in, you are planning a trip to Berlin and want to know which of your friends might be able to recommend attractions or restaurants. Similarly, Facebook highlighted the opportunity to see dentists mentioned by friends. Cool? Yep. Game-changer? Probably, and we can start with its eventual effect on the status quo.

Why should Yelp worry that people can now search for the word dentist on Facebook? One word: trust. Tens of millions of people go to Yelp every month to search for recommendations, but 47% of people using Yelp don’t trust the data they find, and for good reason, since some estimate that between 20-50% of Yelp reviews are bogus. On Facebook, the recommendations shared daily are between friends, generally provided in response to a fellow group member’s need, or a friends’ plea for help. This makes the Facebook platform a highly desirable place not just to ask for, but also to search, high authenticity referrals.

Now, returning to Facebook’s announcement, I claimed above that it is a harbinger for the future, the not-so-subtle implication was that it’s not very impactful for the present. The trend toward social search is clear, but the utility of this new search on Facebook is going to be quite limited – approaching ineffectual – for the time being. The reason, in another one-word answer, is noise. There will be too much data, and not enough information with this new capability.

Searching by keywords leads to lots of results, some will be good and many will be bad. For all you old-timers out there, think Yahoo search versus Google. The former gave you thousands of results, but you still had to spend a lot of time sifting through the noise to find your desired signal. With Google search, they often seem to know the answer before you can fully formulate the question.

And that’s why we believe WhoDoYou’s service is at least several generations ahead of Facebook Keyword Search for local, social search. We’ve been focused on this problem for a few years now, and have become expert at identifying posts which contain legitimate requests for recommendations or referrals. In fact, we’ve processed tens of millions of posts, and have found that fewer than 5% that include common keywords – such as dentist, barber, plumber or photographer – are requests for recommendations from friends. That means you’ll have to sift through, on average, 19 “bad” results for each “good” one. While it’s cool to see who mentioned the word ‘dentist’ in the last five years, I personally would rather get 4-5 good answers, than dig through 80-100 posts that contain the word dentist.

In fact, we’ve been working for a long time to innovate in converting free text recommendations spread across many posts, into aggregated business listings with all the recommendations gathered into one profile. That way you have some ability to order the results based on who is most recommended. This is a pretty challenging proposition for a variety of reasons, but the results speak for themselves. Below are several dentists that have been recommended nearly a dozen times each in response to friends’ requests for help.

We are proud that each month tens of thousands of users contact businesses found on WhoDoYou. Our goal is, and has always been, to help people find answers to their questions. With all the data out there, it’s a delight for our users to have the information pulled, cataloged, inventoried, and presented in a logical and straightforward way.

We look forward to hearing from you, and would be happy to discuss the challenges and opportunities of an effective local, social search.