Why We Need to Keep an Eye on Facebook Graph Search

Since the beta version of Facebook Graph Search was launched just over two weeks ago, everyone seems to be talking about it.  So, not wanting to be left out, I've decided to jot down a few of my thoughts.

For anyone who has managed to avoid all the hype and discussion, Graph Search aims to produce a more personalised form of search than is available elsewhere online.  It's being tested in the USA and already has a waiting list of those people who want to try it out.  Search results will be based on all the information that Facebook users have supplied about themselves and their likes and dislikes over the years.  So, for example, you could search for "an Italian restaurant my friends like" or "a book about English history that teachers like".

To begin with, Facebook is offering searches in four areas - people, photos, places and interests.  And while some of these could be informative and useful, we need to be aware of the potential dangers.  For example, as Carlton Jefferis has pointed out on SmartInsights.com, how happy would a single woman be about admitting her marital status if a complete stranger can search for a list of "single women who live near me"?  Equally, things that we have posted years ago can come back to haunt us.  This is particularly relevant for those who joined Facebook when they were students and now hold responsible jobs.  That picture that you thought was lost in the mists of time could easily resurface if someone were to search for photos from your college days.  Perhaps the answer here is to search for them yourself, as soon as you have access to Graph Search, and try to get identifying tags removed.  In addition, while nothing that you have marked as 'private' will be published in a search, everything else is fair game, so checking what you have set might be a good idea.

Graph Search could be very valuable, though, if someone was looking for, say, a fellow chess-player or a tennis partner, or someone to make up a fourth at bridge in their local area.  And, certainly, Facebook has hopes that Graph Search will come to rival Linkedin when it comes to recruiting.

Another place where Facebook is hoping to challenge rivals (such as Google and Yelp) is in the area of local search.  A recommendation from someone you know is likely to carry more weight than one from a stranger.  So looking for a local restaurant, a hairdresser or a plumber could be made a lot easier . . . or not.  And I put 'or not' because, of course, it all depends on which of your friends are on Facebook (not everyone is), whether the restaurants, hairdressers and plumbers that they would recommend are also on Facebook, and whether your friends have actually 'liked' the pages of those businesses.  While knowing that a friend has 'liked' a business may indicate that the business is good, if a business has no 'likes' it does not indicate that it is any less good.

We 'like' pages for different reasons.  Those of you on Twitter will know that, frequently, people you are following will ask you to like their Facebook pages, often with a promise of reciprocating.  We can thus find ourselves 'liking' all sorts of business pages that we know very little about.  And, indeed, speaking personally, for me it's usually about liking the page, rather than the business.  If the page is attractive and has some interesting posts and is in my sphere of interest, I tend to 'like' it.  It's nothing to do with the business itself because I have no way of knowing if it's reputable or offers good service.  Conversely, I would never think of 'liking' a local restaurant or my hairdresser because that's not what I use Facebook for.

So, will we need to go through the list of pages we've 'liked' and edit it . . . I think so.  Because, while I may find someone's page interesting or attractive, I do not necessarily want to be seen as recommending their business to my friends.

However, local businesses, in particular, need to be geared up to get the most they can out of Graph Search.  Carlton Jefferis recommends checking all the information you have shared to ensure that it's complete and accurate, and continuing to post "timely, relevant and valuable content on your page as this will continue to drive engagement and reach".  He also recommends (as do other writers) focusing on attracting people to your page and giving them a reason to interact with your content.  However, while you can encourage people to post reviews of your business, and even to upload photos, it's important to remember that Facebook forbids any offering of incentives to do this.

A lot of people seem to be getting excited by the idea of Graph Search but I have to admit that, personally, I'm somewhat underwhelmed.  Its success depends so much on the information that people post to Facebook and I'm not convinced that they're going to start 'liking' all their favourite businesses and books and pieces of music just for the benefit of their friends.  In addition, how many people know everyone on their 'friends' list personally?  A lot of people on Twitter will ask you to become a friend on Facebook and you may not know them from Adam.  And so why would we trust their judgement against that of anyone else?  If we're really going to use Graph Search properly, we'll need to prune our list of friends down to those we really know and, perhaps even beyond that, to those whose opinions we really trust.

So could this turn out to be a nine-day wonder?  Danny Sullivan of SearchEngineLand.com quotes Tom Stocky, director of product management at Facebook, as offering the opportunity to find out "all the TV shows that are liked by software engineers" as an example of what Graph Search promises.  But I am left wondering who on earth would want to know that?