Think BR: Facebook's Graph Search - a first impression

One thing we know is that Graph Search is not a direct competitor to Google, or at least not Google as we currently know it. However, Google’s increasing interest in the area - as represented by flirtations with personalized results, and also Google Now for mobile - means that Facebook may have stolen a march on the next big thing.

Palming off traditional web queries to Bing may also prove to be the largest traffic boost the Microsoft engine has ever had, but that’s no reason to think that users will necessarily embrace it any more than they have.

After all, Bing has been Facebook’s mapping tool for a while now and has not exactly caused a great migration from Google’s product.

There is some excitement in some quarters about the potential for ad revenue fromGraph Search. However, I think some people are rather over-egging it, at least for the time being. Before Graph Search can become a viable commercial product, it needs to be something that has real consumer value.

Currently, it is unclear as to how sophisticated the tool really is. It relies very heavily not only on users actively providing information that they do not necessarily always provide - check-in location, employer, reviews - but also the underlying systems that join the dots between multiple, similar-but-different data points.

It’s not a straightforward task. For example, I went to Corsica for a holiday last year, but when labelling my photos I discovered that Facebook thought (and still thinks) it was in Haifa, Israel. 

The other obstacle facing Graph Search is that it does not appear to have a timetable to appear on mobile devices. Given that one of the functionalities that I can see offering real consumer value right off the bat is looking up places friends have been to and liked (restaurants, sightseeing, venues etc), this seems disappointing and somewhat myopic. After all, that particular breed of search would be right at home on a mobile device when out and about. On desktop? I’m not so sure.

The other major user value that I can envisage would be the ability to look up photos of friends in/at/during/doing a particular thing. But then, I’m not entirely sure how that would lend itself to advertising. Perhaps if you are looking at your friends at Disney World, Disney (or indeed someone else) could display a relevant offer.

I suspect that Facebook will rely once again on innovative start-ups and PMDs to jump in and develop the product, in much the same manner as Facebook Apps before. It also represents another incentive for anyone building apps of any description - web, mobile, or Facebook - to incorporate the Open Graph.

After all, if Facebook continue to expand its list of ‘custom edges’ - that’s allowing brands to create their own action verbs such as ‘Listened’ or ‘Watched’, to sit alongside ‘Like’ - then presumably that will feed into Graph Search too. So if I am building an app around cooking, I code in the customedge ‘Cooked’, and then my app and/or brand might show up in the hypothetical search: [My friends] in [London] who [cooked] in [the last year].

Whatever happens, it’s another big push from Facebook into using big data to personalise the life digital. Given that Facebook’s long-term ambition, especially in developing markets, is to be your passport for the web, this makes a lot ofsense.

At this rate, the Facebook of the future may soon start to challenge the definition of what a social network actually is.

Philip Dyte, paid social planner, iProspect