Do Social Media Recommendations Herald The End of Google’s Dominance?

Just read a fascinating if slightly over-egged article on Mashable called “5 Reasons Google And Search Won’t Dominate The Next Decade“. The suggestion is that through a combination of predictive suggestions – i.e. offering you things you might want, based on what you’ve already said or demonstrated you like – and recommendations from friends or strangers, either directly sought or channelled to you via various degrees of connections (i.e. friends of friends of friends) – you will be able to get what you want without searching for it.

The criticism of search is pretty valid. Facebook has just surpassed Google in traffic rankings in the US – possibly evidence that, even though Google has revolutionised how we find information, it cannot hope to be as accurate as a recommendation from a friend or peer. Even Google’s personalised search can’t help you find a book you’ll enjoy. For that you need like-minded others. If you’ve signed up to “follow” a book reader who’s views you respect, getting a message to tell you what book they’ve just bought might meet this need.

But socially driven recommendations have their limits. What percentage of your Google searches fall outside your normal buying patterns or interests? We all buy presents, look up random facts, figures, places and people. I would suggest that at least half of my online activity is not predictable or connected to my typical interests. Were this information to be included as “recommendations” to my friends, it’s likely to confuse rather than help.

Geo-location, it is suggested, will provide yet more options for location-based recommendations. While this is probably true, having worked on mobile social networks – for which having an effective location dimension is the Holy Grail – I’m very aware that geo-location has consistently failed to live up to it’s promise. Let’s see if Foursquare and it’s competitors can prove me wrong.