Social Search Acumen

27 Feb 2011|Leigh Marinner

Robert Scoble talked about something I have been blogging about for years – the unfulfilled promise of social search.

And he put a name on it – malleable social graph. (Although I think we could come up with a better name)  This is the need to get advice or recommendations only from those who know something about what you’re interested in.  It’s the Zagat’s problem – lots of people who contribute reviews aren’t foodies and their recommendations don’t matter to me.

According to Scoble, a malleable social graph “changes based on conditions you set in motion…”  One of his examples is: “One of my favorite sushi places in San Mateo, Yuzu, has a few negative reviews. Why is that? It’s called the “masses are asses” problem on Yelp. Here’s what happens: Yuzu is a place that is awesome for advanced sushi lovers. I’ve eaten sushi all over the world in places like Tokyo, Yokohama, New York, London, and other places. I love advanced sushi. I look for restaurants who do sushi well. But most people aren’t like me. Most people don’t even like sushi. So, if they get dragged to a place like this they try to order “Americanized” sushi like, um, California rolls. Or fried crap I can’t even pronounce. That is NOT sushi. Anyway, these people, er, newbies, get to Yuzu and find that all the other “non-sushi” stuff sucks. So they rate it low. Me? I could care less about all that other non-sushi stuff when I am looking for a sushi restaurant, which is why I rate this place five stars. Now, Yelp does NOT have a malleable social graph. We can’t filter out all the “sushi newbies” who don’t like sushi anyway.”

A year ago Scoble was predicting this is a major way that Facebook can take over from the check-in services like Foursquare.  Well, it hasn’t happened.  And it’s far more complicated than just matching someone up based on location.  Just because a friend of mine is near me doesn’t mean I trust their sushi recommendation.  Almost a year later, Facebook hasn’t solved this problem. Klout (a service that tracks the influence of individuals in social networks and can be imported into other products (e.g. StockTwits) where you really want to know more about the person giving you advice) may be making some progress, but it’s hardly a household word.

Can we come up with a better moniker than “malleable social graph”?  Social search acumen? Discernment within the social graph? Discrimination in social search?

prev next