Location and social technology is creating a whole new race for so-called serendipity applications – centered around discovering people that are of interest to you. Slowly the building blocks to do this have been falling in place.
It’s not rocket sience anymore: It’s not terribly difficult to check two people’s interest graphs (Twitter) or social graphs (Facebook) based on a checkin (Foursquare) into a similar venue and tell them how they are related. That is exactly what what Sonar does.
When services like PeerIndex (and Klout) analyses people’s Tweets and assign an influence score, and on top of all that tell us what topics they talk about – it makes things a little more interesting. Now you can tell not only that this person and you follow the same person on Twitter, but how influential they are, and that they like talking about political-economy. Sexy. We built an app like that ourselves called Woos.at.
Robert Scoble has just done an interview with Whit.li who claims to be a social curation technology. Its soon-to-launch iPad app analyses your Facebook status updates, likes, comments and friends and then creates a psycho-social profile (through that incredibly exact technology called sentiment analysis – not). And they match you with ‘like-minded’ people based on things like religion or politics. The options on Facebook is rather basic. A simple choice between “Liberal” and “Conservative” is a little too banal and unsubtle for our liking. And it will give Eli Pariser (of Filter Bubble fame) heart palpitations. Are we going to end up in a hall of mirrors? Staring at ourselves?
If only there were more FourSquare checkins, because the places you visit says a lot about who you are, and about your identity. Ethan Zuckerman points to Nathan Eagle, who has worked with Sandy Pentland at MIT’s Media Lab on the idea of “reality mining”, digesting huge sets of data, like mobile phone records. They estimate that they can predict the location of “low-entropy individuals” with 90-95% accuracy based on this type of data.
What all these services have in common is the ability to get at the real you: You are to a big extent a product of who your friends are, where you go, and who you share interests with.
Whit.li produces an API which other companies can use to produce better recommendations (for example). Or so they claim. This person that says the sushi is crap at this restaurant, but do they actually know anything about real non American sushi? This information could also be used to build a better Sonar type app.
Then there’s Blendr from the guys that built Grindr, the gay cruising app. Except this one is for straight people, so while they hope it creates lots of serendipitous bumping into each other, it’s not for grinding. Blendr allows you to chat with people, perhaps based on how close they are to you, eschewing the checkin functionality completely and just indicating whether the person is far away or 500 meters from you. That neatly gets away from the problem that checkins have not taken off.
But it’s a little different from Grindr, because the app assumes straight people need more than sexy profile pics to make chat partner choices. So Blendr asks you a whole raft of interest related questions with over 200 options. An interesting choice because they could have tried to get some info from Facebook (or an API like Whit.li ) or services like PeerIndex for Twitter users.
Will Blendr be as wildly succesful as Grindr? We doubt it, simply because it will suffer for the same reason many other straight chat apps have suffered. Over subscription from males. But it will probably get some traction.
But Blendr does point us rather unambigously in the direction in which most of these crosses between location and social tend to point us. The main reason service builders imagine you’d like to meet relevant strangers (serendipity is just that) is for some blending, or even grinding.
Another case in point is Holler. AllthingsD report:
Holler falls somewhere between group messaging apps, event planning tools and “social discovery” sites. The main action is for a user to broadcast — “holler,” as it were — that they are down to hang out, either to specific friends or to people nearby who have said they are interested in a certain activity or event (dog walking, surfing, a conference).
It’s US only now, so I could not try it. The concept reminds a little of the Situationist app, where it put you in contact with other app users when are were close by each other, and asks you to perform some silly task. Which sounds like a great idea but is a bit daunting when it comes down to it. Meeting complete strangers is weird. I certainly never had the courage to actually meet somebody that way. What on earth will you talk about?
Doweet is yet another service that does much the same, but seems to be more focussed on your existing friends, thereby solving the above problem.
And approaching the problem of people discovery from a slightly different slant is Grouper. They organise a meet between two groups (3 girls and 3 guys) of friends (via Facebook) that don’t know each other. Interesting.
Are any of these apps games changers in their current incarnation? Nope. Will they be big business? Dating is not THAT a huge. But it gives you an indication of where things are going. Somewhere somebody is going to hit on a really sexy app.
Posted by Wessel van Rensburg