NEW SOCIAL MEDIA PLATFORM KLAXON.
Well, not really new. Anyone with an eye on the right sites will know about Pinterest, which has been knocking around since mid-2010. However, it stuck itself seewhatIdidthere into the market late last year. And now - well, now it has 10 million unique users, apparently the quickest ever site to do so.
How did it manage this? In my opinion, because it had superb timing. Facebook Timeline, and the recent deluge of apps and API-created edges that are based around 'self-expression', as Pinterest calls it, are getting a certain group of people pretty used to frictionless hyper-sharing. It's no small group, either. However, none of these apps had really been quite so effective in the way that Pinterest is. Half because they were either focusing on other or too many things, and half because Pinterest really nailed the central conceit, right down to the metaphysical name.
You'll find plenty of people to talk about how great Pinterest is, but what's initially fascinating is its ability to audience profile. Take a look at this:
Image credit: Techcrunch
That's an awfully specific geo-demographic in areas not traditionally considered tech hubs. It's also well-documented that the biggest adopters are women aged 25-44. Pinterest can take well-defined audiences like this and package them whole to advertisers, who in turn can reach those people via channels they may often be blind to, and with all the opportunities that go with digital-specific delivery. That alone wouldn't make it special, though, nor even would the viral element that clever brands may be able to tap in to (depending on what products Pinterest is developing).
No, what I really, really like is the underlying associative patterns. Pinterest might call it self-expression, I (and many others) call it an interest graph. ReadWriteWeb had an interesting article that captured it quite well:
A social graph is a digital map that says, "This is who I know." It may reflect people who the user knows in various ways: as family members, work colleagues, peers met at a conference, high school classmates, fellow cycling club members, friend of a friend, etc. Users send reciprocal invites to those they know, in order to map out and maintain their social ties.
An interest graph is a digital map that says, "This is what I like." As Twitter's CEO has remarked, if you see that I follow the San Francisco Giants on Twitter, that doesn't tell you if I know the team's players, but it does tell you a lot about my interest in baseball.
There are plenty of services that can tell you what their customers are interested in on their own site, such as Amazon. What is more difficult is to deliver that at scale and beyond the confines of your own digital properties. Ad networks do it by embedding cookies, which is good but imperfect. As for social networks, it's a great unfulfilled promise. Facebook, for example, can tell you oodles about the social graph, but interests are rocky at best. Practices such as incentivized ads rapidly make user pools homogenous as everyone Likes the same things in exchange for Facebook Credits. Add to this the problems inherent in categorizing fan-made likes, having certain popular items making rankings top-heavy, and the throwaway nature of Likes in general, and soon it can become difficult to filter interests in a meaningful way. Overindexing can help, but it's limited.
However, the Open Graph (more graphs!) may change that, as apps gain the ability to read other public app activity. One would start with Social Graph data on your customer base, gathered through Facebook Connect, and then twin it with Interest Graph data from an app that itself connects to Facebook. And there's nothing to say that Pinterest will be the only one worth turning to: if you were a travel company, you could access public content from all those 'where I've been' apps. The self-curation element of 'Interest Apps', coupled with the right level of specificity, trumps Likes any day. Essentially, Facebook becomes an open information resource, fuelled by APIs and semi-automated profiling.
The question remains, of course, so what? Well, the promise of the interest graph - which is still yet to be delivered to a market that is more than ready for it - is being able to pre-emptively ascertain associative interests without any testing whatsoever. In other words, you know that x indicates a strong preference for y and z. Sure, stuff like this exists in research agencies already, but not particularly granulated and certainly not automated. It's also siloed: analysts might pick through dozens of, say, holiday destinations, but the framework just doesn't exist to link that information to information about, for example, clothing. It's all inferrance and I bet it misses a hell of a lot of detail.
Couple that sort of knowledge with the Social Graph, which helps you figure out interrelationships, attribution and influences, and suddenly you're looking at something that lets you pick any individual demographic profile and flesh it out in any direction, even the counterintuitive or improbable. This combined understanding could then be used to define media homes, comms strategies and niche demographics. And when it came to delivery, cross-referencing this knowledge with Social Graph connections and live multivariate testing would allow you to deliver content for maximum footprint across both digital and traditional channels.
Interesting.