Social Media Proliferation - 5 things you really need to know.

Today every new (and old) web service's must-have is a social network (and being advertising funded of course), and this post on GigaOm called Are Social Networks Just a Feature has spurred us to offer some basic lessons from our emerging experience.

(Related trends are the emerging hype around Digital Lifestyle Aggregators, Trust, ID and, of course, Advertising via Social Media.)

These are for comment and debate, as this is an emerging science (if thats what it is):

1. There is a limit to the number of social networks a person will (willingly) join, and the bad news is that this obeys power laws - i.e. it is hit based - in other words a small number of sites will garner most of the links, and the rest will be much, much smaller. To make it worse, the evidence is that sites go in and out of fashion as well, as this article (also from GigaOm) on Facebook implies

2. For a social network to work, it needs to be....social - small social nets fizzle out, there is not enough to keep people attracted. The Long Tail benefit is limited. There must be enough people to believe it is necessary for that service to exist. Consider Technorati's new Where's The Fire (WTF) digg type social net - despite the huge publicity, the popularity of Technorati (c 200th most visited website globally), and its traffic (a reach of c 3.4m people per day), the WTF social network itself has been fairly slow to take off (c 120 topics with more than 1 post (c 500 posts to date), c 100 topics with 1 post only - and many from the same people and quite a few from staffers/friends before the service was launched 5 days ago (1).

3. Registering with a social network, setting up the user profile etc adds friction to a website, as much (if not more) as the old "Web 1.0" approach of requiring a subscription. In the early days, for the early adopters, such advances tend to be a novelty - but this novelty recedes as it becomes just another mainstream service, and it has to offer a net benefit, as it were. This applies to Digital Lifestyle Aggregators as well - replacing a profusion of Social Nets with a profusion of Social MetaNets is just not an option.

4. A Social Network is not a passport to advertising effectiveness. Just because people trust social networks by nature does not mean they will trust your social network, especially if trust is eroded by clumsy or overenthusiastic attempts to advertise to them. (In addition we would suggest that social nets are "personalised", like mobiles, and all the indication on mobiles is that users resent unsolicited ads far more than web users, the distaste seems to be almost on a par with that for pop-ups)

5. A social net is not the answer to all services, and in fact may be outplayed by database oriented services. Consider alternatives, like Pandora's service which is both analytical and social net based as opposed to Last fm's which is more purely social net based. My personal experience of both of these is that Pandora got me to a more complete % of music I liked faster. In a more competitive future, these sorts of results will start to matter more.

(1) This is not to say it will not take off, just that even given its advantages a social net benefit is not guaranteed. I am curious to see whether a large enough number of topics need to be "seeded" first before it takes off.

Postscript - quite a nice summary of Taxonomy by Liz Gannes at GigaOm here, mapping to an earlier piece of work we did here