Real world enterprise social networks; why quality trumps quantity

I think the uptake of social media within the enterprise is really interesting. There are few independent studies (and lots of sponsored ones) so it was interesting to see the adoption of an experimental, completely unsupported and unpublicised enterprise Yammer implementation – it would be inappropriate to say when, where or how.

What happened was that following an initial burst of activity where the recruitment of new users went viral, it then went silent for about 9 months. There were very few messages, very few new users. Then the activity picked up again of its own accord; users would put up more messages and the join activity started to rapidly increase.

I must admit that having read about the Twitter new user ‘9 month bounce’ last year I was expecting it – basically, users go quiet after joining and then come back. This resonated because it’s exactly what I did with Twitter.

Of course, Twitter’s new user activity is obscured by the continuous near vertical (but linear, apparently) user number growth. In a limited environment any slow down in growth is much easier to see.

So why the dip? Why the pickup?

I think the dip was down to a few things;

1. Public nature of posting – people were a bit shy/wary.
2. Users unsure what to post, or why.
3. Insufficient followers – you post more when you have followers. Probably it's something to do with 9mths being the amount of time it takes to get enough followers as a new user in order for it to be worth tweeting.
4. Growth by spam; it was pointed out that on joining the platform spammed you to invite new users. These users might join but weren’t necessarily engaged.

Why the pickup?

Just my thoughts but:

1. It took 9 months to find the right users; a couple of users popped up out of nowhere and began posting on a regular basis, making the community active and therefore:
1. Useful
2. Accepted behaviour (cf.Gladwell’s ‘The Tipping Point’, particularly sections about ‘permission’.)
2. Personal social media penetration; users are relaxing about posting stuff. Without doubt enterprise users will get burned by posting inappropriate comments – and learn from it.
3. Viable network size; one of the reasons that Yammer wants to drive network growth (see 4 above) is that they’re aware that a big network is more likely to be robust and active, hence useful, hence endorsed by the enterprise.

Now, fascinatingly the relationship between the size of a network in the early stages of development and it’s activity was almost (could be?) mathematical: they correlated precisely. It was almost as if every single new post added a user, even though that user wasn’t addressed.

In fact, the number of messages was far below what we might expect; perhaps the quality of the network and network activity didn’t match the growth; perhaps quality drives quality, as growth drives growth.

The take away for me is that to build a quality network is more difficult and rewarding building a big one.