Monday, August 24, 2009

Twitter in the Land of Power Laws

This is my first post after a long summer holiday and I am glad to say that my blog has still been receiving a reasonable number of visits in the absence of new content, albeit some visits were barely more than a glance.

Gartner has released a collection of Hype Cycles for various technology niches, as reported by Eric Auchard at Reuters for example. The particular Hype Cycle below is for Emerging technologies (double-click to enlarge).


I was immediately surprised to see Quantum Computing registering in the Technology Trigger region, which is used to denote technologies that are 10 years or more out from acceptance. I think this is something of an understatement as I argued here.

Auchard picked up on microblogging (read Twitter) being positioned near the peak of expectations, and therefore about to experience the full G-force of descending into the Trough of Disillusionment. I am not really sure from what perspective Gartner is making this prediction, since a recent study released Sysmos shows that user growth has more than doubled this year – in fact, over 70% of Twitter users joined in 2009.


The report shows that almost every aspect of Twitter is operating under a power law. In the case of new users this power law means exponential growth, whereas for other measures the power law typically means the domination by the few. For example, 92% of people follow less than 100 other people, but 1% of people follow more than a 1000 others. Less than 1% of people have more than a 1000 followers, and more than 90% have less than 100. Interestingly, 21% of people with a registered Twitter account have never made a tweet. Most people make only one Tweet per day but just over 1% make at least 10 on average. More generally, Sysmos observed that there is a 75/5-rule in operation, meaning that 75% of activity is accounted for by 5% of the Twitter user base.


Is Gartner right then? Perhaps in the words of Neil Postman the majority of Twitter users are just “amusing themselves to death”, while a few Twitter users are really tweeting themselves to death. Where is the business case here? But as I argued in The Sub-Timed Crisis in Web 2.0, we are not heading for collapse:

Unlike our current financial structures, the web is not at threat of collapsing, though many foot soldiers will fall by the way (they see themselves as pioneers, but in fact they are easily replaced). Our informational structures are not hierarchical but relational, and as such, are much more resilient to the removal of individuals. It is not the case that there are eager underlings waiting to replace leaders – the underlings are here and functioning already.

Web 2.0 losses will largely go unnoticed. New users/readers, whose information experience begins today, are being added all constantly. They are essentially unaware of what happened last month and will remain that way. Joining is not a generational wait, no corporate ladder to be climbed. Everyone joins and progresses simultaneously. This turnover goes largely unnoticed since leavers are dwarfed by joiners.

I am quite sure that Twitter will weather the Trough of Disillusionment – in fact, it already has I would say. Gartner is as overly optimistic of Quantum Computing as it is pessimistic of Microblogging.

Related Posts

1 comment:

Anonymous said...

Twitter is the most trendy social networking site.

Laby[big and tall suits]