In another sign that Twitter is crossing the chasm, or at least getting close to it, the FT today had a full page analysis on the company and its prospects under the heading Sweet to tweet, which made me think again about how the business will scale.
That thought process took me back to the thinking of Stephen Johnson as expressed in his fine book Emergence. I’ve written before about his theories applied to social communities. The following is an excerpt from that post:
per Emergence the success of communities is determined by:
- The number of members in the community
- The rules that govern the behaviour of members
- The feedback loops which re-enforce certain behaviours and filter out others
Johnson does a case study on Slashdot which illustrates this well:
- Size – when the community was small everyone could read all the posts, but when it became larger the number of posts became too large and sophisticated rules and rating systems were required to keep the site usable – the optimum rule set changed as the community grew
- Rules – the introduction of a 1-5 rating system, vertical subdomains, the formation of a cadre of people with sufficient status to review and critically the emergence of an unwritten rule or community norm that “slashdot status” is desirable – the evolution of this sophisticated rule set was critical to the success of Slashdot
- Feedback – the rating system they devised has a complex feedback process which encourages quality submissions and fosters the emergence of an elite group of moderators and to maintain a position in that elite your contributions to Slashdot have to be highly rated by the community generally – the ruleset incorporated sophisticated feedback mechanisms
As another example of this remember how Facebook has continually changed the rules for applications in order to maintain the right balance between spam and viral growth.
All this is interesting for Twitter because unlike Slashdot and Facebook it doesn’t control the user interface and therefore can’t change the rules which govern the service in the way that they can. Instead it is up to the Twitter clients (Tweetdeck, Twhirl etc.) to develop tools that get users over the problems that will inevitably emerge with scale. (The first of these problems was keeping track of DMs and @replies as the volume of general tweets grew which Tweetdeck solves by capturing them in separate columns. I suspect the next problem will be keeping track of conversations.)
This trick of outsourcing the evolution of the rules for Twitter to third party developers is at once clever and risky. Clever because it absolves Twitter from the need to impose changes on the community which might backfire and instead allows them to benefit from the successful experiments of others without suffering if they fail. Risky because in some sense the rules are the community and outsourcing them to third parties means you are depending on them getting it right (or at least one of them) and also because you risk the community becoming more resident in the third party client than in Twitter itself (remember Photobucket and Myspace).