Playing with Ratios

futurelab default header

Fred Wilson writes an interesting post about what he calls the common ‘web/mobile laws of physics’, making the observation that many of the tech services and companies they see as VCs exhibit similar ratios between the number of registered user or downloads (if it’s a mobile app) the service has, the number of monthly and daily active users, and the maximum number of concurrent users (for those with a real time component). The 30/10/10 ratio runs:

  • 30% of the registered users or number of downloads will use the service each month
  • 10% of the registered users or number of downloads will use the service each day
  • The maximum number of concurrent users of a real-time service will be 10% of the number of daily users

Fred goes on to talk about how often these ratios are incorrectly interpreted as poor performance, and how smart tech services and apps change these ratios more in their favour through judicious use of email and mobile notifications (such as Facebook photo tagging or Instagram ‘Like’ notifications) which can increase visit frequency and materially increase the proportion of registered users that use the service daily or monthly. It’s possible to turn off these notifications on many services of-course but from personal experience, it’s notable that whilst I’ve made efforts to avoid notification overload, I haven’t turned them off on every service. There are some things it seems, that I want to be notified about.

Of-course, perhaps the best known ratio of this type is the 90-9-1 principle of participation inequality amongst online communities – the idea that (accepting slight variations due to subject matter) such communities typically have somewhere around 1% of people who create content (creators), 9% who edit or modify that content (contributors), and 90% who view the content without contributing (lurkers).

At last weeks Wikipedia conference, Jimmy Wales said he was concerned at the steady decline in the number of contributors to the site (cue sensationalist Gawker headlines about Wikipedia ‘slowly dying’). With almost 90,000 active Wikimedia editors and 381 million unique users to Wikimedia projects the participation inequality ratio is (perhaps inevitably given the utility of the content) skewed lower, though in fairness the definition of an ‘active’ editor used is one that demands 5+ edits a month so the real number of editors is likely to be a lot higher. Still, the total number of active editors has gone down by 2.6% y-o-y.

There may be many reasons for this. Wales himself has suggested that the profile of editors is slowly changing as the priorities of the typical longstanding contributor change over time and that, given how established the site now is, there are simply fewer new entries to add. But it is the decline in the number of new editors per month (at 13% y-o-y) that is notable, which is perhaps what Wales is referring to when he says that they are not replenishing their ranks (‘It is not a crisis’ he says, ‘but I consider it to be important’). So Wikimedia have set about addressing the problem through simplification of editorial guidelines which have become convoluted over time, providing new tools to the community, and partnerships with Universities.

I don’t think for a minute that Wikipedia is slowly dying. But it’s a reminder of a few things: that unique users are not created equal; of just how important a relatively small proportion of users can be to any company, service or application that relies on the regular interaction, participation and contribution of its users; and that, as indicators of the health of such services, ratios of these kinds matter.

Original Post: http://neilperkin.typepad.com/only_dead_fish/2011/08/playing-with-ratios.html