Mobile Apps and the Power Law

Ok, enough is enough. People have foolishly ignored the importance of power laws for too long; and I will strike them down with great vengeance and furious anger for doing so.

Flashback to February, when Pinch Media's app store study began to proliferate. Back then I argued, and Pinch Media's analyst Josh Reich agreed, that speaking of averages was neither very interesting nor very informative because app usage no doubt fell into a power law distribution.

The shorthand for the power law is the 80/20 rule. 80% of blog traffic goes to 20% of blogs, 80% of links go to 20% of websites, and so forth. The power law is ever-present in all social arrangements. This fact is incompletely appreciated by most of the analysts who ought to be paying attention, to say the least.

My original criticism of the Pinch Media study has been on my mind lately, as many are concerned over whether the Palm Pre could possibly catch up with some 50,000 iPhone apps. This should seem less insurmountable when you take into account the fact that 20% of the apps very likely account for 80% of the downloads, and 20% of those downloaded account for 80% of those still being used regularly a month later. In fact, Josh Reich said that the data Pinch Media was looking at suggested it was more like 1% of apps accounting for 99% of the given activity.

Not only does that mean that newcomers like the Pre need only come up with a fraction of the iPhone's apps in order to be competitive, the mere existence of the iPhone provides a helpful set of examples of precisely what kinds of apps have already proven successful. If there is something exciting about the newcomer, something that sets it apart from the iPhone in a way that entices developers, it should not take long before the number of apps relevant to the user are on par and the real competition can begin.

There are already signs that this occurs in practice, though again they are misinterpreted because people just can't bring the power law into the picture.

A new study is making the rounds that indicates that Android apps on average have a higher retention rate than iPhone apps. Of course, there are significantly fewer apps for Android than there are for the iPhone. One theory advanced is that this means that iPhone users flit from app to app because there are so many of them coming out all the time.

Maybe. I think it's more likely that Android, a younger platform, saw much of its initial development taking the form of applications similar to the handful of the most successful ones on the iPhone.

I'm no expert on this subject by any means, and am mostly relying on theory and other people's empirical work. But I think that ignoring the power law when attempting to analyze any social phenomenon is a mistake--if you goal is to understand it.

UPDATE: The mobile apps analytics firm Medialets announced that there had been over 700,000 app downloads off the Palm Pre. This even though there currently are only 30 apps available. Hat tip: Engadget

This work is licensed under a Creative Commons Attribution 3.0 United States License.