Estimating Android tablet market share: 30%, 40%, 10%?

Consider the following: 

  • According to the platform statistics that Google provides for developers, 5.8% of Android devices connecting to the Android Market (now ‘Google Play’) have screens that are 7 inches or larger, and a further 2.5% are 5-7 inches. 
  • Google disclosed 300m cumulative device activations in February. At the current run rate, that would be 325-330m now. 
  • Almost all of those sales are within the last 24 months, so there’s very little replacement in there - almost all those devices should be in use

Hence, it is possible to say that 8.3% of 330m devices are tablets - so somewhere between 25m and 30m Android tablets have been sold. 

Meanwhile, Apple has sold 67m iPads since launch, and again almost all of these are in use. That would mean that Android tablets are almost a third of the install base. Indeed, since Android sales only really started in the last 4 quarters and Apple sold 47.6m iPads in that period, run-rate share might be closer to 40%. 

But, but, but…

I’ve hardly seen ANY. Even at MWC in Barcelona, I saw perhaps 3 Android tablets being used by delegates as against several dozen iPads. The Galaxy Touch Note has sold well, but that’s part of the 5-7 inch category and certainly hasn’t sold double digit millions. 

At this point, it’s worth looking at the share of different versions of Android. While ‘7 inch plus’ has 5.8% share, Honeycomb (the version of Android designed for tablets that is now being superseded by ICS) only has 3.2% share. There are hardly any ICS tablets out there. So the implication is that almost half of the ‘iPad-sized’ tablets out there are models that are running Android 2.3 and, by implication, being sold, probably cheap, in China and emerging markets, as this story suggests. These are not really iPad competitors. 

So, if we take only the Honeycomb segment, Android tablet sales fall to ~10m (3.2% of 330m) - 8% share of tablet sales to date. Somehow that seems more reasonable than 30%.

However, there is a further consideration here: the representativeness of the ‘devices connecting to Android Market’ that are the basis of the Google platform stats. Not all Android devices do connect to Android Market, though no-one quite knows how many - Google itself knows how many do connect but not how many don’t. 

In other words, do Android tablets make up the same proportion of devices sold as they do of devices connecting to the Android Market? It is quite possible that Android tablets are more likely to be bought by technically-minded people and therefore more likely to connect to Android market - OR that since there are no Android tablet apps, Android tablet owners don’t bother with Android Market at all and so are under-represented in Googles’s stats. In this question is the potential to swing the sales figure estimate by double digit percentages with no difficulty at all.

Then, of course, there are the Nooks and Kindle Fires, which can run Android apps but which don’t show up in Google’s activation figures at all and (unless an enterprising hacker has side-loaded it) don’t have Android Market either. Sales of these are in the millions, but aren’t publicly disclosed.  

As Groucho Marx almost said - these are my estimates, and if you don’t like them I’ve got others.