The latest controversy in the world of mobile devices is the market share of Android tablets; in particular, when analysts claim that Android has, say, 26% of the tablet market share, are they talking about sales to real live people, or shipments to the channel?
One technique for getting a very vague idea is to take the total number of Android activations (190 million) and multiply it by Google's version share data for the tablet OS, Android 3.x (1.8%), giving 3.4 million total Android 3.x tablets (as compared to about 40 million iPads sold thus far). Clearly less than 26%. I used this technique before it was cool.
However, it should be cautioned that it is a very, very vague technique. There are a few problems. First of all, the Google version share numbers are based on Android devices which have been in accessed the market in the last two weeks. Does this mean all devices in use and in communion with the Android market, or does it mean all devices where the user has hit the marketplace button in the last two weeks? If the latter, it might be the case that there are lots of Android tablets, but their owners don't bother much with the marketplace.
Second, what is an activation? For instance, if a phone is sold to person A, returned, and sold to person B, is that one activation, or two? If person B then buys a new phone, is that another activation? That is, does the figure relate to a specific device, or to a specific Google account, or to a specific device with a specific user account?
Then, even if there have been 190 million Android devices 'activated', undoubtedly many are now out of use. Say there are 100 million now in use; that's only 1.8 million Android 3.x devices now in use.
And finally, is it possible that Google's reported figures are just wrong? A cap of 3.4 million 3.x tablets sold seems bizarrely low, as compared to shipment numbers of 4.6 million in the most recent quarter alone. It should be borne in mind that these numbers include Android 2.x tablets, and the Barnes & Noble Nook, which, while it can be used as a tablet when rooted, is really intended to be used as an ebook reader, of course.
Anyway, I mention this because of this GigaOm article, by the same author who cited me a while ago on the above technique. I think he's relying on it too much; he seems to almost be stating as a matter of fact that there are 3.4m 3.x tablets, and further muddying the waters by comparing that estimated lifetime sales number to a claimed Q3 shipment number. All a bit messy, I fear.