I’ve come to the depressing conclusion that I can’t trust comScore’s share numbers any more.
The problem isn’t that they’re showing a second month of effectively no change in userbase share for Android. That, in isolation, I could manage to believe (though I would find it very puzzling). The problem is that I can’t reconcile that with other lines of evidence.
One of my commenters has pointed out that Android activation numbers are up to 900K a day. That’s 27 million devices a month, and I can’t imagine any plausible percentage of tablet activations to discount that by that would leave Android smartphone activations unable to swing the needle in a market that comScore estimates at just 110M users.
Other market-research outfits were already quoting Android global market-share figures ranging from 56% to 60% two months ago. Yes, global market share isn’t the same as U.S. share, but historically those trendlines have been distinguished by timelag rather than having different slopes. And just to drive that point home, NPD already had Android at 61% of U.S. share in May.
Finally, there’s the curious fact that when you multiply out comScore’s own numbers it looks like Android is still gaining smartphone users a little faster than iOS (though not by a statistically significant amount.)
I don’t know what’s going on here. I trusted comScore’s numbers for a long time because they showed a continuous and regular set of measurements that was in sync with more sporadic indications from other sources. Overall it made a coherent picture. Now the other sources are still suggesting rapid Android growth, but comScore thinks it ain’t happening.
This is a bummer. It means I’m going to have to be a lot more skeptical of comScore’s numbers in the future even supposing they turn happy for Android again.