Thoughts on Google’s Android version charts

Google regularly updates the data it provides to developers on Android versions in use, screen sizes and screen densities, and I’ve been diving into this data today for a report I’m working on. In the next few days, Google will update the data again and there will no doubt be the usual flurry of blog posts and news items about Android fragmentation. But I wanted to share some thoughts that occurred to me as I looked through this data that go beyond the usual rhetoric. Some of these are original, some of them likely aren’t.

Firstly, a fairly predictable pattern has emerged in the adoption of the versions of Android, as follows (and as illustrated by the chart below) 1:

Android Major Version Distribution Continue reading


  1. It’s worth noting that the methodology Google uses for all these numbers changed in April 2013, and now reflects only devices actively using the Google Play store, and not all devices as previously. From what I can tell, it hasn’t made an enormous difference, but has slightly increased the representation of newer versions while decreasing or eliminating the representation of older versions.

Why Sprint – T-Mobile makes sense

There were rumors today – not for the first time – that Sprint might be interested in making a bid for T-Mobile. This is not all that surprising given recent remarks from T-Mobile execs and Dan Hesse that they would be open to a merger. But there’s been a predictable outcry about the possibility of the US’s three major carriers being whittled down to two, and especially about the presumed loss of T-Mobile’s recent disruptive approach to the industry.

There are several good reasons to take this view:

  • T-Mobile has indeed been disruptive, and has caused real change in the industry. Its shift away from 2-year contracts and towards easier, more frequent upgrades sparked the other major carriers to follow suit. It has won subscribers from Sprint and AT&T in particular as a result.
  • There’s an instinctive reaction to a reduction in the number of players in any industry, and it would follow years of consolidation in the US wireless market. It’s easy to argue that a market dominated by three players would be less competitive than one with four major players.
  • The US has a huge population, and it seems like it ought to be able to support four or more players without too many problems, given that there are other markets around the world with more players and much smaller populations.
  • The two carriers use incompatible network technologies. After Sprint has worked so hard to eliminate iDEN and WiMAX and focus on its core CDMA, EVDO and LTE networks, it would be a shame to complicate things by adding T-Mobile’s GSM-based networks to the mix. Given the focus on LTE this might be less complicated than it once was, but it’s still a non-trivial issue.

However, I think this knee-jerk reaction opposing any consolidation among the big four may be misguided, and here are the reasons. Continue reading

What John Chen needs to tell customers (and investors)

Earlier this week, BlackBerry’s CEO John Chen posted a letter to customers. While he provided some sense of his strategy going forward, he unfortunately continues the tradition started by his predecessors of failing to answer the most compelling questions customers (and investors) have:

Is there any reason to believe the atrocious trend in device sales will turn around? If so, what?

Device shipments have now dropped 75% from their peak in 2011, and although it’s possible we’ll see a small blip next quarter from steep discounts on the Z10, the trend is likely to continue downward. Chen needs to explain what, if anything, will cause these same poorly-selling devices to start selling better, or allow any future devices to be more appealing to users. As of right now, there’s no evidence of either of those things, and as such we have to assume shipments will continue dropping, and with them what has historically been the largest chunk of overall revenues.

In the absence of that, is there any reason to believe service revenues won’t follow suit very soon?

Service revenues make up most of the rest of BlackBerry’s overall revenues, which is why some people seem to think it’s the most promising avenue for BlackBerry going forward. But the reality is that these service revenues are directly tied to the installed base of BlackBerry devices, each one of which generates a few dollars every month for the company. But, if device shipments go down dramatically and existing BlackBerry users churn to other platforms, this service revenue will merely lag falling hardware revenues by a few quarters but generally follow the same path. BlackBerry has already stopped reporting subscriber numbers, which started falling late last year, and had dropped from 80 million to 72 million by the time BlackBerry closed the door on that metric.

Continue reading