Will Android Overtake Palm & Windows Phones Soon? Silly Linear Assumptions or Forecast?

If you’ve visited this blog before, you may have caught some of the fun I have when various statistics (or what passes for that) and predictions are made in the mobile market space. Here’s the latest one that caught my attention…

ReadWriteWeb: Android Market Share Doubles – Will Overtake Palm Soon

First, where does this data come from? It came from this comScore press release: comScore Reports December 2009 U.S. Mobile Subscriber Market Share. Next, how reliable is this information? How much validity does it have? That is difficult to say without a much deeper analysis of comScore’s sampling methodology and assumptions. Third, does comScore itself make any trend analysis on their “share (%) of smartphone devices” results? No, they do not.

So, where does the prediction that Google Android will overtake Palm’s marketshare (Palm OS + webOS?)? I believe it comes from an assumption that we will see linear market growth or loss carry through the current quarter (Q1 2010). That would indeed have Google Android overtake Palm by the end of March. Using this predictive model, let’s take it to an illogical extreme and take the numbers all the way through the end of 2010. You can see resulting graph above. If we assume a steady linear growth or loss for this year, Google will also overtake Microsoft’s Windows Phone before the end of the year (and that may happen). Palm, on the other hand, completely disappears by September.

Do I actually believe that my chart predicts anything? Not really. Palm’s introduction to the Verizon Wireless customer base may pump up its market share a bit. Android has a lot of mindshare right now (expecially following last year’s aggressive Verizon Droid campaign) and the recent Nexus One hoopla. But, that hasn’t translated into iPhone like sales yet. Google’s and Palm’s extremely small existing base makes small fluctuations seems a lot larger than they probably are. Their relative growth and loss are probably much flatter than what you see in my graph based on oversimplified assumptions.