App Store Growth Curves & Forecasts: webOS vs. Android

Dan Ramirez (@vara411) and I have been monitoring the webOS App Catalog and recording app counts, and we now have data clear back to its launch.  Dan just posted a terrific write-up on the overall data, along with important events over webOS' short life so far, on his blog: Totally Palmed.  He also compares it to Android's App Market over its 15-month existence and comes to some interesting conclusions. (Go read that first and then come back and finish this.)

But, being the statistics and operations guy I am, I thought it'd be interesting to see what happens if we forecast these app counts out into the future.  So, using our data, I did just that.

First, I plotted the data versus months after each platform's app store launch.  The orange are all Palm webOS and the green are, obviously, Android. Note the two different y-axis units in the graph:

Then, using Excel's quick-and-dirty trendline tool (I just didn't feel like breaking out SAS for this), I looked through several forecast models and settled on second-order polynomials as the best-fitting solutions (per the R-squared values).

Using those trend formulas, I projected out the app count for each platform and got the pair of graphs below.

First is the graph that shows how each platform's app count grows relative to the calendar.  Since Android got an 8-month head start on webOS, it obviously gets bigger sooner. The projected data (estimates) are shown in alternate colors: dark green for Android and light orange for webOS.


Then, to look at this in a more apples-to-apples fashion, I removed the time lag for webOS and overlaid the two app counts on each other.  This is the same approach as the first two graphs above and shows total counts versus months after the launch of that platform's app store.

It looks pretty clear that both platforms are seeing exponential growth, which is a really good sign of healthy uptake by developers.  Also, it seems that Palm webOS will likely have 1/3rd to 1/4th the number of apps, at the equivalent time into its existence, that Android will.  And that should be fine, as any platform with 20,000+ apps (and a healthy growth rate) should be well-positioned for maintaining its position in the market. Android should pass that bar any day now, and webOS should, if these numbers are reasonably accurate, hit 20K apps before the end of next year.  Recent announcements that webOS will come to Verizon and AT&T in the first half of 2010 may accelerate webOS's app count considerably.

Granted, these charts are based on a variety of assumptions and should not be considered anything more than an interesting thought-experiment with pictures.  Google or Palm could come along with a new development tool or device or innovative program that greatly excites the dev community.  Conversely, either could screw up on something and drive them away.  Only time will tell, but it'll be fun to watch.