Lies, Damned Lies, and Statistics

I was looking over some app store metrics tonight, in particular the breakdown of apps by category, and wondered if you could look at the breakdown of the top 500 paid apps by category and determine which categories were over or under served. The results were pretty interesting.

First, I started with this category breakdown over at 148Apps. The breakdown looks something like this.

Category Percentage
Games 16.7%
Books 11.8%
Entertainment 10.4%
Education 9.3%
Lifestyle 7.8%
Utilities 6.0%
Travel 5.4%
Music 4.0%
Business 4.0%
Reference 3.7%
Sports 3.1%
News 2.7%
Productivity 2.6%
Healthcare & Fitness 2.4%
Photography 2.0%
Finance 2.0%
Navigation 1.9%
Social Networking 1.8%
Medical 1.8%
Weather 0.4%

Of course this is a simplification, but if supply and demand were perfectly balanced, we would see the same breakdown by category in the top 500 paid apps. Instead it breaks down like this:

Category Percentage Abs Diff Rel Diff
Games 40.2% +23.5% +140.3%
Entertainment 9.2% -1.2% -11.4%
Utilities 7.8% +1.8% +29.7%
Photography 6.8% +4.8% +234.4%
Productivity 5.2% +2.6% +101.6%
Lifestyle 4.6% -3.2% -41.2%
Healthcare & Fitness 4.2% +1.8% +74.1%
Social Networking 4.2% +2.4% +130.3%
Music 3.8% -0.2% -5.9%
Education 2.2% -7.1% -76.4%
Business 2.0% -2.0% -50.3%
Navigation 1.6% -0.3% -14.3%
Weather 1.4% +1.0% +235.2%
Reference 1.2% -2.5% -67.3%
Sports 1.2% -1.9% -61.8%
News 1.2% -1.5% -55.8%
Books 1.0% -10.8% -91.5%
Travel 1.0% -4.4% -81.6%
Finance 0.8% -1.2% -59.2%
Medical 0.4% -1.4% -77.5%

Games dominate the makeup of the app store submissions, so you would expect it would occupy a large percentage of the top 500 apps, but I wouldn’t have expected it to be so out of balance. Before really thinking about it, I wouldn’t have put much thought into the game category because on the surface it seems that category is over saturated. But with games representing almost 1/2 of the top 500 and only 1/6th of the submissions, it may not be unreasonable to expect to crack the top 500 with a game. What about some of the other categories? Sorting by relative difference between overall and top 500 representation might make this easier to see.

Category Percentage Abs Diff Rel Diff
Weather 1.4% +1.0% +235.2%
Photography 6.8% +4.8% +234.4%
Games 40.2% +23.5% +140.3%
Social Networking 4.2% +2.4% +130.3%
Productivity 5.2% +2.6% +101.6%
Healthcare & Fitness 4.2% +1.8% +74.1%
Utilities 7.8% +1.8% +29.7%
Music 3.8% -0.2% -5.9%
Entertainment 9.2% -1.2% -11.4%
Navigation 1.6% -0.3% -14.3%
Lifestyle 4.6% -3.2% -41.2%
Business 2.0% -2.0% -50.3%
News 1.2% -1.5% -55.8%
Finance 0.8% -1.2% -59.2%
Sports 1.2% -1.9% -61.8%
Reference 1.2% -2.5% -67.3%
Education 2.2% -7.1% -76.4%
Medical 0.4% -1.4% -77.5%
Travel 1.0% -4.4% -81.6%
Books 1.0% -10.8% -91.5%

Weather and Photography apps make up 0.4% and 2.6% of apps submitted to the app store, but they represent 1.4% and 6.8% of the top 500 apps. That’s pretty huge. iPhone users love them some weather and Photography apps! Social Networking, Productivity and Healthcare & Fitness also seem to be under represented in the app store.

At the other end of the spectrum are Books. Books make up 11.8% of app store submissions, but only 1.0% of the top 500 fall in that category. The app store is pretty well set on Books. Seriously, no more. Medical is 1.8% of submissions, but is the teeniest sliver of the top 500 at 0.4%. Other categories are over represented are Travel, Education, Reference, and Sports. Turns out the overlap between sports enthusiasts and the typical geeky iPhone user is a bit smallish. Who’d’ve thunk it!?

So the other thing I was curious to know was: suppose all apps are created equally (they’re not, I know, but just for the sake of curiosity let’s suppose they are) and you submitted an app under each category. Using these ratios, what are the chances that your app would crack the top 500? (we’re going to assume that the 41% of free apps is constant across all categories, so we won’t be competing with those.)

Category Odds
Weather 1 in 121
Photography 1 in 121
Games 1 in 169
Social Networking 1 in 176
Productivity 1 in 201
Healthcare & Fitness 1 in 233
Utilities 1 in 312
Music 1 in 430
Entertainment 1 in 458
Navigation 1 in 473
Lifestyle 1 in 689
Business 1 in 816
News 1 in 918
Finance 1 in 993
Sports 1 in 1,060
Reference 1 in 1,240
Education 1 in 1,719
Medical 1 in 1,803
Travel 1 in 2,207
Books 1 in 4,779

If you’re looking to crack the top 100, it’s not just 5 times less likely. The category breakdown is slightly different here.

Category Odds
Photography 1 in 589
Games 1 in 652
Weather 1 in 846
Productivity 1 in 871
Utilities 1 in 1,356
Music 1 in 1,363
Healthcare & Fitness 1 in 1,629
Navigation 1 in 1,890
Entertainment 1 in 3,006
Social Networking 1 in 3,695
Lifestyle 1 in 3,963
News 1 in 5,505
Sports 1 in 6,361
Reference 1 in 7,442
Business 1 in 18,911
Finance 1 in ∞
Education 1 in ∞
Medical 1 in ∞
Travel 1 in ∞
Books 1 in ∞

Notice, for example, how Social Networking is 4th easiest to break into the top 500, but drops down to 10th position when it comes to the top 100. Social Networking is popular, but not so much top 100 popular

Anyway, I realize that I’ve vastly oversimplified this analysis, but I think some of the generalizations are probably accurate. And besides, I was mostly just having fun playing around with numbers and thought I’d share. Maybe someone else will find them interesting as well.