Welcome! If you're new here, you might want to check out the best posts on GAMESbrief or subscribe to find out how the games industry is changing.

The average iPhone app sells 25,000 units

At the IBC legal conference I attended last week, I received some interesting iPhone download statistics.

The key point was that even as the number of iPhone applications has proliferated to 75,000, and over 1.8 billion apps have been downloaded, the average number of downloads per app has remained remarkably constant at around 25,000. The charts below show the story (click to see them at full size).

Number of iPhone applications and downloads

The average doesn’t tell the whole story. It’s a mean, which assumes that downloads are evenly distributed. It is much more likely that some applications have a million downloads and some have almost none. A median would be more useful, but I don’t have the data for that.

Nevertheless, the trend line shows that despite the proliferation of titles on the AppStore, the average app has, consistently, had 25,000 downloads.

Which should give heart to aspiring iPhone developers the world over.

You should follow Gamesbrief on Twitter here.

1/3 of smartphone users have never downloaded an App – the figure for iPhone is close to zero

20% of Facebook gamers spend money, an average of $35 per month

Ad-funded or paid-for? An iPhone case study

How many units does a succesful XBLA game sell?

New business model alert: Lite is dead as freemium comes to iPhone

Posted by Nicholas Lovell on October 6, 2009 · Filed under Statistics, iPhone

Julio Medina

I like this info...

WireBear

Good Article! It would also be interesting to find out how the "average" paid app does in sales.

Luke Halliwell

Um, I just did it with these 2 lines of Python :)

total = sum([(1.0 / r) for r in range(1, 25000)])
print 1800000000 * (1.0 / 37500) / total

I'm sure there must be some statistical software that lets non-programmers look at distributions, but I don't know of it. It's probably possible in Excel somehow too.

Nicholas Lovell

That's a really interesting analysis. I hadn't heard of the Zipf Law before. Where exactly do you "plug the numbers in"?

Luke Halliwell

Of course, the median is really far more important to "the average" developer, so it would be interesting to see that. As a quick thought experiment, I tried plugging in these numbers to the "Zipf Law" (downloads inversely proportional to rank) - which is commonly observed in "long tail" situations, and that suggests the median would be about 4000 downloads.

Would be good to know the true figure though :)