We’ve published the presentation we gave tonight at the New York iPhone Developers Meetup, based off of our first 30,000,000 application downloads, here on SlideShare.
We’ve published the presentation we gave tonight at the New York iPhone Developers Meetup, based off of our first 30,000,000 application downloads, here on SlideShare.
This presentation was incredible — really rich data and insights. I especially liked the logic behind why ad supported free apps don’t make financial sense.
Very interesting stuff guys. Thanks for sharing.
Interesting! Helpful! Many thanks for posting it.
wow! Is this number for real – 30,000,000 downloads? or is this a typo?
No typo. We’ve got a lot more data than that now – that’s what we had when we started the analysis.
Best,
Greg
Can you comment more about how you collected this data?
For instance how many apps make up the 30 million downloads? Do each of the “biggest apps” have more than 3 million downloads or are the “biggest apps (how many?) responsible for more than 3 million downloads? What is the average number of downloads per app? Are you counting application updates in the number of downloads? Or can I take your average to represent the average number of paid customers (I guess I am asking for the averages broken out by free versus paid) for which I can reasonable strive?
Keith -
The data’s aggregated from the few hundred applications that use Pinch Analytics.
We’re pretty quiet about details about specific applications, of course. Only a very small number break the three million user barrier.
Application updates aren’t counted. These are new installs – and we only count one install per user, so reinstalls don’t count either.
Best,
Greg
Great data. Read through it 3′xs and will have it link to it from my blog! We’re just about to launch our 2nd app on the app store and have been thinking through the differences between paid & free. All that work for $.70 per user or even less with ads. Wow.
Thanks,
Brook
Pinger
Thanks for sharing.
Great info!
Many of your slides site USAGE or TIME SPENT in minutes (e.g. the Engagement slides).
Is that minutes/day?
Thanks
In your analysis of free vs paid apps, it’s actually worse than you indicate. You don’t have 80 sessions to make the 70 cents/user of paid apps. According to your listed assumptions, you actually only have 12 sessions. “Assume free applications are run, at most, a dozen times per year.” — That’s 12 sessions. Not 80.
Your “Doing the math” point on slide 23 is invalid. You are multiplying the number of sessions for _free_ apps, by the number of times _free_ apps are run more than paid apps. You are doing an invalid conversion. One that you don’t need to do anyway.
Biff – I agree that for *revenue per user*, you’re getting $0.70/user from a paid application, while only the revenue you can earn with 12 impressions from a free application. On a revenue per user basis, free can’t compete with paid under any circumstances.
That said, free applications do have more users than paid applications, so we’ve tried to account for that in our model.
That’s why we do the multiplication – if for every paid-application user, you’d get six to seven free-application users, maximizing revenue requires you to compare the revenue of one paid-application user to the revenue from six-to-seven free-application users.
Hi Greg,
Thanks for the data- it was very useful. I had one problem though -
it seems like your analysis attempts to directly compare free and paid apps as though they’re made of the same stuff. I’m not sure they are of the same quality. For example I might create a game with awful graphics and offer it up for free with ad support, even though I might spend more time on the UI and design of an app that I mean to sell. This means that you couldn’t simply compare revenues between the two markets (paid and free apps) unless that decision was being used to determine whether to develop a free app (developed with free app quality) vs a paid app (developed with paid app quality).
Developers should recognize that offering apps for free is comparable to offering that app for a fee that is less than minimum revenue per paid app (70 cents) but equal to the revenue gained from advertising.
Then they should decide whether it makes more sense to make a paid or a free app- given the expected time/costs involved for each and the expected revenues that one would obtain from each.
My point is this: it might very well be that the free apps that sold 6 times the number of paid apps sold would actually not sell 1/6 their current sales if they themselves were offered for a price.
Am I on to something, or did you already account for this effect in your statistical analysis?
Stephen,
Thanks for your comment. You bring up a very valid point – free applications are not the same as paid applications. Thus for our analysis, we primarily focused on examining situations where an application changed its price. By holding the type of application constant we sought to determine the effect of price on distribution. We then were able to cross check this data against data for free vs. paid applications overall to minimize the confounding effects of application type, ranking in lists, launch date, etc.
The ultimate goal is to put together a price elasticity curve so that we can better help developers optimize the pricing of their applications. However, before we share this data with the public we want to get a better sense of other factors at play.
Josh
Great – Thanks Josh!
Can you share any data on the average amount of time that an iPhone user spends interacting with apps (free or paid)?