Friday, January 13, 2012

Breakdown of Windows Phone App Download Statistics

Line Graph by Scott Lewis, Noun
In this blog post I want to share the download graph of my quite successful Windows Phone app Pictures Lab and my conclusions. Some of you might have seen a similar graph already on Twitter or on Facebook, but I thought an update and a bit more explanation might probably make sense.





The graph below shows the download statistics of my Pictures Lab app during the last 9 months. The green line represents the daily downloads and the orange are the accumulated downloads. I blurred the numbers at the scales, since I don't release the download number for various reasons. Please note, Pictures Lab offers a trial mode which means users can try the app for free. Therefore the download numbers aren't equal to the number of sold licenses! Although the app has a pretty good conversion rate from trial-to-paid of ~32% worldwide average and in December it even was 97% in Norway! I noticed that all the Scandinavian countries seem to have a great conversion rate in general. Worst seems to be Hong Kong, but that's a story for a different blog post. Just to be clear, Pictures Lab provides me with a solid extra income each month and the trend is positive, but it's not enough to make a living out of that yet.
The absolute download numbers actually don't matter that much for this blog post. The important part of the graph are the text labels I manually added. Those labels mark events I tracked and noted at that time and are usually followed by a download peak.

Pictures Lab download graph from 04-01-2011 to 01-07-2012 (click on the image for the original size).

My conclusions
  1. Steadily releasing updates with new functionality and being featured at the Windows Phone sites has a big impact on the downloads. Keep in mind the major Pictures Lab updates were featured at least at 3-5 of the top Windows Phone sites.
  2. Being featured at the Marketplace also has a significant positive impact on downloads. Of course this depends on how big the local Marketplace is. Unfortunately does the AppHub not provide the information when an app has been featured. There is the Distimo service which tracks that too, but that service needs your live id credentials and I won't give a 3rd party site my live id to store it in their database. Fortunately some of my friends ping me when they see my apps being featured in their local marketplace. I think it's pretty much safe to say that all the graph peaks without a label nearby are related to a Marketplace feature.  
  3. The base line increased not only after the 3.0, but also after the peak of the 4.0 update. The 4.0 update brought the multi-language support with 10 languages. The steady download jump after 4.0 is a pretty good indication that localization can help to increase downloads.
  4. The price drop to $0.99 was followed by the highest peak in the graph, but after that it went down pretty quickly. I think the price point is the hardest thing to get right and it largely varies for different kind of apps, so the above Pictures Lab pricing information shouldn't be generalized.   
  5. Of course Christmas, the holidays and New Year also resulted in nice peaks. Customers have time to browse the marketplace or just want to fill up their brand new Windows Phones with some apps.

What about the Windows Phone 7.5?
The first Windows Phone 7.5 (previously known as Mango) devices and Nokia hit the market in October and November 2011. This fact also needs to be considered when interpreting the graph. Although from looking at the statistics of my other apps I see that Windows Phone 7.5 didn't have such a huge impact like the updates had. The downloads definitely increased due to the Windows Phone 7.5 launch, but the tracked events align very well with the peaks in the graph and the increase of the base line after.
Let me try to clarify this and my conclusion #3 with the graphs of my successful Helium Voice Free app and my not so successful Benchmark Free app.

Helium Voice Free got a nice update in November, but I didn't inform the Windows Phone sites, so there are no peaks in the graph. You can see that the release of Windows Phone 7.5 slightly increased the downloads and the downward trend turned into an upward trend.

Helium Voice Free graph from 04-01-2011 to 01-07-2012 (click on the image for the original size to see numbers).

















In the Benchmark Free download graph there's also this small increase of the average downloads around the Windows Phone 7.5 device launch, but it's not as much as after the updates of Pictures Lab.

Benchmark Free graph from 04-01-2011 to 01-07-2012 (click on the image for the original size).


















Considering all that information, I think this means the Windows Phone 7.5 launch of course raised the downloads base line, but the Pictures Lab updates had a bigger impact on the downloads and it's not only the Windows Phone 7.5 launch which raised the average downloads of Pictures Lab.
Keep in mind that such a statistical analysis isn't bullet-proof at all, esp. when there's not all the information available like the date of all marketplace features and more. Please also note not all apps are the same and things that work for Pictures Lab or Helium Voice Free don't necessarily need to be valid for other apps.

Makes sense?
What are your conclusions?
What is your experience?


Thank you!
I would like to thank all the users of the app, the great group of translators who helped me to translate Pictures Lab, the beta testers and not at last all the great Windows Phone sites which keep us informed about all the things happening around Windows Phone and help us developers by informing the world about our apps. Keep it up!

8 comments:

  1. Great information - have you considered also having a free app who's primary job is to get people to upgrade? Optionally, you can put advertising in the free app (both for revenue and to give another reason to upgrade). I had posted a blog on the combination of trial/paid and free and in my specific scenario it made sense.You can see that entry here http://apps-are-people-too.blogspot.com/2011/06/increasing-app-sales-with-analytics.html My paid app does not have the numbers that yours does and so what made sense for me might be completely wrong for you. Cheers!

    ReplyDelete
  2. I also have a couple of other apps where I use that model you describe (free+ad and paid version). Works great for certain apps like http://bit.ly/HeliumVoiceFree. Pictures Lab is a different kind of app.

    ReplyDelete
  3. Nice insights! Not sure why you don't think that Mango had an impact though. Looks to me that the average daily download trendline was flat until ~mango release timeframe and it's obviously looking up since about October. Obviously you had your own huge updates during that timeframe, but looking at the graph it's easier to argue that Mango had an impact than that it hadn't, imho :)

    ReplyDelete
  4. I looked at my other apps and I see a slight increase at that time, but not in relation. To be clear, it had an impact, but it was not that huge like I wrote.
    Thanks for the feedback, I will update the post later and maybe add some graphs of my other apps for comparison.

    ReplyDelete
  5. And today - Feb 18 - is the upward trend continuing?
    The way I see it here in Europe, with the Nokia Lumia now launched almost everywhere, Windows Phone usage should continue to go up. And your app downloads as well. Do you confirm?
    If yes, I'll buy myself some Nokia shares! :-)

    ReplyDelete
  6. I will have to check it. The trend is continuing to go up, but not that steep. AFAIK, Lumia is selling OK, but not like cake.

    ReplyDelete
  7. That's how I also see it: Lumia is selling, but not like cake. But that's OK - for me the big question is/was: will it sell at all, ie. 1) do consumers still like the Nokia brand 2) can consumers learn to like/love Windows Phone?

    For me, based on (amongst others) your stats, the answer is yes (to both questions). And that's very good news for Nokia & Microsoft. They now 'just' need to execute in 2012 & 2013.... and build up that 3rd ecosystem.

    Thanks in advance for your update(s)!
    Jan

    ReplyDelete
  8. Just checked it, the baseline stays at the level of the last weeks in the above charts, so it hasn't dropped.

    ReplyDelete