"Keep-Rate" goes down and down?

Hey guys. At the start of my app (5000 downloads), about 50% of the users deleted my app. I thought that was ok. At 10’000 downloads, this value went to 60%. Now, at 25’000 downloads, I have 800 downloads, but 650 uninstalls/day, which seems pretty much to me.

I this development normals as an app becomes “older”???

Monitor reviews in Developer Console. Do you use intrusive ads? (notification ads for example) They tend to cause more uninstalls.

Could point to another problem. Your users don’t know enough from the description to understand your app. So some people install it, then decide it was not what they wanted.

Bonus question: would you rather have lots of installs and uninstalls … or fewer installs, but users keep your app ? How does Google treat each app in each case ?

I will use them soon because I got pissed off, but I didn’t use them yet.

Well I got one MIDI controller app and just every 20th user understands the app but that’s something I can understand.

Another app plays audio itself and the description is clearly. I mean, I’d understand if the rate was such a bad value whole the time and it just stays like that. But the “break” was 7 days ago. There is another new app out now that does pretty much the same but it has a worse rating (I think it’s worse, too) and not so many users like I had in the first days. Additionally I think, the other app can’t have a lot to do with me because then the download rate would sink in first instance, but the download rate is nearly the same like before.

Google Developer Console really should include a daily install/uninstall ratio graph - and seems like the obvious thing to have there (could it be that because it is very close to one of the factors which Google uses to rank an app, that Google deliberately not put that metric in the graphs ?).

I did however draw some graphs like that earlier and I noted something similar - i.e. the retention rate goes down.

Some factors that could be very neutral (i.e. nothing wrong with your app) - is that if the download rate (per day) is increasing, you maybe getting more people coming to your app (compared to the early days when fewer people visited or visited only because your description matched their search terms). As your app becomes more established it can start to appear in other apps’ Users also visited/downloaded section - this may raise the number of users being attracted to your app, but may no longer be the same “quality” (i.e. matched to what your app actually does). Or some such reason.

I think the existence of a competing app can make a difference also - esp. if you appear in that apps’ “Users also downloaded” list - in which case you may immediately do a fast comparison and rapidly dump the slightly lesser app. This may happen (slightly) even when the other app is worse - as some users will for some reason think that app is better (has better colors etc.).

Another thing to consider (and this is more significant probably) - is that with an “Active User” base of say 10K users, and 1K users coming in daily, your uninstall rate for that day will include the 1K users getting tired of your app, PLUS it will include the attrition from the 10K users “Active User” base (!)

This is one of the major reasons why nearly all apps experience stagnation - as EVENTUALLY no new install performance can outrun the attrition you are experiencing from your installed base (i.e. the “Active User” base).

Take an extreme example - if you are getting 1K app installs per day, and have 1M installed base - you can see how the behavior of your installed base will DOMINATE what happens.

For most apps (that are unchanged, or not radically get super-great) there will eventually be an “equilibrium” as the attrition from the installed base matches the daily retained installs (i.e. new installs minus new uninstalls).

If the app is good, it’s equilibrium point (for the “Active User” base) will be 5M, but if it is not so great the equilibrium maybe at 1M.

In fact, if your new installs minus new uninstalls daily figure is not big enough (or starting to dwindle following Christmas season) then your Active User base (installed base) may actually start to decline (since you are not able to replace the attrition from your installed base by the new net-installs that you get on daily basis).

Hope this helps.

P.S. - it is for this reason, the Active User base is essentially irrelevant (unless you have some way of galvanizing your installed base - perhaps by issuing notifications etc. so users revisit your app - the more installed base you have the more you can do this). Since many users install an app and fail to revisit it again much.

For ad revenue etc. - most of the time what is relevant is the “daily active user” or DAU.

If you are using push notifications, then the Active User base maybe important - as you can potentially make ad revenue from all the users who get push notifications (even if they haven’t used your app in a while).

Thank you for your detailed answer, it really helped me understanding and confirming my own thought (I had the 10k example in my head as well).

Yeah I will kick those notifications in, as far as I think this business is over, it’s like you said, an app to have fun with 2 minutes/day one week long then it’s just boring.

Can you tell me what is normal eCPM with push notifications? Is it more than with banner clicks?

Greetings, Alex

I don’t have direct experience with push notifications - others here would know more about it.

The point is that you should not bother about Active User base - for a game perhaps it is important (esp. a social one where it helps if more people use the app) - but on a daily basis the most important thing is DAU (Daily Active User) - from ad revenue perspective and because it more closely indicates real use of the app.

DAU will include the daily new users (perhaps including the ones who uninstall ?) - and it will include a proportion of the Active User base (i.e. installed base).

For example for my apps I have seen that (looking at AppBrain stats - which you get for free if you use their SDK) - that the new installs maybe 1500 but the DAU are 3000 or so - which means 1500 are coming from existing installed base who opened the app that day.

So if Active User base starts going down - then you can see that the 1500 from installed base will start to be affected - while the 1500 from new installs are almost completely dependent on how “searchable” your app is (it’s description and if people searched for that stuff today) and whether your app is mentioned in other apps’ “Users also downloaded” etc. i.e. all the funnels via which you can get traffic to your Google Play app page.

The use goes down as well. I use admob as an indicator for it, had always between 20k and 25k requests, it is now slowly dropping, got 17k yesterday :confused:

Yes, it is normal

Hm ok… A bit disappointing, but at least I kept the industrial country users so my eCPM is now almost .50 :confused:

Now the moment came where the active installs go down… fuck