New User Retention

196 views
Skip to first unread message

JPhilippe

unread,
Aug 17, 2018, 10:06:20 AM8/17/18
to Firebase Google Group
Hi, 

I'm working on downloads and retention figures for an iOS app.
I have compared the itunes figures (App Units) and the New User figures (from Firebase > Retention), and there is 100% of difference between both.

Did you already see a so huge difference?
How can I explain this gap?

Thanks a lot

JPhilippe

Ian Barber

unread,
Aug 20, 2018, 4:27:37 PM8/20/18
to Firebase Google Group
Hard to say without a few more details. Are you checking first_opens? What is the direction of the difference?

--
You received this message because you are subscribed to the Google Groups "Firebase Google Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to firebase-tal...@googlegroups.com.
To post to this group, send email to fireba...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/firebase-talk/aacf55b5-f6aa-4d6f-a345-748b44b7ad0d%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

JPhilippe

unread,
Aug 21, 2018, 9:47:03 AM8/21/18
to Firebase Google Group
Thanks for your answer.

I use the "new user" figure coming from the retention section.
It's around twice the number of download (app units) on iTunes.
This figure on itunes sounds good.

By the way the "new user" figure is consistent with the "new users" (in Firebase)

I'm checking the sdk integration too, cause it's really strange to have such a difference.

JPhilippe

unread,
Aug 21, 2018, 9:47:05 AM8/21/18
to Firebase Google Group
I'm checking new users in the retention section.

Capture d’écran 2018-08-21 à 11.02.01.png


New users are twice the number of App Units, on itunes.

This figure sound good on itunes.



Le lundi 20 août 2018 22:27:37 UTC+2, Ian Barber a écrit :

Jean-Philippe BRIGUET

unread,
Aug 24, 2018, 10:55:47 AM8/24/18
to fireba...@googlegroups.com
Hi Ian,

We check the firebase implementation, and everything sounds good.

Have any idea about this huge difference?
Did you already notice this kind of behavior?

Thanks!

Jean-Philippe

Ian Barber

unread,
Aug 24, 2018, 12:53:39 PM8/24/18
to Firebase Google Group
Sorry, I'm still not sure what the difference is! The retention graph is showing percentages of cohorts over times, so I'm not sure how you're comparing it against a downloads number? Would you be able to give specific examples of numbers that don't match up?

Jean-Philippe BRIGUET

unread,
Aug 28, 2018, 10:14:05 AM8/28/18
to fireba...@googlegroups.com

Hi Ian,

 

Thanks for your answer.

 

In iTunes Connect, I can see the number of downloaded apps.

It is called « app unit » :

« The number of first-time app downloads made on the App Store using iOS 8 or tvOS 9, or later. App updates, downloads from the same Apple ID onto other devices, and redownloads to the same device are not counted. Family Sharing downloads are included for free apps, but not for paid apps. »


Capture d’écran 2018-08-28 à 05.57.48.png


I have compared this figure with the number of new users in the cohort analysis of user retention.


firebase.png

26 new users (Firebase) VS 14 app units (iTunes)

This would mean the firebase sdk is initialized twice per download.

This is strange…

 

And this ratio is stable for a bunch of weeks:

Capture d’écran 2018-08-28 à 06.13.21.png

I thought it was maybe the SDK implementation, but we checked it and everything seems good. There is only one Instance-ID generated by app.

 

The problem may come from iTunes.

But it’s a really core data, and they generate it for a while.

 

What about you?

 

Thanks a lot for you precious help.

 

Jean-Philippe



Ian Barber

unread,
Aug 29, 2018, 1:17:41 PM8/29/18
to Firebase Google Group
Yeah, that seems pretty consistent. I don't have a good explanation for the difference there - is there any other potential source of users (e.g. versions of the app distributed over testflight, use of Firebase Test lab or other testing tool, or similar)?

Reply all
Reply to author
Forward
0 new messages