Author: 
Frédéric Block

If you have an app available on a store, such as the AppStore or the PlayStore, you probably want to improve the 'user retention'. But what's a normal retention rate? How can you ensure the users who installed your app will actually use it... and keep using it...

Retention : 90% of my users are gone...

Many users install an app... and drop it. Did you know 70% of users only keep a new app one day max before removing it ?
It is today a 'industry-accepted-standard' that user-retention rate can be considered as valid after 90 days.
Data collections also indicate that a very successful app will have a 20% retention rate after these 3 months. Meaning 80% were lost in the process... Most successful apps will actually tend to achieve 5 to 10%.
So if you have (only) 1 user out of 10 still using your app after this time period, be happy, you are good :-).

Improve your app

But how can you improve your app to ensure the best possible retention ?

  • Onboarding

Most users quit your app because they do not understand it. And they did not want to take time to try to. There are enough other apps to test...
So you have a -very short- time to convince.
You have to offer a very appealing and self-learning onboarding. User should be able to get the point and the added value of the app during the first 2 minutes... At the end of onboarding, end-user should feel self-confident, wanting to try out what he just enjoyed...
Do not forget : most of the users who leave the app during onboarding will never start the app again...

  • Push notifications & App awareness

There is a risk users who installed your app never open it again. Maybe because the app is not meeting their expectations... or maybe because they simply forgot about it being there.
So your app should be able to remind the user about its presence, and push notifications can be a way of doing this.
Of course, such notifications have to be appealing enough (and not too intrusive) to be read and followed... Try to make these context-sensitive (based on location, time, content,...), and as much personalized as possible : one user is not another...
Besides, do not forget almost 50% of users opt-out of push notifications... You might need to remind them using another way (such as mail, general awareness through re-marketing,..)

  • Screen intuitiveness

Everything should be simple. Easier said than done.
You might want to use this tip : take somebody who never used your app. if you need to explain him how the major driver (=why a person would use your app) works, your app will fail. Simply because the new user will have no one next to him to explain...
So keep the basic stuff very accessible and understandable. Hide the advanced features. They're scary. And the advanced users will discover these soon enough...

  • Comparative testing

How can I know whether users will like the new way my app works ? One of the best method to determine this is to 'ask' them. Since you will probably never be able to really ask this (and get an answer), you have to find it out by yourself.
If you hesitate between 2 ways of doing something, you should build both. And compare the results. If the users with one solution tend to remain longer, it probably means you will go further with this one... You can also build your app in such a way both solutions co-exist. Some will see / use one, others will have the other. Without even knowing a different way of working -already- exists.
Did you know Facebook has constantly dozens of app functionalities being user-tested this way ? Grab 10 persons you know and compare their app's screens...  you could be amazed :-)

  • Use your data

Mobile apps can ( / should always) collect usage data. Besides, if you have server-side data, exploit these to discover what's used ... and what's not. The unused features are either not seen as you would have thought... or these are simply useless. So make these more visible / appealing or remove those... they are maybe part of the scary part of your app...

These are of course not the only ways of improving an app, but they are probably the most efficient ones for doing so...
... and these were the principles we also applied on InboxZero these last months in order to define the forthcoming roadmap. I will explain this in a future post.
 

Newsletter