[asimjalis]

I think you are fundamentally asking the wrong question.

Instead of trying to understand why 80% of the people are leaving, try to find out why 20% are staying.

Instead of contacting the people who are uninstalling the app, contact the people who install it and stay with it. Ask them why they are buying and what they are using it for.

The people who are uninstalling are not your customers. The people who install and stay are your customers. Find out why they like the app and then use these insights to attract more customers.

[tarun_anand]

Well - in my opinion - having done this day in and day out for last 4 years - the key thing is to look at both users who are leaving and who are staying. We understand the difference between the two and then try to tilt the uninstalling users towards the installing users behaviour. However, sometimes the uninstalling behaviour is specific to something that cohort only and hence both cohorts need to be looked at.

[chupasaurus]

>Instead of trying to understand why 80% of the people are leaving, try to find out why 20% are staying.

That's a survivorship bias. You have to analyze both groups in order to improve the ratio.

[dnbgfher]

Their comment wasn't directed towards improving the ratio. They were suggesting that the focus should instead be on the current users. I imagine this is a result of everything being so new, they think it makes sense to take some time and figure out why people are actually using the app before trying to tackle more directed problems like the uninstall ratio.

In that context, it's really not a survivorship bias. If you want to know about the survivors, it's not wrong to consider the surviors only.

[chupasaurus]

But the OP's question is about improving the ratio.

[dnbgfher]

Right... and then they got a response that they might want to ask a different question along with a suggestion of how to answer it.

Now you are trying to treat that solution that is explicitly meant to answer the new question as if it were an answer for the original question, going so far as to apply logic issues from the old onto the new.

Discussions, particularly ones of this nature, change and shift direction. It doesn't mean that later topics inherit logical traits or something from earlier ones.

[Nightshaxx]

Agreed. If a minority of people like the app and you want to appeal to the majority, don't ask the minority what they want because those might be reasons the majority don't care about/dislike.