Let me tell you another story.
A company we work with discovered they had a very low first week retention rate for the users who created a new account: 20%.
Instead of panicking and acting on it, this is what they did:
First, they tried to validate the data.
They realized they were monitoring the users, and not the accounts.
And since an account can have more than one user, and it is the accounts that are paying, not the users themselves, this led to false results.
After this discovery, they came to the conclusion that the first week retention was 30%. But it was still very little.
Second, they tried to explain why the situation was so bad.
They realized that the app had an on-boarding rate of 40%.
This meant that 60% of those who made an account never followed through. They were not relevant in calculating the retention rate.
So, they applied the retention reports to those users who created an account and also finished on-boarding. And the result for the first week retention was 50%.
Third, they made an estimate of what could happen if they registered an increase from 50% to 60%. This increase seemed realistic and it would generate 15% more revenue / month for the existing customer base.
Forth, they looked for a possible solution.
In their case, if they could identify the actions that made the difference between the users who came back and those who churned, they could create an automatic process to alert the support team, who could contact these potential clients before it was too late.
Last but not least, it was high time he broke the news to the manager.
Armed with all this data, the analyst decided to not to leave any essential details to be guessed. He:
- confirmed the validity of his discovery.
- explained the situation in detail.
- presented the forecast he made.
- and offered the manager a potential solution.
Do you think his approach got him the expected results? - Of course it did!