Cool Tool: DAU calculator based on users retention

Published by Digital Limbo on

By Farhan Haq

When your product team comes to you and tells you they need 10k DAU after one month of (soft) launch, it can be pretty hard for a marketer to know or understand how many users they need to acquire over that period in order to hit that target.

This model (Retention DAU tool) allows you to do just that, with the only pre-requisite being that you need to know your D 1-30 retention data.

  • Step 1 – Gather your D 1,3,7,15,30 retention data from your internal BI, and plug in to cells B2 to B7 – (you can use retention numbers going further than 30 days, but this specific model will predict your DAU based on 30 days of activity).
  • Step 2 – Now you have the Retention Points on your graph, you have to choose a regression line (“y”) that best matches these points. Right click on the current regression line and click “format trendline”.


I have pre-selected the exponential curve for the dummy data, but a logarithmic curve may also be valid here. Don’t use a linear trend, as retention is NEVER linear.

The Key is to have a strong R squared coefficient– i.e. the goodness of fit of your model. This should be above 90% if you have valid retention data that can be modeled.

  • Step 3 – Copy the equation formula into A9, and separate the exponential and the X numbers into C9 and D9. This will populate column J, but be wary that you have to change the formula in this column if you swap from an exponential curve to a logarithmic one.
  • Step 4 – Now plug in the number of users you expect to acquire per day from 1-30 (including any perceived organics you expect). In row 33 you will see the total installs you expect to acquire over this period and your expected DAU after 30 days.


Note – You can also play with the retention numbers, i.e. to see what an uplift in retention will do to your D30 DAU, but remember that your equation for the regression line will change, and artificial retention numbers will harm your R squared coefficient.


Farhan is a growth mobile marketing enthusiast, and has more than 7 years’ experience in performance marketing within the games industry. Starting his career as a statistician for global media agencies Mediacom and then Starcom Mediavest, he went on to have user acquisition roles at the likes of Plumbee, Socialpoint, and is currently Head of UA at Nanobit in Zagreb, Croatia.


1 Comment

Predicting DAU for Mobile Game - Game Development News · December 7, 2019 at 15:41

[…] retentions for few days and predicting remaining days retention. The model was first discussed by Digital Limbo. For the tool be used effectively you need following […]

Leave a Reply

Avatar placeholder

Your email address will not be published.