Don’t be tricked by the “Monthly Active Users” engagement myth

May 28

Don’t be tricked by the “Monthly Active Users” engagement myth

Metrics for Engagement

Lots of App developers just operate on “how many installs” – thats a vanity metric.
So smarter App developers talk about “Actives”, this can be:

  • MAUs (Monthly Active Users)
  • WAUs (Weekly Active Users)
  • DAUs (Daily Active Users)

 

WAUs and DAUs tend to be too volatile for smaller organisations because its got statistically insignificant changes.

WAUs and DAUs are also too much detail for enterprise Apps (Brands, Retail etc) because the reporting cycles are too slow to use the data anyway.

So most smart App owners drop back to MAUs as the default top-level metric for “mobile app engagement“.

When the Metric Fails

This is sad tale that could have been avoided if two tools were used:

  1. cohort analysis and
  2. segmentation

But first the story. AcmeApp [not their real name – obviously!] is (was) a venture funded mobile startup going after a big consumer opportunity. It was social and viral and MAU=32-37%.

That sounds OK – right?

Nope.

AcmeApp was spending significant money each month on customer acquisition and so new installs a major contributor to the MAU number. In other words Acme was buying MAUs. This is the classic leaky bucket I’ve written about before – its incredibly important from a company survival perspective that you are not obscuring the engagement and retention qualities of your app with your own marketing spend.

The result is that when¬†AcmeApp went for more funding, the due diligence process drilled into retention just by asking the obvious question “Do people really love the App enough to keep coming back?” The subquestion to this is “How many people still use the App in Month 2”, “….Month 3”, “….Month 4” etc – you get the picture.

So this brings us to cohort analysis. In this chart, we measure the how many users are retained (“ACTIVE”) from each month, the total in any months is the MAU figure.

Basic Cohort Example

Basic Cohort Example

So what you really have is a great comparative display of how sticky your App is. Of course if you are developing games, then weekly might be a better cohort granularity.

So full credit for this simple example goes to Andrew Chen and Christoph Janz who wrote the blog post. For any entrepreneurs who want to pull the wool over their investor’s eyes about retention, then don’t try pitching to Christoph!

Now whilst we think that cohort analysis is great, we actually like segmentation better.

Segmentation Visualization

Segmentation Visualization

In this chart we display the relative size (population) and value of segments (Y-Axis). In this visualisation you see the users who are potentially churning off the application. Its more explanatory of where users are heading (up to the right is more valuable!)

So you can think of:

  • cohort analysis as WHAT
  • segmentation analysis as HOW

 

If only AcmeApp had used one or more of these tools and acted to solve their retention problems.

  Get our Referral Program Guide