How to measure retention in mobile apps using Google Analytics?
New users, app speed and crashes are all fine metrics. But if you’re really looking to measure if your users love your app, you need to measure engagement and retention. How does one measure if the app quality has really improved over time?
In this post i’ll show you how to use Google Analytics to track engagement and retention for mobile apps.
Retention: If customers find an app valuable, they will keep using it. That’s retention. We can measure retention over days, weeks and months using Cohort Analysis in Google Analytics.
What is Cohort Analysis?
Let’s use a definition from Wikipedia:
A cohort is a group of people who share a common characteristic or experience> within a defined period (e.g., are born, are exposed to a drug or vaccine or pollutant, or undergo a certain medical procedure). Thus a group of people who were born on a day or in a particular period, say 1948, form a birth cohort.
Basically a cohort is a segmentation of users based on date. For Flipkart a cohort can be customers who made a purchase on the big billion day. You can use cohort analysis to find out if these customers are more loyal than others.
Some other cohort examples:
- All the people who installed the app in January
- All the people who signed up via an ad campaign
With the new Cohort Analysis feature in Google Analytics, it’s really simple to get the answer.
Before we begin cohort analysis, we need to understand the following points
Acquisition date: App installation date. Currently we can make cohorts in Google Analytics based on acquisition date only.
Cohort size: Size of cohorts to be compared. Available options are day, week and month. A day comparison can be used to measure metrics of users who signed up on 10th with the following days.
Metric: Retention % is defined as the number of users who came back in this time period/initial number of users. This percentage is tracked over days/weeks/months.
In the above graph you can see that only 12.29% of the signed-up users came back to use the app in the 2nd week.
Defining retention metrics:
D1: Number of users who used the app a day after installation
D7: Number of users who were still using the app after 7 days. Similar Day 30 retention can be defined.
If we look at daily retention numbers, we get a chart like this.
Out of the 709 users acquired on Jan 30th, only 14% remained engaged by day 1, by day 5 this rate had dropped to 4.5%
We can track the day 1 retention rate over time to determine if the app quality is improving(more users stay back). However attributing the improvement to a single factor can be a tough proposition. Correlation doesn’t always mean causation.
We can combine custom segments with cohort analysis to get more interesting results.
Some more useful comparisons:
- Users acquired via which channel have the best retention rates?
- How many purchases does a highly engaged user perform in his life-time?
Metric 2: Loyalty
Loyalty and Recency can be used in combination to define engagement for your mobile app.
What is Loyalty?
Let’s use the definition from Wikipedia.
Loyalty is faithfulness or a devotion to a person, country, group, or cause
A more technical definition from Google analytics help.
The number of sessions that are the nth occurrence in your app.
Let’s understand what a session instance means. So session instance 1: A new user who opened the app for the first time. session instance 2: A user who opened the app for the 2nd time today in the entire history of his app usage.
So you can see that 159 people have opened the app between 26–50 items in it’s lifetime.
The above data is for a single day. Assuming I open the facebook app at least 3 times a day in the past 3 months, i’ve used the app a staggering 2700 times already. You can accordingly arrive at a suitable be benchmark for your app.
You can see that 2nd day sessions are relatively high. But that’s expected if you have a high acquisition rate. However the rate drops off drastically after 2 sessions. We can conclude that the chart for a highly engaging app would look like this.
Success can mean different things for different categories of apps. Let’s take for Makemytrip app. Since people travel less often, a user with 5 session instances might be inclined to continue using the app.
Along with sessions we can use avg session duration and goal completion rate to get more meaningful insights.
Metric 3: Recency
It’s great to find out the number of times users typically use the app in it’s lifetime. An interesting data analysis would be to take this data and use the app-uninstall information to find out the average number of items an app is used.
However how often do your users open your app? Do they use it day after day like whatsapp? The recency reports gives us the number of days between sessions.
Fig 5: Recency chart for sample mobile app
The first bunch is the new users. From the above chart we can conclude that users usually delete the app after 30 days.
To summarize, if you measure retention and engagement for your users, you’re more likely to build value in the long term for your users and yourself. So start tracking!
Some useful links: