When it comes to tracking user engagement, there are two main approaches:
- measuring a user's engagement within the app in terms of stickiness, duration, and frequency, and
- looking at engagement in terms of retention and the user's lifecycle.
In addition to these two approaches, we recommend to track how well specific activities intended to increase engagement are performing. These activities may look very different depending on the product and service, but messaging campaigns using e-mail, push, in-app or text messages will usually play a major role.
Measuring user engagement within your product
So let's have a look at how to measure a user's engagement within your mobile app. Ideally you will want to create a dashboard in your analytics tool of choice that displays
- Stickiness: how often your active users are returning to use your app (e.g. DAU to MAU ratio)
- Engagement duration: how much time your users spend within the app on average (e.g. average time spent per daily/weekly/monthly active user)
- Engagement frequency: how much users interact with your app (e.g. average number of sessions per user)
It is important to measure these engagement metrics not just across all users, but per user cohort. For instance, your core user cohort (e.g. frequent listeners from Sweden) may have very different engagement metrics than your average user.
Just looking at daily active users or general engagement durations and frequencies will not be enough for most products. You need to define the most important action that you want your users to perform over and over again and measure engagement in terms of this action only.
For instance, for uber it is not as interesting how often their users open the app but rather how often they search or book a ride in order to determine engagement. Now, let's have a look at the best way to measure each of these metrics.
Measuring the stickiness of a product is one of the key engagement metrics to look at. VCs usually will want to understand the stickiness of your product before making an investment decision. Consequently, let me explain to you in a bit more detail how to measure it and why it is important.
The usual stickiness metric is calculated by the ratio of DAU to MAU, i.e. by dividing the number of daily active users by the number of monthly active users. Depending on the frequency your product is used, you may want to calculate different ratios such as WAU to MAU or DAU to WAU. This ratio allows you to understand how engaged your users are. For instance, a DAU to MAU ratio of 30% means that your users on average engage with your app on 9 out of 30 days (9 divided by 30 = 30%).
In comparison, simply looking at the DAUs or MAUs individually does not tell you anything about engagement as you do not know whether the DAUs are the same users day over day. Even though looking at DAU and MAU is much better than looking only at daily or monthly users, DAU and MAU metrics are often already considered vanity metrics.
As mentioned in the beginning of the section, make sure to look at these ratios for your main actions and events that drive engagement and segment it per user cohort to get the most out of it.
Measuring a user's engagement duration is pretty straightforward and can be achieved by looking at the average time spent per day, week, or month per user within the app. Depending on the product, the duration may or may not be important for engagement. For instance, social networks like Facebook are eager to increase the duration spent within the app in order to increase the number of times ads can be shown to their users while companies like uber might care less about this particular engagement metric as it does not increase the number of rides ordered.
The starting point for measuring engagement frequency is to have a look at the average number of sessions per user within a certain time interval. This average number of sessions can, again, quickly be considered as a vanity metric. As a result, we recommend looking at the average number of times a specific event is triggered. For instance, in the case of a music app like Spotify, you may want to track engagement frequency of your user cohorts based on the average number of songs played per user session. In the case of Uber, it might be the average number of rides per month. As with all engagement metrics that we are talking about, make sure to slice and dice the engagement metrics per user cohort to get the most out of it.
Retention and the user’s lifecycle
Something that is often overlooked when talking about user engagement is that engagement and retention are tightly connected with each other. High engagement increases retention and high retention increases revenue and referrals. Since engagement is impacting retention rates, it provides a lot of value to look at retention curves and the user’s lifecycle since this is to some extent the output of engagement and therefore helps you analyze your users’ engagement.
Your product's retention curves and user lifecycle analysis will provide you great insights on how engaging your product is.
Let's have a look into retention curves and user lifecycle charts.
Retention curves are one of the most powerful analyses out there, and it would go beyond the scope of this article to describe them in detail. In short, they show you how many of your users (or what % of your users) are using your services on a recurring basis. If you are expecting a high frequency in the usage of your app, you look at daily retention rates showing you how many of your users that used your app on day 0 returned to use the app also on the following days. If you are expecting a lower frequency in the usage, you may look at weekly or monthly retention curves. And again, you will want to look at this differently per user cohorts; for instance, uber may have power users for whom they look at daily retention curves while for the average user a monthly retention curve may be more suitable to look at.
Retention curves start on day/week/month zero at 100% usage since you are only looking at users who performed the specific event your are investigating. From there on, you will usually see a first sharp drop because a good chunk of your user base will not return to perform the same action again on the following day/week/month. After the first drop, the retention curve usually behaves in one of the following three ways:
- Regular: Your retention curve flattens after a certain period. This is normal behavior. Your goal should be to increase the point where it flattens (i.e. the percentage of users retaining) as this indicates an increased retention rate. A flat curve symbolizes healthy retention. However, you want the percentage of users where it flattens as high as possible.
- Declining: If the curve does not flatten but continues to fall, then you are running into some serious issues because you are not retaining the users that you are acquiring at all.
- Smile: If the curve flattens and then increases after some time again, you have a very healthy retention curve as this indicates that your most loyal customers are more likely to stick around than your average users.
Like for the other engagement metrics, you will want to look at retention not only from an app usage perspective, but also from the perspective of your most important use cases. For instance, Spotify will be interested in the retention of users who listen to music. By looking at retention curves per customer segments and cohorts, you may identify behaviors that increase retention and engagement - to stay with the Spotify example, listeners in a certain age group or users with a Spotify Premium Family Plan may have a much higher retention rate than regular users. As a result, Spotify may target these users specifically in their acquisition and activation campaigns.
User lifecycle analysis
The user lifecycle analysis chart provides a great overview of how your active users are interacting with your service over time. It distinguishes between active users in the following categories:
- New Users: are newly acquired users that used the service for the first time (e.g. a New User in July used the service in July for the first time)
- Current/Retained Users: have been active users in the current month and previous month (e.g. a current user in July used the service at least in June and July)
- Resurrected Users: users that became active in the current month that were not active in the previous month but had been active users of the service some time in the past (e.g. a resurrected user in July was active both in July and in May but had not used the service in the months in between)
- Dormant Users: are users that were active users in the previous month, but were not active in the current month (e.g. a dormant user in July was using the service in June, but not in July)
The following figure explains this concept more visually.
The user lifecycle chart shows how your active users are distributed among these categories and how this distribution changes over time. The sum of all active users from the previous month equals the sum of Current Users and Dormant Users in the following month; i.e. users either continue using the service or become dormant (which does not necessarily mean that they churned, they can still become resurrected users again).
The user lifecycle chart allows you to understand the growth of your service among your active user categories. If you have a highly engaging service, then your Current User base should be continuously increasing. In addition, highly engaging services have a higher chance of resurrecting users.
Like with retention curves and other previously discussed engagement metrics, we recommend to look at the user lifecycle by specifying the use case you are most interested in (e.g. listening to a song or following an activity) instead of looking at the general usage of your service.
Engagement is usually considered as one of the rather difficult areas to understand and analyze. Depending on your service’s offering, it may also be quite difficult to influence. For instance, a tax reporting service will never achieve high engagement rates during the year and it might not even be required for such a service unless they want to stay top of mind also during the year in which case they should think about diversifying their offering by adding further (more frequent) use cases. For other services, such as Facebook or also Netflix, engagement metrics are more important to understand as they directly impact their revenue streams.
By analyzing the metrics and charts described throughout this article, you will not only be able to understand your base engagement rates but also which events impact your engagement and how to best segment your cohorts. Once you understand how you can increase engagement across one cohort, you can define and run experiments to increase engagement for similar cohorts.
Are you interested in setting up and evaluating your engagement metrics for your business but need further assistance? Get in contact with Kasva.