Five app engagement metrics that connect usage to subscription revenue
Downloads alone do not tell you whether a subscription app is working. What matters is whether people come back, what they do when they return, and whether that behavior turns into paid subscriptions that last. The five metrics below cover that chain: active users, session quality, conversion, churn and realized lifetime value. Read them together rather than in isolation, because each one answers a different question about the same user journey.
How app engagement is measured
Engagement measurement has two sides. One side is behavioral: how often people open the app, how long they stay and what they actually do inside it. The other side is commercial: whether that behavior leads to trials, paid subscriptions and retained revenue. Product analytics tools such as Google Analytics for Firebase, Mixpanel or Amplitude can capture the behavioral side; subscription data supplies the commercial side.
A useful way to organize the work is to ask four questions:
- Frequency and duration: How often and for how long do users engage?
- Depth: Which features and content hold attention once the app is open?
- Retention: Do users return over time, and does that pattern hold across cohorts?
- Monetization: Does engagement translate into conversion, lower churn and higher realized value?
Reviews and ratings add qualitative context, but they are a signal to investigate, not a substitute for behavioral data.
1. Daily active users and monthly active users
DAU and MAU are the baseline health indicators. For a subscription business, a consistently high active-user base is what feeds every downstream metric: trials, conversions, renewals and referrals.
The hard part is defining “active.” That definition is a product decision, not a universal constant. Early on, a broad rule such as “opened the app” may be appropriate. As the product matures, a stricter rule — completing a core action — usually produces a more meaningful signal. Tools like Mixpanel and Amplitude let you configure that definition precisely, which matters because a loose definition can make engagement look healthier than it is.
Once the definition is stable, use product analytics to find which features and content correlate with repeat visits. Personalization based on observed behavior is one lever for increasing return frequency.
2. Session length and session depth
Active-user counts tell you how many people show up; session metrics tell you what happens while they are there. Session length is the clock time a user spends in one sitting. Session depth counts the meaningful in-app actions taken during that sitting — screens advanced, features used, tasks completed. For subscription apps, both measures are worth watching because they describe the quality of the experience rather than just its frequency. Longer and deeper sessions tend to accompany a stronger perceived value proposition, which in turn can support conversion from free to paid and reduce churn. Session data also helps you locate which features or content keep people engaged, and where the interface or content may be creating friction. If your subscription platform documents analytics integrations, you can view these behavioral measures alongside subscription metrics rather than reconciling two separate reports; check the platform's integration documentation for the supported analytics destinations and setup steps.
3. Conversion rate
Conversion rate asks how effectively the app moves people from one funnel stage to the next. The stage you should focus on depends on your model. For apps with a trial, track trial-start conversion (downloaders who begin a trial) and trial conversion (trial users who become paying subscribers) as separate steps. For apps without a trial, track download-to-paid conversion directly, which is especially important for freemium products. Trial length can influence conversion, and shorter trials do not automatically outperform longer ones. The right length depends on the app's context and on how much time a user needs to reach the value moment. Treat trial length as a variable to test, not a setting to copy. On download-to-paid conversion, a 2023 State of Subscription Apps report placed most apps below 2% and top performers around 4%. That figure is a dated reference point from that third-party report, not a current benchmark; check the latest published data before setting targets. The practical takeaway is the size of the gap: post-download engagement — through a trial, free features or a well-constructed paywall — is where the conversion work happens.
4. Churn rate
Churn rate is the share of active subscriptions lost in a period that have not resubscribed. A stable or falling churn rate is one of the clearest signs of a healthy subscription business, because retained subscribers are what keep revenue steady.
Two different losses are often mixed together:
- New-user churn: users of the app who stop using it.
- Subscriber churn: subscribers who stop being subscribers.
Both matter, but they call for different responses. Onboarding is the main lever for new-user churn, since it shapes whether a new user reaches a trial or a first conversion. For subscriber churn, segmenting the data — by country or by subscription duration, for example — can show which groups retain best, which may also hint at where product-market fit is strongest.
Some cancellations are unavoidable. Win-back tactics such as offers for lapsed subscribers, or in-app reminders of the premium features a user is about to lose, give you a way to recover part of that loss.
5. Realized lifetime value per customer
Realized LTV per customer measures the revenue a customer cohort has actually generated over its relationship with the app. It is a backward-looking, realized figure rather than a projection.
The calculation: take the revenue generated by a cohort, subtract refunds, and divide by the number of customers in that cohort. Because it is cohort-based, the number evolves as the cohort's behavior and monetization patterns change.
Realized LTV is often used alongside average revenue per user (ARPU), but the two are not interchangeable. ARPU measures revenue over a defined period; realized LTV follows the cohort across its whole relationship with the app.
Two practical uses follow. First, identify which features or content correlate with higher realized LTV, and refine the product toward what users actually value. Second, use realized LTV as the yardstick for pricing experiments, since it captures long-term revenue impact rather than a short-term spike.
Which metric answers which question
The five metrics are easiest to use when you map each one to the funnel stage it describes. The table below is a reading aid, not a scoring system.
| Question you are asking | Metric to check | What a change usually means |
|---|---|---|
| Are people coming back at all? | DAU / MAU | Shifts in habit strength or in your definition of “active” |
| Is the time spent valuable? | Session length and depth | Changes in feature appeal, content fit or interface friction |
| Are users reaching paid status? | Trial-start, trial conversion or download-to-paid conversion | Paywall, onboarding or value-demonstration effects |
| Are subscribers staying? | Subscriber churn rate | Retention, pricing fit or segment-specific weakness |
| Is the relationship worth money over time? | Realized LTV per customer | Long-term monetization and pricing impact |
A useful discipline: when one metric moves, check the adjacent stage before concluding that engagement itself changed. A conversion drop with flat session depth points somewhere different than a conversion drop that follows a fall in session depth.
What counts as a good engagement rate?
There is no universal threshold. A social app built for daily use and a meal-planning app used weekly are not comparable, and the same numeric rate can mean very different things in each. Three factors shape the answer:
- App category and usage rhythm: expected frequency differs by product type.
- Your own definition of engagement: daily logins, completion of a key action and content sharing are different behaviors.
- Benchmarks: industry figures are useful as rough orientation, not as targets to copy.
A more reliable approach is to compare your app against its own history and against similar apps in the same category, using a consistent definition over time. Published benchmark reports on subscription and monetization metrics are a reasonable starting reference, provided you note their publication date.
A worked example: reading a conversion benchmark
Suppose your app's download-to-paid conversion sits at 1.4%. A 2023 State of Subscription Apps report put most apps below 2% and top performers near 4%. Two readings are possible, and they lead to different actions.
- If your trial-start rate is healthy but trial conversion is weak, the value moment may be arriving too late in the trial. Test trial length and the onboarding path that leads to the paywall.
- If trial starts are low in the first place, the problem is earlier: the post-download experience is not convincing users to try the paid product at all.
The benchmark tells you roughly where you sit; the funnel split tells you where to work. Treat the 2023 figures as a historical reference from that third-party report and confirm the current numbers in the latest published edition before you set a target.
Checklist before you change trial length or pricing
- Confirm the metric definition has not changed since the last measurement period.
- Split the funnel: trial starts, trial conversion and download-to-paid conversion are separate numbers.
- Decide in advance whether you will judge the test on conversion rate alone or on realized LTV as well.
- Segment the result by country and subscription duration before generalizing.
- Note the reference date of any external benchmark you compare against.
- Check whether the change affects onboarding, the paywall or both, since those are different levers.
Putting the five metrics together
Each metric in this guide describes a different part of the same relationship. Active-user counts show whether the habit exists. Session metrics show whether the habit is worthwhile. Conversion shows whether users reach paid status. Churn shows whether they stay. Realized LTV shows what the relationship is worth over time.
Read together, they answer two questions that matter more than any single number: is the app useful, and is it valuable? Where the metrics disagree — strong usage but weak conversion, or healthy conversion with high churn — that gap is usually the most informative place to start investigating.
| Question you are asking | Metric to check | What a change usually means |
|---|---|---|
| Are people coming back at all? | DAU / MAU | Shifts in habit strength or in your definition of “active” |
| Is the time spent valuable? | Session length and depth | Changes in feature appeal, content fit or interface friction |
| Are users reaching paid status? | Trial-start, trial conversion or download-to-paid conversion | Paywall, onboarding or value-demonstration effects |
| Are subscribers staying? | Subscriber churn rate | Retention, pricing fit or segment-specific weakness |
| Is the relationship worth money over time? | Realized LTV per customer | Long-term monetization and pricing impact |
Descriptive reading aid; not a scoring system or a set of universal thresholds.
Frequently asked questions
What does the DAU/MAU ratio tell you?
It compares how many users are active on a given day with how many are active across a month, giving a rough sense of how habitual the app is. The number is only as meaningful as your definition of “active,” so fix that definition before comparing periods or cohorts.
Does a longer trial always convert better?
No. Shorter trials can convert at lower rates than longer ones, but the ideal length depends on your app's context and on how long a user needs to reach the value moment. Test trial length against your own funnel rather than adopting a default.
What is the difference between new-user churn and subscriber churn?
New-user churn is the loss of app users overall. Subscriber churn is specifically subscribers becoming non-subscribers. Onboarding mainly addresses the first; retention and pricing fit mainly address the second.
How is realized LTV different from ARPU?
Realized LTV per customer divides a cohort's revenue minus refunds by the number of customers in that cohort, following the cohort across its whole relationship with the app. ARPU measures revenue over a specific period instead.
Is there a standard “good” engagement rate for apps?
No single threshold applies across app types. Usage rhythm, your own definition of engagement and category benchmarks all change the answer, so compare your app against its own history and against similar apps using a consistent definition.