App engagement metrics: formulas, benchmarks by app type and a worked case

A spike in daily actives can mean more people found value, or that a push notification got more people to tap and leave. This guide lists the app engagement metrics with formulas, explains how to judge them by app type, and walks through a notification change where the headline numbers and the real engagement disagreed.

What are app engagement metrics?

App engagement metrics measure how often people open a mobile app, how long and how deeply they use it, and whether they keep coming back. Downloads say how many people tried the app; engagement says whether it became part of their routine. They are the leading indicators for retention and revenue: an app whose engagement is falling will see churn a few weeks later.

The core app engagement metrics

Two quality metrics belong next to them, because crashes and freezes destroy engagement: crash-free users (1 − users with a crash ÷ active users) and the ANR or freeze rate. Google Play treats a user-perceived crash rate above 1.09% or an ANR rate above 0.47% as bad behavior that can reduce an app's visibility.

Judge against the natural frequency of your app type.

What "good" engagement looks like

Benchmarks depend heavily on category. A messaging app should be opened several times a day; a travel app a few times a year. Rather than chase an industry number, set the expected frequency for your app's core job:

Worked example: more opens, not more use

A fitness app added daily streak reminder notifications. Thirty days before and after (sample figures):

The headline numbers (DAU +20%, stickiness up) suggested a big win. The engagement detail told a different story: many of the extra opens were people tapping the notification, closing the streak, and leaving. Engaged sessions per active user rose only 8%, sessions got shorter, retention did not move, and more than twice as many users turned notifications off, which removes the channel for everything else.

The team kept the reminder but sent it only to users who had worked out at least twice in the previous week, and changed the copy to suggest a specific workout. Opt-outs fell back to 3.8% while engaged sessions per DAU held at 1.21.

How to improve app engagement

1. Shorten time to value in onboarding: get new users to the core action in the first session.
2. Find the habit-forming action. Compare retained and churned users: which action in week 1 separates them?
3. Use notifications sparingly and personally, triggered by behavior, not by the calendar.
4. Fix stability first. A crash in the first session is often the last session.
5. Remove friction in the core loop, using screen paths and drop-offs to find where users leave.
6. Measure engaged sessions, not starts, so a notification campaign cannot inflate the metric.

Measuring app engagement in TapCub

App analytics › Users › Users shows new users, active users (daily, weekly and monthly), starts and one-time starters, split by version and channel. App analytics › Users › Engagement shows the distributions of session duration, starts per user, interval between starts, screens per session and starts per day, plus a weekday-by-hour grid. In TapCub an engaged session lasts at least 10 seconds, includes at least two screens, or includes a key event. Retention is under Users › Retention, and crash-free users and freezes under Quality & attribution › Stability (Crash-free rate, Track app versions). Related reading: Mobile app KPIs and Engagement metrics.

See also: Session timeout.

See also: Increase app engagement, Mobile app analytics.

Sources

• Android Developers, Android vitals: user-perceived crash and ANR rate thresholds.