User engagement metrics measure how much, how often and how deeply people use your product. They sit between activation (did people get started?) and retention (did they stay?), and they're usually where trouble shows first: engagement fades weeks before a customer cancels.
This guide explains what user engagement metrics mean, the core metrics for web products and apps, how to calculate them, how they differ from the "engagement" metrics in Google Analytics, and how to build an engagement view your team will use.
What do user engagement metrics mean?
Engagement isn't one number. It has three dimensions, and good metrics cover each:
- Frequency: how often people come back. Daily, weekly or monthly active users; days active per month.
- Depth: how much of the product they use. Features used per user; core actions per session.
- Breadth: how widely use spreads. Active users per account; invitations sent; work shared.
A product can score high on one and low on another. A tool opened daily for one small task has high frequency and low depth; a monthly reporting tool can have low frequency and great depth. Knowing which dimension matters for your product is the first decision. See engagement rate and customer engagement score.
Core user engagement metrics
| Metric | Measures | How to calculate |
|---|---|---|
| DAU / WAU / MAU | Frequency | Unique users active in the day, week or month |
| Stickiness (DAU/MAU) | Frequency | Average DAU ÷ MAU |
| Days active per month | Frequency | Distinct active days per user, averaged or as a distribution |
| Feature adoption rate | Depth | Users of a feature ÷ active users |
| Core actions per user | Depth | Count of your key event ÷ active users |
| Active users per account | Breadth | Unique active users grouped by account |
| Invitation rate | Breadth | Users who invited someone ÷ active users |
Active users
Count unique people who did something meaningful in the period. The definition of "active" matters more than the period: logging in to dismiss a notification shouldn't count the same as finishing real work. Our DAU, WAU and MAU guide covers choosing the period and definition.
Stickiness
DAU divided by MAU estimates what share of the month a typical user is active. Read it as a trend in your own product rather than a benchmark against others, since daily-use and monthly-use products naturally differ.
Days active per month
More revealing than an average is the distribution: the power-user curve, showing how many users were active on 1, 2, 3 and up to 30 days. A second hump on the right means a habitual core. See power users.
Feature adoption
For each important feature: the share of active users who used it in the period. Low adoption of a feature you invested in is either discovery or value, and a funnel into the feature tells you which. See feature adoption rate.
Breadth in accounts
For B2B products, count active users per account and the share of accounts with more than one active user. Single-user accounts depend on one person and churn more readily.
User engagement metrics for apps
Mobile and desktop apps add a few metrics worth tracking:
- Session frequency: sessions per user per week. Many apps track sessions as an event when the app comes to the foreground.
- Session length: time between session start and the last event. Useful for content and media apps, misleading for utility apps, where shorter can be better.
- Notification engagement: the share of notifications that lead to an open, and whether those opens lead to a core action.
- Version adoption: engagement by
app_version, to see whether a release improved or hurt behaviour.
The mobile app metrics guide goes into more detail.
User engagement metrics in Google Analytics
Google Analytics 4 reports "engaged sessions", "engagement rate" and "average engagement time". These describe website sessions: a session counts as engaged if it lasted long enough, had a conversion or viewed more than one page. They're useful for content and marketing pages.
They don't describe product engagement: whether an identified person used your product's features this week, or whether an account is spreading. For that you need events tied to users and accounts. The two complement each other. For the difference, see product analytics vs web analytics. And note that web engagement figures include automated browsers unless they're separated, which can distort averages; see bot traffic in GA4.
How to build an engagement view
- Pick your primary frequency metric. DAU, WAU or MAU, matched to your product's natural rhythm.
- Define "active" as a meaningful action, and write the definition next to the chart.
- Add one depth metric: adoption of the two or three features that matter most.
- Add one breadth metric if you sell to teams: active users per account.
- Show the power-user curve monthly.
- Break everything down by plan and cohort. Engagement that's healthy overall but falling in the newest cohort is an early warning.
- Set an alert on the primary metric so a drop reaches you the day it happens.
A simple engagement score
Some teams combine the dimensions into one score per user or account, so they can sort a list and act on it. Keep it simple and explainable:
- Frequency: days active in the last 28, scaled to 0–40 points.
- Depth: number of core features used, scaled to 0–40 points.
- Breadth: for accounts, active users beyond the first, up to 20 points.
An account active on 20 days, using three of four core features and with four active users scores roughly 29 + 30 + 20 = 79. Review the lowest-scoring paying accounts weekly, and the fastest-rising trial accounts daily. The exact weights matter less than keeping them stable so scores compare over time. When a score drops sharply, look at the account's timeline before reaching out: the reason is usually visible there.
Turning engagement into action
Engagement metrics earn their place when they trigger something:
- A drifting segment. Users or accounts whose activity fell compared with their usual level. Review weekly and reach out.
- Feature discovery. Users who never tried a feature that power users rely on. Show it to them in context.
- Expansion signals. Accounts with rising active users approaching their plan limits.
- Release checks. Engagement by cohort before and after a release.
Our guide to user behavior analytics covers the analysis behind these.
Common mistakes
- Counting logins as engagement. A login is the start of engagement, not proof of it.
- Averaging away the story. An average of six active days hides a split between people active on 1 and on 20.
- Tracking too many metrics. Seven engagement charts nobody reads are worse than two charts everyone reads and acts on every week.
- Ignoring accounts. User engagement can rise while the number of engaged accounts falls.
Measuring user engagement with LogZuno
LogZuno calculates engagement from your events: weekly and monthly active users, unique users and accounts for any event, the frequency chart, feature adoption with a breakdown, and built-in segments for engaged, power and drifting users. Put them on a dashboard, alert on drops with threshold alerts, and read more in product analytics.
Key takeaways
- Engagement has three dimensions: frequency, depth and breadth.
- Define "active" as meaningful action and match the period to your product's rhythm.
- Web engagement (GA4) and product engagement measure different things.
- Make engagement actionable with segments, alerts and per-cohort views.