Quick definition: Monthly active users (MAU) is the number of distinct users who perform a predefined meaningful activity during a specified monthly measurement window. Its value depends entirely on the activity, identity rule, eligibility population, and calendar definition.
What are monthly active users?
MAU is an active-user metric commonly used to describe the reach of a product over a month. A user is counted once if they meet the activity rule during the window, even if they take the action many times. For a messaging product, sending or reading a message may qualify; for a finance product, reconciling a transaction may be more meaningful than opening the app. A login, page view, or push-notification open can be useful diagnostic telemetry, but often makes a weak definition of active use.
“Monthly” needs a precise meaning. A calendar MAU counts activity from the first through final day of a named month, usually in a documented time zone. A rolling 28- or 30-day MAU counts a moving period ending on each reporting date. Calendar windows are simple for planning; rolling windows reduce boundary effects but overlap heavily from day to day. Never compare the two as though they were the same series.
MAU is related to daily active users, retention, and engagement, but it is not any of them. DAU captures daily reach; retention follows an earlier cohort into later periods; engagement can require frequency or depth. MAU measures whether a distinct identity crossed a chosen activity threshold within a month.
Defining and calculating MAU
The basic formula is MAU = count of distinct eligible user identities with at least one qualifying activity in the month. A rate can be more decision-useful: monthly active rate = MAU / distinct users eligible to use the product during the month. The denominator should fit the question. Registered users, paid seats, provisioned users, trial users, and users with access to the relevant feature are different populations.
| Definition choice | What to document | Example |
|---|---|---|
| Unit | Person, device, account, or workspace? | Authenticated person, merged across devices. |
| Qualifying activity | Value-linked server-confirmed action | Completed transaction reconciliation. |
| Window | Calendar or rolling dates and time zone | Calendar month, UTC. |
| Eligibility | Who could reasonably perform it? | Provisioned paid users with feature access. |
| Identity policy | How anonymous and merged IDs work | Deduplicate after verified account merge. |
Keep a versioned definition and a reproducible query. Changes to authentication, event names, bot filters, account mergers, consent collection, or activity semantics can move MAU without a real change in customer behavior. Backfill only when it is valid and transparent; otherwise annotate the break and preserve both series. Reconcile a sampled set of counted users with source-of-truth product records.
Report counts with composition. Break out new versus returning users, platform, plan, region, acquisition source, and account size when those distinctions explain a decision. Include the numerator and eligible population rather than presenting a growth percentage alone. A rise in MAU after a free campaign can be valuable, neutral, or harmful depending on activation, cost, and product capacity.
Using MAU in a product framework
MAU is usually a reach or health indicator, not a complete north-star metric. Pair it with a value-quality measure, such as completed workflows, successful outcomes, error-free use, or customer satisfaction. Pair it with retention to distinguish a growing top of funnel from durable adoption. Pair it with monetization or capacity measures where use has cost. The appropriate companion metrics reflect how the product creates value.
The DAU/MAU ratio is sometimes called stickiness: average DAU during the month / MAU. It approximates the share of monthly active identities active on a typical day. It is useful for products expected to be used frequently, but it can mislead for monthly or event-driven products. A tax tool or payroll system can have excellent customer value and low daily frequency. Do not treat a ratio as a generic quality score.
Cohort views are essential. Compare people by first eligibility, activation, or acquisition month, then examine what share becomes active in later months. This separates a larger new-user inflow from improved continued use. Retention definitions should be aligned with the MAU activity rule when the question is ongoing value; otherwise a person can be “retained” based on a shallow event while MAU reflects a deeper action.
Experiment scenario: reactivation reminders
A budgeting product finds that many previously active users have not completed a monthly review by the third week of the month. It hypothesizes that a personalized reminder, sent only to eligible users with verified notification consent, will help them return to complete their intended review. The team randomizes eligible users before scheduling the reminder and excludes people who have already completed the review before the send decision.
The primary outcome is the share of all assigned eligible users who complete a server-confirmed monthly review by month end. It does not use MAU alone as the primary metric, because a generic open could increase active status without completing the job. MAU is a secondary reach measure. Guardrails include unsubscribe rate, notification complaints, help requests, app errors, duplicate reminders, and later satisfaction. The analysis records delivery failures and uses assigned treatment rather than only delivered notifications.
If the reminder increases MAU but not completed reviews, it may generate superficial opens. If it raises reviews but also causes a substantial opt-out increase, the net benefit is uncertain. Report the absolute effect, eligible denominator, uncertainty, and pre-planned duration. A month-end metric inherently needs time to mature; avoid declaring a winner based on partial-window snapshots.
Interpretation and data limitations
Identity fragmentation is a central problem. One person can use multiple devices, reset a browser, share an account, or remain anonymous until conversion. Merging identities may reduce MAU; splitting identities may inflate it. Decide whether the product question concerns people, devices, or accounts, and document the merge logic. Do not compare an unauthenticated web MAU directly with an authenticated application MAU without explaining the difference.
Opportunity and cadence also matter. A user may be inactive because they lack data, permissions, teammates, or a reason to return that month. Seasonality, holidays, billing cycles, and calendar length change observed activity. Late-arriving events can revise a recently closed period. Bots, employee traffic, automated integrations, and background application events can create activity that is not customer value.
MAU is not causal evidence. A product release, marketing campaign, or pricing change can coincide with movement for many reasons. When assessing a change, use a valid comparison group where possible and avoid defining the analysis population using post-treatment activity.
Common mistakes
- Counting logins as meaningful use: choose an action close to delivered value.
- Leaving “month” undefined: publish dates, time zone, and window type.
- Reporting a count without opportunity: show an eligible denominator and composition.
- Ignoring identity changes: version merge logic and annotate trend breaks.
- Using DAU/MAU as a universal target: match expected cadence to the product job.
- Calling a trend a treatment effect: use experiments or an appropriate causal design.
FAQ
What counts as an active user for MAU?
A predefined qualifying action taken by a distinct eligible identity during the measurement window. The correct action depends on the product’s value.
Is MAU a count or a percentage?
Usually a count, but an active rate can be more comparable across eligible populations. Report both when possible.
Can a user be counted twice?
Not within a MAU calculation after the identity policy is applied, though poor deduplication can cause the same person to appear as multiple identities.
Why did MAU change after an analytics release?
The activity event, identity merge logic, filtering, consent population, or time-zone handling may have changed. Audit definitions before concluding behavior changed.
Summary
MAU counts distinct eligible identities that complete a defined meaningful activity in a monthly window. Make the action, unit, identity policy, eligibility, and calendar explicit; interpret the count with cohorts, quality, and opportunity. Use it as one component of product health, and use controlled comparisons for claims about a change’s effect.