What are DAU and MAU?

DAU (Daily Active Users) is the count of unique users who fired a qualifying action in a calendar day. MAU (Monthly Active Users) is the same over a calendar month or a trailing 30 days (the period must be stated). WAU (Weekly Active Users) uses a 7-day window. The three metrics describe a product’s active audience — usually a mobile or web app — not an ad account.

“Active” is not “installed” and not “a process woke in the background.” The team picks an event: session start, a key action (message, bet, order), sometimes any analytics hit. Different thresholds produce incomparable DAU. Changing the threshold changes the metric; it is not product growth.

Formulas

DAU = unique user_id (or device_id) with ≥1 qualifying event that day. MAU = uniques in the month. One person opening the app 20 times in a day is still one DAU. The sum of daily DAU over a month is not MAU: the same people appear on multiple days.

WAU uses the same uniqueness rule on a 7-day window. WAU helps when a month is too coarse (a new feature, weekly seasonality) and a day is too noisy.

The DAU/MAU ratio (stickiness)

DAU ÷ MAU is called stickiness or return frequency. The range is 0…1 (sometimes shown as a percent). If DAU = 80,000 and MAU = 400,000, the ratio is 0.20: a typical active user shows up on about 20% of days in the month (a rough ~6 of 30 if activity is even — a calendar simplification, not an identity).

Rough category ranges, not norms:

  • messaging and social — often 0.4–0.6 and higher;
  • utilities, content, e-commerce — often 0.10–0.25;
  • games vary by genre: mid-core usually stickier than hyper-casual after week one.

Rising MAU with a falling ratio means more new users who come back less often. A rising ratio with flat MAU means the same audience got denser, not necessarily larger. For UA those are different calls: buy more installs vs fix retention.

Relation to retention and cohorts

DAU/MAU are stock snapshots on a date. Retention (D1, D7, D30) is a cohort share of users acquired on day T who return later. You can post high DAU from fresh CPI traffic with weak D7: “active today” are yesterday’s installs, not habit. The reverse also happens: a stable product with strong D30 and modest DAU growth.

DAU does not replace cohort reports or LTV. LTV and ARPU say how much an active user is worth; DAU/MAU say how many such users exist and how often they appear. Subscriptions and gambling also split paying DAU/MAU (or payer DAU), otherwise “activity” mixes whales with people who only opened the app.

Apps and paid UA

In mobile UA, DAU and MAU are the frame for CPI, ROAS, and payback. A campaign can look fine on day-0 install cost and still inflate MAU with low-quality installs: next-day DAU does not hold, the ratio drops. In-app and rewarded inventory often buys cheap CPI and weak stickiness if the product does not retain.

Pushes and lifecycle campaigns lift DAU without creating new users: the same MAU show up more often. That is neither good nor bad by itself — the question is whether revenue per active rises or only the session counter. In-app inventory itself is measured by eCPM and fill; DAU/MAU belong to the product that shows that inventory or that you are promoting.

How comparisons break

  • Calendar month vs rolling 30. February is short; rolling is cleaner for charts, worse for accounting.
  • User vs device. Reinstalls and multiple devices inflate MAU without a login.
  • Bots and prefetch. Unfiltered “uniques” include crawlers and SDK noise.
  • Time zones. The product’s “day” is not the ad account’s day.

In arbitrage, DAU/MAU matter when the offer is an app you can see in product analytics (white-label, in-house app, MMP report). On a classic CPA landing with no app, the metric barely applies: you watch CR, approve, and repeat conversions — not “monthly active users of a website” in the product sense.

See also: retention rate, churn rate, LTV, ARPU, CPI, in-app ads.