Blog

DAU, MAU and Stickiness: How to Calculate Them

Lakshith Dinesh

Lakshith Dinesh

Head of Growth, Linkrunner

Last updated on 7 min read

DAU, MAU and Stickiness: How to Calculate Them

Short answer: DAU is the number of unique users active in one day, MAU the number of unique users active in a month (a calendar month or the last 30 days), and stickiness is average DAU divided by MAU. A 20 per cent ratio means the typical monthly user shows up on about 6 days in 30. Public 2026 data puts the median app at around 14 per cent, but your own trend, on a fixed definition of "active" and split by acquisition channel, matters more.

Every app team reports DAU and MAU. Far fewer can say what their own number counts. Two dashboards in the same company often show different DAU for the same day, and the gap is rarely a bug: one tool counts any app open, another counts users who did something, a third uses a different day boundary.

This post covers the definitions, the formulas with a worked example, 2026 benchmarks, the mistakes that inflate the count, and why DAU only becomes useful once you split it by acquisition channel.

What are DAU, WAU and MAU?

Active users are unique people (or accounts) who did something you define as "active" within a time window. The window is the only thing that changes between DAU, WAU and MAU.

  • Daily active users (DAU): unique users active on a given day.
  • Weekly active users (WAU): unique users active in a 7-day window. Some reports label it "7-day active users".
  • Monthly active users (MAU): unique users active in a calendar month or a trailing 30 days.
  • Stickiness (DAU/MAU ratio): the share of your monthly users who are active on an average day.

Window conventions differ by tool, so write yours down:

  • GA4 defines DAU as active users in the last 24 hours, WAU as the last 7 days and MAU as the last 30 days, and calculates DAU/MAU, DAU/WAU and WAU/MAU for you, per its user stickiness help page.
  • Calendar month MAU suits finance and month-on-month reporting; trailing windows suit daily trend charts.

What counts as "active"?

"Active" is a definition you choose, and it moves your DAU more than anything else on this page.

There are two broad options:

  • Lenient: any app open. Easy to collect and closer to most public benchmarks, but it counts people who tapped a notification and left two seconds later.
  • Strict: a core action. A message sent, a round played, a trade placed. Lower numbers that track the value people actually get.

GA4 sits in between: Google's metrics reference counts users with an engaged session (10 seconds or more, a key event, or two screen views) or a logged first_open or engagement event.

Report the lenient number for comparability, run the business on the strict one, and restate past months whenever the definition changes.

How do you calculate DAU, MAU and stickiness?

Count unique users per window, then divide average DAU by MAU for the same period.

The formulas:

  • DAU = count of unique user IDs active on day d
  • Average DAU = sum of daily DAU across the period ÷ number of days
  • WAU and MAU = count of unique user IDs active in the 7-day or monthly window
  • Stickiness = average DAU ÷ MAU
  • DAU/WAU = average DAU ÷ average WAU
  • WAU/MAU = average WAU ÷ MAU

Worked example

Take an app over a 30-day month:

  • Daily DAU, added up over the 30 days, comes to 360k. Average DAU is 360k ÷ 30 = 12k.
  • 60k unique users opened the app at least once in the month. MAU is 60k.
  • Stickiness is 12k ÷ 60k = 20 per cent. The average monthly user was active on about 0.2 × 30 = 6 days.
  • Average WAU across the month was 30k. DAU/WAU is 12k ÷ 30k = 40 per cent (about 3 days a week for weekly users), and WAU/MAU is 30k ÷ 60k = 50 per cent.

Two shortcuts break this maths: adding up DAU or WAU to get MAU (a user active on 20 days is one MAU, not 20), and dividing a single peak day's DAU by MAU instead of the period average.

The same maths answers the "50k DAU" question in reverse: at 20 per cent stickiness, 50k average DAU implies around 250k MAU; at 10 per cent, around 500k.

What does a good DAU/MAU ratio look like?

In 2026 public data, the median app sits at around 14 per cent, the top 10 per cent at around 31 per cent, and the spread by category is wide.

The most detailed recent source is MWM's DAU/MAU benchmark, built from US iOS apps with at least 1k downloads in 30 days and at least 100 daily active users (data labelled April and May 2026):

Category (US iOS, 2026)Median stickinessTop 10%
All appsaround 14%around 31%
Social and communicationaround 21%around 41%
Productivity and toolsaround 18%around 39%
Lifestyle and well-beingaround 15%around 30%
Media and entertainmentaround 15%around 30%
Education and knowledgearound 13%around 25%
Gamesaround 12%around 24%

Now compare games with Mixpanel's 2026 mobile gaming benchmarks, published in March 2026, which put gaming stickiness in APAC and North America at around 32 per cent. That is nearly three times MWM's games median.

Neither is wrong. They measure different samples and likely different definitions of "active". Both are third-party sources, and MWM covers US iOS only, so Android-heavy apps in India should treat these figures as a sense check, not a target.

How to set your own baseline instead

  1. Fix the definition. One "active" event, one user ID, one time zone, one MAU window.
  2. Backfill six months of average DAU, MAU and stickiness on that definition.
  3. Split by platform and acquisition channel (see below).
  4. Set a band, not a point. Your trailing three-month range is normal; a move outside it is worth investigating.
  5. Pair it with retention. Stickiness says how often monthly users return; retention rate says how many new users are still there. See a good Day 7 retention rate and a good Day 30 retention rate.

What does stickiness tell you (and what does it miss)?

Stickiness measures habit: how much of your monthly base comes back on a typical day. It does not measure growth, value or revenue.

  • Natural ceilings differ. A messaging app expects daily use; a travel app is opened a few times a year, so a low ratio is normal there. For weekly-use apps, WAU/MAU is the better habit metric.
  • Fast growth pulls it down. A big install push adds users to MAU who have not formed a habit yet.
  • Shrinking can push it up. If casual users churn, the ratio rises while the business gets smaller. Always read it next to MAU.

What are the most common DAU counting mistakes?

Most inflated DAU comes from counting the wrong thing, counting the same person twice, or counting things that are not people.

  • Counting opens, not actions. Push taps and accidental opens inflate lenient DAU. Check how many "active" users never reach your core action.
  • Bots, emulators and test devices. QA devices and automated test runs can appear as users. Exclude internal device IDs and test builds.
  • Reinstalls and new devices. A reinstall or new phone can mean a new device ID, so counting by device rather than logged-in user ID makes one person two. Relatedly, our reattribution benchmark found around 3 per cent of installs in mid-2026 were returning users credited to a fresh ad touch, not new people.
  • Day boundaries. A UTC day and an IST day split users differently. Use one time zone across every tool.
  • Mixing tools. GA4's trailing 30 days and another tool's calendar month rarely agree. In Linkrunner support, when a customer's ROAS differs between Linkrunner and an ad network, the cause is usually a definition gap (window length, cohort versus in-period counting) before a setup fault. Active user counts behave the same way: agree the definition before you debug the data.

Why should you split DAU by acquisition channel?

Blended DAU hides which spend built your active base. Splitting it by the channel and campaign that acquired each user shows which installs became habits.

An illustrative split of the worked example above:

Acquisition sourceMAUAverage DAUStickiness
Organic25k6k24%
Meta20k3k15%
Google10k2k20%
Referral5k1k20%
Blended60k12k20%

The blend reads as a healthy 20 per cent. Underneath, a third of MAU comes from one paid channel whose users show up far less often. If that channel is also the cheapest per install, CPI alone would push budget the wrong way.

What to do with the split:

How to set this up with an MMP

Your analytics tool counts DAU; your mobile measurement partner knows which campaign earned each install. Join the two.

  • Linkrunner is a mobile and web attribution platform that attributes each install to its network and campaign. Its Mixpanel integration writes the campaign (lr_campaign) and ad network (lr_ad_network) onto each user's profile, so DAU and MAU in Mixpanel can be broken down by acquisition source. The GA4 integration sends each install's media source and campaign ID into GA4.
  • Retention per campaign: Linkrunner reports classic and cumulative retention at Day 1, 7, 14 and 30 per campaign, with rolling retention to Day 7, as described in the retention documentation. Cohorts start at install, and cohort revenue and ROAS run to Day 180.
  • Cost and setup: 25k attributed installs free, one time, then ₹1.20 to ₹0.80 per install by monthly volume. Most teams integrate the same day and are fully live within a week.

FAQ

What is a good DAU/MAU ratio?

There is no single good number. In MWM's 2026 US iOS data the median app is around 14 per cent and the top 10 per cent around 31 per cent, with social apps higher and games lower. Compare against your own trailing months on a fixed definition first.

Should MAU be a calendar month or the last 30 days?

Either works if you are consistent. GA4 uses the last 30 days. Calendar months suit finance and month-on-month reporting. Never compare a trailing-30-day figure from one tool with a calendar-month figure from another.

Why is my DAU different in two tools?

Usually because the definitions differ: what counts as active, the day's time zone, device ID versus user ID, and whether test devices are excluded. Align those before treating the gap as a tracking fault.

The takeaway

Pick one definition of "active", calculate stickiness as average DAU over MAU for the same period, and judge it against your own six-month band before any public benchmark. Then split it by acquiring campaign, which is where DAU turns from a vanity number into a budget decision. If you want campaign-level attribution flowing into the tool where you count DAU, request a demo from Linkrunner, then run the five-step baseline above on your last six months.

Start measuring the installs your team cares about

Bring attribution, deep links, SKAN, cohorts, and campaign intelligence into one workflow your growth team can trust.