“We have two lakh monthly active users” sounds like a fact. It is really a definition followed by a count, and the definition does most of the work. Change what counts as “active” and the same app, on the same day, can report numbers that differ several times over. Before discussing DAU, MAU or stickiness, it is worth being precise about the word underneath them.

The counts themselves

DAU (daily active users) is the number of distinct users who were active on a given day. WAU and MAU are the same idea over a week or a month. The word “distinct” matters: someone who opens the app eight times on Tuesday counts once in Tuesday’s DAU, and someone active on 20 days in November counts once in November’s MAU.

A month can mean a calendar month or a rolling 28- or 30-day window ending on each day. Rolling windows give a smooth daily series and avoid the awkwardness of February being shorter than March; calendar months line up with financial reporting. Neither is wrong, but mixing them in one report is a common source of mismatched numbers.

Choosing the action that defines “active”

The candidate definitions usually form a ladder from easy to meaningful:

  1. Opened the app or loaded a page (including background refreshes, if you are careless).
  2. Logged in.
  3. Performed a core action: searched, viewed a product, read an article.
  4. Completed a value-creating action: placed an order, made a payment, sent a message, completed a lesson.

The lower rungs inflate numbers and can be moved by things that do not matter, such as a push notification that tricks people into opening the app and closing it. The higher rungs track genuine value. A sensible rule is to define “active” by the action that, if users stopped doing it, the business would actually suffer.

Context changes the right answer. For a bank, “logged in during the last 12 months” is far too generous: a dormant savings account whose holder checked the balance once in January is hardly an active relationship. Something like “at least one customer-initiated transaction in the last 90 days” is closer to the truth, and regulators and internal policies sometimes define dormancy separately, which your analytics definition should not contradict. For an e-commerce marketplace, “placed an order in the last 90 or 180 days” is common because purchase cycles are longer than a month. For a messaging app, daily usage is the natural rhythm.

Also decide who is excluded: internal staff and test accounts, bots and scrapers, users who only hit an error page, and users of a deprecated version that can no longer transact. These exclusions should be part of the written definition, not ad-hoc filters that each analyst remembers differently.

Stickiness: what DAU/MAU actually measures

The ratio of average DAU to MAU is called stickiness. If an app averages 60,000 DAU across November and had 2,00,000 MAU in November, stickiness is 30%.

There is a neat interpretation hiding here. Add up DAU across every day of the month and you get the total number of user-days of activity. Divide by the number of days to get average DAU, then by MAU, and what you have is the average number of active days per monthly user, divided by the days in the month. So 30% stickiness in a 30-day month means the typical monthly user was active on about 9 days. That reading is far more intuitive for stakeholders than a percentage.

The same average can come from very different user bases, though. Nine active days on average might mean everyone uses the app about twice a week. It might equally mean 20% of users are active every single day while the other 80% show up for under four days each. Those are different products with different risks. Look at the distribution of active days per user (sometimes called an L-curve or power-user curve) rather than relying on the ratio alone.

When stickiness is the wrong lens

Not every product should be used daily. A travel booking app, an insurance app or a tax-filing service may be excellent and still have very low DAU/MAU, because people need them a few times a year. Judging such products on daily stickiness pushes teams towards notifications and gimmicks that annoy customers. For them, the right lens is whether users return when the need arises: repeat booking rate, renewal rate, or the share of customers who come back within their natural cycle.

Similarly, rising MAU can mask decay. If marketing brings in many new users who try the app once, MAU climbs while stickiness falls. Splitting MAU into new, retained and resurrected users (returning after a gap) shows whether growth is coming from people who stay or from people passing through.

When active users drop

A drop in weekly actives has a standard checklist. First rule out measurement: a tracking SDK change, an app release that broke an event, a change in bot filtering, or a data pipeline delay. Then check calendar effects: a festival week, exam season, a long weekend, or simply comparing a week with a public holiday against one without. Only then decompose by platform, app version, city, acquisition channel and user type. A drop concentrated in one Android version points to a bug; a drop spread evenly across everything points to something external or seasonal.

Whatever you choose, write the definition down, give it a name and an owner, and use it everywhere. A single agreed definition of “active” saves more meetings than almost any dashboard.

Written by the RP Strativa Team. Figures in the examples are illustrative.

Leave a Reply

Trending