Product Engagement metric

DAU/MAU. A near-useless number in aggregate — and a sharp churn signal at the account level.

DAU/MAU is the classic stickiness ratio: daily active users divided by monthly active users, the share of your monthly base that shows up daily. The consumer-app world treats it as gospel, but I'll give you the honest operator take after sixteen years running a daily-use CRM: in aggregate, it's close to a vanity metric. If stickiness across 3,000 accounts ticks up or down 2% week to week, what do you actually do with that? There's no action you can take in aggregate. The signal that matters lives one level down — at the individual account, where a falling login pattern tells you exactly which customer is drifting toward churn while there's still time to do something about it.

What it is

Daily active users divided by monthly active users — what percent of your monthly base shows up on a given day. Only meaningful if your product is daily-by-design. For a daily-use tool it's a real signal; for a weekly or monthly product it just punishes you for being infrequent on purpose.

Measurement period

Account level.

The aggregate ratio is a vanity number; the value is per-account. A single account's login pattern falling off is an actionable churn signal in a way the company-wide ratio never is.

Formula

Daily active users

Monthly active users

× 100

A 20% ratio is the consumer benchmark. For a daily-use B2B tool you'd expect far higher — and you should read it per account, not company-wide.

When to review

Weekly, per account.

Watch login patterns at the account level, weekly, so customer success can intercept a drifting account the moment it slips — not the aggregate trend, which tells you nothing you can act on.

Why it matters

What do you do when 3,000 accounts tick up 2%?

Let me give you the contrarian version, because I lived it. PipelineCRM was a genuinely daily-use product — reps lived in it every workday — so stickiness was a legitimate signal for us, not a vanity number to wave at. And yet I'm still skeptical of the aggregate ratio, because I could never answer the basic question: if usage across 3,000 accounts ticks up or down by 2% in a week, what action does that let me take? Nothing. There's no lever you pull in aggregate. A company-wide stickiness number that drifts a couple points week to week teaches you almost nothing and changes no decision.

Where the same data becomes genuinely valuable is at the account level. A daily-use product gives you a gift: when a specific account's daily logins start falling off, you're watching disengagement happen before it shows up as churn. That falling login pattern needs to be reviewed at the account level, the moment it happens, so customer success can intercept and intervene. I'll be honest about our own miss here — we had some signals and a churn-prediction model, but it came too late in our business cycle, and our early-warning system was never as good as it should have been. The lesson isn't "ignore stickiness." It's "stop reading it in aggregate and start reading it one account at a time, in time to act."

And there's a real dollar consequence underneath the engagement data. Take a twenty-seat account where four people are daily and sixteen never log in. That's not a stable account — it's a downgrade or a churn waiting to happen. At some point the buyer or admin notices the sixteen dead seats and either trims them (contraction MRR) or questions the whole contract (a churned account). Neither is good. Seat-level utilization inside an account is a real input to churn risk, which is exactly why the aggregate ratio hides what matters: it averages the healthy accounts and the hollow ones into a single number that describes neither.

If usage across 3,000 accounts ticks up 2% in a week, what action does that let you take? Nothing — there's no lever you pull in aggregate. The signal that matters lives at the account level, where a falling login pattern is a churn warning you can still act on.

Worked example

Same healthy-looking aggregate. Three accounts telling different stories.

A company-wide stickiness of, say, 55% looks fine on the dashboard. But it's an average hiding three completely different accounts underneath. The aggregate would have you celebrating; the account-level view tells you which customer to call today.

Embedded

18/20

Seats20
Daily users18
Login trendSteady
ReadWoven in — safe

Nearly every seat is in the product daily. This account has woven the tool into how the team works — the stickiest, safest kind of customer, and the one whose renewal you barely have to think about.

Hollow

4/20

Seats20
Daily users4
Login trendFlat-low
ReadDowngrade risk — 16 dead seats

Four power users, sixteen seats nobody touches. The admin will eventually notice and trim them — that's contraction MRR coming. A success play to activate those seats now is worth real money later.

Slipping

12 → 5

Seats20
Daily users12, falling to 5
Login trendDropping fast
ReadChurn signal — intervene now

Was healthy, now the daily logins are falling off a cliff. This is disengagement happening in real time, before it's churn. The account a CS person should be calling this week — the whole point of watching logins per account.

Benchmarks

First decide if the metric even applies to your product.

Unlike most metrics here, the first question isn't "what's a good number" — it's "should I be measuring this at all?" The bands below are the usage-frequency reality check. Stickiness is only a health signal if your product is meant to be used daily; for weekly or monthly tools, a low ratio is the design, not a problem, and feature adoption or activation is the better engagement read.

Daily-by-design Use it (per account)

A genuinely daily-use product — a CRM reps live in, a comms tool, a daily workflow app. Here stickiness is a real signal, and a healthy daily-use B2B account runs far above the consumer 20%, often 50–70%+ of seats active daily. Watch it per account as an early churn warning.

Frequent Useful, with context

A product used several times a week but not strictly daily. Stickiness still carries information, but read trends rather than absolute levels, and pair it with feature adoption — a lower ratio here is expected and isn't cause for alarm on its own.

Weekly/monthly Use a better metric

A tool used on a weekly or monthly rhythm by design — reporting, planning, billing. A low DAU/MAU here is correct, not broken, and chasing it would mean optimizing for the wrong behavior. Reach for feature adoption or activation rate instead; they'll tell you what stickiness can't.

Vanity trap The aggregate number

However daily your product is, the company-wide ratio is the trap. A 2% move across thousands of accounts is unactionable — you can't pull a lever on an aggregate. If the only DAU/MAU you look at is the blended one on a slide, you've got a vanity metric, regardless of how sticky the product actually is.

When stickiness is slipping

Three plays that actually move it.

There's nothing to do with a sliding aggregate — so every play here pushes the same direction: down to the account, where action is possible. Confirm the metric applies, watch logins per account, and intervene on the specific customers slipping.

— 01 First, confirm the metric fits your product

If you're not daily-by-design, use a different metric.

Before optimizing stickiness, ask whether your product is even supposed to be daily. A CRM reps live in, yes. A quarterly planning tool, no — and a low ratio there is the design working as intended. If your usage is naturally weekly or monthly, DAU/MAU will only punish you for being infrequent, and feature adoption or activation rate is the engagement metric that actually tells you something. Don't measure stickiness just because the consumer playbook says to.

— 02 Watch logins at the account level, weekly

The falling pattern is a churn alarm — wire it up.

For a daily-use product, a specific account's daily logins falling off is disengagement you can see before it becomes churn. The play is to surface that the moment it happens, account by account, so customer success can intercept. Our own miss was leaning on an aggregate churn model that fired too late; the better system is an account-level login alert that trips while there's still time to call. Build the early-warning system around the account, not the company average.

— 03 Chase seat utilization inside accounts

Dead seats are contraction MRR waiting to happen.

The four-of-twenty account isn't healthy just because four people love the product. Sixteen dead seats are a downgrade or a churn the admin hasn't gotten around to yet. The play is to drive activation on those dormant seats — onboarding, a champion, a reason to log in — before the renewal conversation, because seat utilization is a real input to churn risk and the cheapest contraction to prevent is the one you head off early.

Common mistakes operators make with DAU/MAU.

Treating the aggregate ratio as actionable.

The big one. A company-wide stickiness number that moves 2% week to week tells you nothing you can act on — there's no lever you pull across 3,000 accounts at once. Aggregate DAU/MAU is a board-slide vanity metric. The same data is genuinely valuable one account at a time, where you can actually intervene. If the only version you look at is blended, you've turned a useful signal into a number that changes no decision.

Applying the consumer 20% benchmark to B2B.

The 20% "good stickiness" figure comes from consumer apps fighting for attention against everything else on the phone. A daily-use B2B tool that reps work in every day should run far higher — and a B2B product that's weekly by design might sit well below 20% and be perfectly healthy. Importing the consumer benchmark either makes a healthy infrequent product look broken or sets a daily product's bar far too low. Judge against your product's intended rhythm, not a consumer number.

Measuring stickiness on a product that isn't daily-by-design.

If your tool is meant to be used weekly or monthly, a low DAU/MAU isn't a problem to solve — it's the correct behavior. Chasing daily usage on a planning or reporting tool optimizes for the wrong thing and annoys customers. Understand your product's natural usage frequency first; if it isn't daily, feature adoption or activation rate will tell you far more about engagement than a stickiness ratio ever will.

Ignoring seat utilization inside accounts.

A twenty-seat account with four daily users looks like an active account in any blended number, but it's sixteen dead seats away from a problem. Eventually the admin trims the dormant seats (contraction MRR) or questions the contract (churn). Per-account, per-seat utilization is a real input to churn risk — averaging it away into a company-wide ratio hides exactly the accounts you most need to act on.

Building the early-warning system too late.

A daily-use product hands you disengagement signals before churn — but only if you're watching account-level logins in real time. A churn-prediction model that fires after the account has already gone quiet is too little, too late; that was our own miss. The point of stickiness data is the head start it gives customer success, and you only capture that head start if the alert trips the week the logins start falling, not the quarter after.

Confusing high stickiness with breadth of value.

A high daily-login number tells you people show up; it doesn't tell you they're getting deep value or using the features that drive retention and expansion. Stickiness is a presence signal, not a value signal. Read it next to feature adoption — a customer who logs in daily but only touches one shallow feature is more fragile than the raw ratio suggests. Presence and value are related but not the same thing.

Read alongside

Stickiness says they show up. Adoption says it's worth it.

DAU/MAU is a presence signal — and for a non-daily product, often the wrong signal entirely. Feature Adoption tells you whether customers are using the things that actually drive retention, which is the engagement read that holds up no matter your product's usage rhythm.

Feature Adoption guide →

How Upbeat helps

The signal where it's actionable — per account, not in aggregate.

The aggregate stickiness number belongs nowhere near a decision; the account-level login trend belongs in front of your customer success team every week. Upbeat surfaces engagement at the account level on your scorecard, so a customer whose daily logins are falling off shows up as a name to call — not as a two-point wobble in a blended ratio nobody can act on. The data turned into an intervention list.

See how it works → See pricing

Related metrics

Presence is one signal. These tell the rest.

Stop reading it in aggregate. Read it one account at a time.

Upbeat surfaces account-level engagement on your scorecard — so a customer drifting toward churn shows up as a name your success team can call this week, instead of a two-point wobble in a blended ratio that changes no decision.

Email Nick directly →