A 20% DAU/MAU ratio means roughly one in five monthly users is active on a typical day, or about six days of use across a 30-day window when activity is evenly distributed. That's a useful reference point, but it isn't proof that users value the product. A meditation app and a payroll tool could reach the same ratio while serving completely different routines, needs, and levels of dependency. (Gainsight)
After running stickiness reports at three startups, I've learned the expensive lesson: DAU/MAU alone lies. It can count shallow opens, reward a small group of power users, and make a product look healthy while new users disappear after onboarding. The right question isn't how often people open the product. It's whether they return within a meaningful window, complete the action that creates value, and expand their use as their needs grow.
Table of Contents
- What Product Stickiness Measures
- Calculating the DAU/MAU Ratio
- Reading Retention Cohorts Over Time
- Going Beyond Daily Opens With Session Depth
- Building Your Stickiness Dashboard
- Combining Stickiness With Churn and NPS
- When Higher Stickiness Is the Wrong Goal
What Product Stickiness Measures
Product stickiness measures the strength of a user's return pattern. The foundational metric is DAU/MAU, calculated by dividing Daily Active Users by Monthly Active Users and expressing the result as a percentage. A high ratio can indicate repeated use, but it is only useful when “active” means an action connected to the product's value, rather than any visit or notification open.
The same ratio can describe very different products. A meditation product may generate brief daily sessions because users follow a routine. A payroll tool may attract visits when a specific operational task requires attention. Their similar return rates would not show whether users are satisfied, whether a workflow is embedded in team operations, or whether each visit produces meaningful progress.
A habit signal, not a satisfaction score
Define stickiness as the probability that a user who has shown up once returns within a window that fits the product's job. A daily workflow may warrant a day-based window. An asynchronous B2B tool may be better measured over a week or after completion of a meaningful task.
That distinction separates habit-forming stickiness from dependency-driven stickiness. Habit-forming users return because the product repeatedly helps them achieve an outcome. Dependency-driven users return because a required process, notification, approval, or reporting obligation gives them little practical choice. Both patterns raise usage, yet only the first clearly signals growing product value.
A dependency-driven product can hold a high ratio while subscription revenue stalls, because obligation-driven logins do not necessarily produce renewals. That is the mechanism worth examining when connecting repeat behavior with recurring revenue: users must keep receiving enough value to continue paying, not merely keep appearing in the activity log.
The three layers of an honest measurement system
Use three layers instead of treating one headline number as product health:
- Frequency: DAU/MAU, or a cadence-adjusted alternative, shows how often active users return.
- Durability: Cohort retention shows whether users continue returning after acquisition.
- Quality: Session depth, core-action completion, and feature breadth show whether activity represents value.
The ratio belongs at the top of the dashboard, not at the center of every roadmap decision. A rising score may reflect a useful routine, but it may also reflect a shrinking customer base dominated by power users. A falling score may indicate weak engagement, or it may fit a product whose value arrives through substantial weekly or monthly tasks. For irregular-use products, measure return against the job's natural cadence before deciding that lower daily activity is a problem.
Calculating the DAU/MAU Ratio
DAU/MAU is useful only when “active” means something specific:
DAU/MAU = unique daily active users ÷ unique monthly active users × 100
Use one activity definition and one time window for both values. A rigorous implementation counts unique daily active users and unique monthly active users across the same rolling 30-day window, then divides DAU by MAU. (Mixpanel)
Suppose a product has 12,400 DAU and 58,000 MAU. The result is approximately 21.4%. That figure demonstrates the calculation, not a transferable benchmark. The ratio describes return frequency inside the selected active population. It does not prove that users completed valuable work, retained, or paid.
The event definition can change the result substantially. A social application may count an authenticated session that includes a feed view. A B2B application should usually require a workflow action, such as creating, approving, editing, or exporting something. A marketplace might count a search, listing interaction, message, or transaction, depending on which behavior represents meaningful progress.
Practical rule: If a notification opens a screen but does not create or advance value, do not automatically count that user as active.
The definition controls the metric
Put the active definition beside the ratio in the dashboard. Analysts should be able to see whether the metric counts passive opens, meaningful events, or completed outcomes without searching an analytics specification.
| Product type | Active definition | DAU | MAU | DAU/MAU |
|---|---|---|---|---|
| Social app | Opens a session and views the main feed | 12,400 | 58,000 | 21.4% |
| B2B workflow tool | Completes a core workflow action, excluding passive opens | 8,900 | 58,000 | 15.3% |
| Marketplace | Sends a message or completes a transaction action | 6,700 | 41,000 | 16.3% |
These rows demonstrate why counts need context. The social app treats feed consumption as activity, while the B2B tool excludes an open without workflow progress. The marketplace uses a smaller monthly population and a narrower value event. A lower score after removing notification opens may be a measurement correction, not an engagement collapse.
For an app launch supported by backers, define activity around behavior that shows actual use rather than counting every campaign or notification visit. That distinction also helps teams report backer activity for a crowdfunding app launch without inflating demand.
DAU/MAU also misfits products with irregular usage. Payroll, compliance, procurement, and project-review workflows may be better evaluated with WAU/MAU, which follows their weekly cadence without penalizing users for skipping daily opens. (Statsig)
Choose the cadence that matches the job. Document the event, actor, qualifying conditions, identity rules, and time window. If the definition changes, mark the break in the series and reset the baseline. Otherwise, the team will compare incompatible ratios and mistake a reporting change for a product trend.
Reading Retention Cohorts Over Time
DAU/MAU tells you about frequency within an active population. Cohort retention tells you whether that population survives. Cohort analysis groups users by signup week or month and follows each group through its lifecycle, which prevents new acquisition from masking early churn. (Product Metrics)
Build the table from the user's first meaningful signup or activation event. Assign each user to a signup-week cohort, then check whether that cohort remains active on day 1, day 7, day 14, and day 30. “Active” should use the same value-based event you documented for the stickiness calculation.
What each checkpoint tells you
Day 1 is primarily an onboarding test. If users don't return after their initial session, look for setup friction, unclear positioning, missing data, or a weak first-use experience.
Day 7 tests the first habit loop. Users have had enough time to encounter a reminder, recurring need, collaboration moment, or product outcome. A weak day 7 result often means the product delivered a one-time answer but failed to create a reason to return.
Day 30 tests whether the product has earned a durable place in the user's routine. For some products, that means weekly workflow integration. For others, it means repeated task completion or continued consumption. The window matters more than the label.
| Signup week | Day 1 | Day 7 | Day 14 | Day 30 |
|---|---|---|---|---|
| Week A | ||||
| Week B | ||||
| Week C | ||||
| Week D |
Leave the cells blank until your event data is reliable. A clean-looking table built on inconsistent identity resolution is worse than no table because it invites false confidence.
Read the shape, not only the endpoint
A sharp cliff after day 1 points toward an onboarding or activation problem. A curve that declines and then develops a flat tail suggests the product has found a retention ceiling among users who reach value. A late uptick can indicate that a triggered re-engagement feature, collaboration event, or recurring workflow is bringing users back.
Low-volume products need wider cohorts. Aggregate by signup week rather than day when daily counts make the percentages unstable. Keep the window consistent across cohorts, and compare cohorts by acquisition channel, plan, persona, and first core action.
A cohort table beats a single ratio because it shows the decay rate. Two products can share the same current stickiness while one loses users immediately and the other retains a stable base. Those products need completely different fixes.
Going Beyond Daily Opens With Session Depth
A team dashboard can report 30,000 DAU against 90,000 MAU, producing a 33% DAU/MAU ratio, and still have a serious value problem. Those figures are a realistic analytical example, not a verified benchmark. The warning appears when revenue stalls and event-level analysis shows that 40% of daily users spend under 10 seconds on a single page view, reporting-module adoption sits at 6%, and the median user returns only after nine days. Because these figures are illustrative, don't use them as industry thresholds.
That pattern is what I call zombie traffic. Users are present in the dataset, but many aren't completing the work that justifies the product. A high ratio can reflect a dashboard tab left open, a notification click, a login obligation, or a narrow group of heavy users carrying the average.
Measure the quality of the return
Track events that indicate progress:
- Events per session: Useful for spotting whether sessions contain one passive event or a sequence of meaningful actions.
- Core-action completion rate: The share of active sessions that complete the action tied to the product's primary value.
- Key-screen depth: Scroll depth or interaction depth can help on content-heavy screens, but it shouldn't replace outcome events.
- Time between meaningful actions: Return cadence reveals whether users repeat the workflow on the product's natural schedule.
- Feature breadth: Adoption beyond the entry feature shows whether users are expanding their reliance on the product.
The thresholds in your operating model must come from your own baseline. A core-action completion rate above 25% and a return cadence under 72 hours can be useful internal warning markers in a daily-use SaaS example, but they aren't universal standards and shouldn't be presented as proof of health.
Two cohorts, one ratio
| Metric | Cohort A, Healthy | Cohort B, Zombie Traffic |
|---|---|---|
| DAU/MAU | Same ratio | Same ratio |
| Session behavior | Multiple meaningful events | One passive page view |
| Core action | Completed regularly | Rarely completed |
| Feature adoption | Expands beyond entry feature | Remains concentrated in one screen |
| Return cadence | Matches the workflow | Long and inconsistent |
| Interpretation | Habit supports value | Activity inflates the ratio |
You can turn these signals into a simple session-quality score, but resist false precision. Start with separate components, such as whether the session included a core action, whether the user used a second relevant feature, and whether the next meaningful return fits the expected cadence. Report the components alongside the score so stakeholders can see what changed.
The product decision follows the weakest component. If users return but don't complete the core action, fix activation or workflow design. If they complete the action once but don't return, investigate recurring value. If they return frequently and complete value but never expand, look for feature or packaging gaps.
Building Your Stickiness Dashboard
Ship the first dashboard around five events, in this order:
- Sign-up: Record account creation with a stable user identifier and acquisition context.
- First core action: Capture the first completed activity that represents delivered value.
- Repeat core action: Record the same action again within the product's defined cadence.
- Feature expansion: Track adoption of a second key feature beyond the entry workflow.
- Churn trigger: Record meaningful inactivity or another behavior that signals departure risk.
Don't begin by instrumenting every click. If these five events are trustworthy, you can answer more useful questions than a large event catalogue full of ambiguous page views.
Match the tool to the job
Use Mixpanel or PostHog for event-based product analysis and fast behavioral cuts. Amplitude is strong for cohort visualization and retention exploration. Heap can help with autocaptured retro analysis when you need to investigate behavior you didn't anticipate instrumenting. For warehouse-grade queries, send a simple export to BigQuery or Supabase.
A dashboard for an indie product doesn't need an elaborate data platform. A practical layout looks like this:
- Top row: DAU/MAU or WAU/MAU, the active definition, and the weekly trend.
- Middle row: Retention curves segmented by acquisition channel, plan, and activation path.
- Bottom row: Core-action frequency, session quality components, feature expansion, and churn triggers.
Teams that need broader operating visibility can borrow layout ideas from these real-world GTM dashboard examples, then replace sales measures with product behavior.
A five-day build plan
Day 1 is taxonomy day. Write the event names, properties, identity rules, and qualifying conditions. Decide what “active” means before anyone opens a chart builder.
Day 2 is instrumentation day. Add the five events, validate duplicate handling, and test anonymous-to-authenticated identity merges. Ask a teammate to complete the primary workflow while you verify the raw event stream.
Days 3 and 4 are assembly days. Build the ratio trend, cohort curves, acquisition-channel breakdown, and session-quality views. Add filters for plan, persona, platform, and signup cohort only after the unsegmented reports reconcile.
Day 5 is review day. Put the dashboard in front of product, engineering, growth, and customer-facing teammates. Ask each person to name one decision the dashboard supports and one event they don't trust. Fix instrumentation gaps before adding another visual.
If you're already paying for Stripe, Notion, and Linear, a lightweight stack can stay simple. Use free PostHog for product events where it fits, export the necessary data, and build one Metabase dashboard without creating a new operational system. Fundl can also publish connected traction signals such as Monthly Active Users on a shareable page, which gives founders a public way to show activity alongside their product metrics.
Combining Stickiness With Churn and NPS
A stickiness percentage becomes useful only when it is read beside churn and NPS. These metrics describe different parts of product health: stickiness measures return frequency, churn measures loss, and NPS captures willingness to recommend. Treat them as a composite signal, not three competing targets.
Review the three metrics weekly, using identical segment definitions and cohort windows. Compare direction first, then investigate causes. A falling stickiness trend can surface weakening habits before churn changes. Rising churn alongside stable stickiness may mean a specific segment is leaving while retained users continue frequent use. NPS adds a satisfaction signal that event data cannot provide.
Three patterns that fool teams
Rising stickiness with falling NPS often indicates that power users are receiving more value while casual or newer users face a worse experience. Examine onboarding, performance, pricing friction, and unmet needs outside the heavy-use segment.
Flat stickiness with falling churn can conceal weak acquisition quality. A stable retained core may offset new accounts that never activate and soon disappear. Break down the ratio and retention curves by acquisition source before accepting the aggregate trend.
Rising stickiness with rising churn usually means a loyal minority is carrying the product while other users leave. A popular workflow may also be overloaded, slow, or difficult for less experienced users.
Use thresholds as investigation triggers, not verdicts. If NPS falls substantially during a quarter while DAU/MAU rises above 25%, examine segment mix and the user experience before adding engagement prompts. A high ratio can reflect concentrated usage rather than broad product health.
Founders deciding how to finance improvements can use a guide to funding a startup on verified traction to identify the evidence investors will expect.
A practical investigation order
- Stickiness down, churn up: Start with retention cohorts and the first core action.
- Stickiness up, NPS down: Start with user segments and qualitative feedback.
- Stickiness flat, churn up: Start with acquisition quality and activation.
- All three deteriorating: Start with the core workflow, reliability, and recent product changes.
When Higher Stickiness Is the Wrong Goal
A rising DAU/MAU ratio isn't automatically progress. Above 25%, treat the score as a yellow flag worth investigating rather than a trophy. The ratio may be rising because power users are hitting the same narrow workflow, the total user base is shrinking, or a feature is being overused before churn appears elsewhere. (Livesession)
Use supporting signals to choose the next investment.
| Stickiness range | Supporting signal | Recommended focus |
|---|---|---|
| Below your product's established baseline | Weak activation or steep early cohort decay | Fix onboarding and first value |
| Above 25% | MAU shrinking or acquisition weakening | Shift attention toward acquisition and audience quality |
| High ratio with NPS below 30 | Frequent use concentrated among dissatisfied users | Improve the casual-user experience |
| Flat weekly retention above 60% | Expansion revenue stalls | Build new capabilities and broaden value |
These thresholds are decision triggers, not universal benchmarks. A payroll workflow may be healthy with weekly engagement, while a messaging product may need far more frequent interaction. Compare the score with the product's natural cadence, active definition, cohort behavior, and business outcome.
A 30-day correction plan
Week one: Audit cohorts by acquisition channel. Find out whether the ratio is supported by recent users, older power users, or a single source of high-intent accounts.
Week two: Interview five churning users and ask them to reconstruct the last successful workflow, the first point of friction, and the moment they stopped expecting value. This is a defined research sample for your product, not a statistical claim about churn behavior.
Week three: Ship one feature that addresses the clearest capability gap. Don't launch several engagement experiments at once, or you won't know what changed the return pattern.
Week four: Re-measure DAU/MAU alongside MAU growth, cohort retention, core-action completion, and feature expansion. The intervention worked only if it moved the right number for the right users.
Stickiness is a health signal, not a scoreboard. Optimize for durable value, not more opens.
Fundl helps founders publish live traction signals, including Monthly Active Users, by connecting data sources to a shareable evidence-based traction page. Visit Fundl to turn your stickiness and retention evidence into a clearer funding story and invite support around measurable progress.