Measure what happens after the sign up
The AARRR framework is useful when a product has plenty of registrations and little evidence that people receive value. Acquisition, activation, retention, referral and revenue give the team a way to inspect what happens next. The difficult work is defining those events honestly. A welcome screen is easy to count. A customer completing a useful task is a more demanding, and more informative, starting point.
Understand the five measurement jobs
Dave McClure’s original presentation describes acquisition, activation, retention, referral and revenue as startup metrics.1 Acquisition concerns how people arrive. Activation concerns an initial useful experience. Retention concerns whether they return or continue using the product. Referral concerns customers bringing others. Revenue concerns the commercial result. The order of payment and product use can vary by business; the acronym should not force an unsuitable transaction sequence.
Use the model to organise questions, then define events for the product. A workplace tool might activate when an eligible account completes a real workflow with another team member. An ordering app might activate when the first fulfilled order arrives. A registered account that never receives value should not count as activated merely because it accepted an email address. The definition needs product and customer evidence. Our guide to RACE framework covers the next question.
Acquisition needs identity and cost rules
Decide whether the unit is a person, account, household or business. Deduplicate where the data permits and state the limits. Count eligible new users within the chosen window rather than mixing returning users with newly acquired ones. Separate paid, unpaid and partner routes with consistent source rules. Unattributed activity should remain visible instead of being assigned to whichever channel makes the report more comfortable.
Acquisition costs also need a boundary. Advertising cost divided by eligible sign-ups is not the same as full customer acquisition cost. The latter requires a customer definition and an agreed set of sales and marketing costs. A product can acquire registrations inexpensively while acquiring paying, retained customers at a poor cost. Show both the early response and the later cohort outcome.
Activation should represent a first useful outcome
Talk to customers about the moment the product became useful. Then examine whether a recorded event can represent that moment reliably. The definition should be observable, reproducible and resistant to accidental triggering. Completing an empty setup form may be less informative than finishing a genuine task, but a task event also needs rules to exclude tests or duplicate activity.
Within the AARRR framework, activation rate is activated eligible new users or accounts divided by eligible new users or accounts, with a stated observation window. Give each cohort enough time to complete that window. Comparing a thirty-day-old cohort with a three-day-old cohort will often manufacture a decline. Record changes in product instrumentation so an apparent improvement does not merely reflect a changed event definition. If that is your next question, read our guide to RFM analysis.
Retention reveals whether the value continues
Specify what retained means. A daily-use tool, a monthly service and an occasional booking product need different usage windows. Cohort retention follows an eligible starting group and measures the proportion meeting the retention definition later. Aggregate active-user growth can hide weak retention if new acquisition continually replaces departing users.
Revenue retention is a separate question. A subscription business may retain fewer accounts while expanding revenue from those that remain. Net revenue retention includes changes within the existing recurring-revenue cohort; new customer revenue should not be inserted to improve the result. Product usage can diagnose behaviour, but the commercial measure must use the appropriate billing records. Explain the difference rather than placing both percentages under an ambiguous retention label.
Connect acquisition to continuing value
Share the activation definition, cohort windows and the stage where useful customers disappear. We can discuss the communication and research needed before the business increases acquisition spend.
Make referral and revenue honest
A referral needs a defined outcome. Sending an invitation is different from an invited user joining, activating or paying. Revenue should come from an agreed financial source and show discounts, refunds and the measurement window. The goal is a dependable account of the business, not the most flattering version of the funnel.
- Record acquisition by eligible new user or account, with deduplication and source rules. Show unknown sources instead of forcing attribution.
- Define activation as a meaningful product event within a stated window. Check the event against actual customer experience before treating it as value.
- Measure retention within a starting cohort and appropriate interval. Keep newer cohorts from appearing worse simply because they have had less observation time.
- Separate invitations sent, referred users acquired and referred customers retained. Each answers a different question about whether referral produces useful growth.
- Use financial records for revenue, refunds and recurring changes. Include relevant contribution and acquisition economics before deciding that the product is ready to scale.
Choose experiments that answer the constraint
If acquisition is strong and activation weak, investigate audience fit, onboarding and product usefulness before commissioning a larger campaign. If activation is strong but retention weak, talk to customers who leave. The model locates a question; it does not automatically diagnose the cause. A retention email cannot repair a product that fails the customer’s task.
For an experiment, predefine the primary outcome, eligible population, assignment method and observation period. Avoid declaring a win because one of many secondary numbers improved. Consider the effect on other stages: a simpler sign-up may raise registrations while attracting fewer suitable accounts. Keep the original cohort definitions stable, report uncertainty and make the next product decision from the whole result. Customer lifetime value is explained in a separate guide.
Questions people ask
What counts as activation?
A defined first useful product experience, supported by customer and product evidence. Choose an observable event and a time window. Account creation alone is often too weak, but the right definition depends on the product.
Must referral happen before revenue?
No. Businesses have different transaction patterns. The categories organise measurement questions; they are not a requirement that every customer pass through the events in the acronym’s order.
Is a high activation rate enough to scale acquisition?
No. Check retention, revenue, acquisition cost and the delivered experience. An initial action can look healthy while the business loses customers or contribution later.
Discuss the work your team needs next
If sign-ups are growing while retained customers are not, bring the cohort definitions and product journey to Quiamo. An AARRR framework review can connect acquisition communication with activation and continuing value, while separating product questions from media questions. Start with the event your team currently calls success and the evidence that customers actually benefit from reaching it. That is a useful foundation for the next growth decision. Explore our brand strategy engagement to discuss the scope and deliverables.
Discuss the next step with Quiamo
Share the plan, the decision it needs to support and what you can already measure. We will tell you plainly which approach fits, what to track and what the evidence can and cannot show. The first call is free.
Sources
- Dave McClure Startup Metrics for PiratesCreator presentation of acquisition activation retention referral revenue.
Worked examples on this page are fictional and labelled. Search phrases and figures from external sources are cited; nothing here describes a Quiamo client result.
