What needs to be finished first
A defined North Star Metric or primary outcome metric exists so the AARRR stages can be aligned.
A basic funnel logic with defined conversion points is implemented in tracking.
What needs to be ready before start
Whiteboard or Miro board with five columns (Acquisition, Activation, Retention, Referral, Revenue); stage template with fields for definition, metric, baseline, target, source; tracking schema for the product.
One owner (Head of Growth or Product Lead); one data analyst with product analytics access; one person each from marketing, product, customer success, and sales; scribe.
Current data from the last 90 days per stage, if available; definition of activation moment; known drop-off points; cohort reports on retention.
3-4 h
Create columns and define AARRR order. Prepare one definition card per stage. Data analyst works in parallel in BI tool and streams values live into the workshop.
The one question this method answers
Which metric represents each AARRR stage, where is the largest leverage at the bottleneck, and which experiments address it?
Marker: Phase
| Step | Duration | Action | Hint |
|---|---|---|---|
1Phase 1: Stage definitions | 30 min | Define a product-specific meaning for each stage: what counts as Acquisition, Activation, Retention, Referral, Revenue. Run a sample user path. | Activation is often set too early (sign-up) or too late (first purchase). Ask: after which action does the user become significantly more likely to return? |
2Phase 2: Metrics and baselines | 45 min | Define one primary metric and 1-2 secondary metrics per stage. Enter current values from BI. Calculate conversion rates between stages. | If a stage has no metric, either tracking is missing or the stage is irrelevant to the business model. Make both explicit. |
3Phase 3: Identify bottleneck | 30 min | Compare conversion rates with benchmarks (industry, own history). Mark stage with highest relative weakness. Form a hypothesis on why it is the bottleneck. | The largest absolute drop is not always the main bottleneck. If 80% drop at Acquisition but the highest value stage is retention, retention can still be the best leverage point. |
4Phase 4: Hypotheses and experiments | 45 min | Create 3-5 hypotheses per bottleneck, prioritized by ICE or RICE. For each top hypothesis draft an experiment (setup, success metric, duration, owner). | Hypotheses must target the identified conversion rate, not improve a different stage. Distinguish clearly what measures success of the experiment. |
5Phase 5: Cadence and rollout | 20 min | Set weekly or bi-weekly AARRR-stage review cadence. Dashboard layout with five tiles plus conversion rates. Name an owner per stage. | Without cadence, the model is filled once and then unused. Create recurring calendar cadence immediately. |
What comes out at the end
AARRR dashboard with five stages, main metric per stage, secondary metrics, baseline, target, owner and trend, plus hypothesis backlog with priorities and ongoing experiments.
Track dashboard configuration as code, for example through Looker LookML. Archive a quarterly snapshot as markdown report. Version stage definition changes and keep change date visible.
Compact working template for Pirate Metrics AARRR with context, input, output artifacts, and next step.
Structures growth across Acquisition, Activation, Retention, Referral, and Revenue.
When and for what do we use this method?
Which data, observations, decisions, or materials are available?
Short notes along the runsheet.
Owner, date, and success signal.
Concrete filled scenario, fictional example
Concrete filled scenario, fictional example
| Stage | Definition | Metric (90D) | Previous period | Trend |
|---|---|---|---|---|
| Acquisition | First visit to landing page | 18,400 visits | 16,200 | +13% |
| Activation | First verified booking in 7 days | 1,260 (6.8%) | 1,180 (7.3%) | -7% |
| Retention | At least one booking in the following week | 38% | 41% | -3 pp |
| Referral | Person invited who reaches Activation | 7% | 6% | +1 pp |
| Revenue | Paying users per month | 412 (33%) | 388 (33%) | flat |
Bottleneck: Activation drops from 7.3% to 6.8% despite traffic growth. Hypothesis: onboarding email sequence was changed in week 14 and the new variant did not explain day-of-booking.
Top experiment: revert onboarding email to previous version and add inline tutorial on first search. Success metric: raise Activation rate to 7.5% within 4 weeks. Owner: @lisa. Start: 25.05.
Cadence: Wednesdays at 10:00, 30 min, each stage owner gives 2-minute update.
Recognize symptoms and steer against them
Activation rate looks good while retention still declines.
Identify the Aha moment empirically: which action correlates strongest with retention after 4 weeks? That action should become Activation definition.
Multiple teams work on all stages at once; learning rate falls and effects overlap.
Set focus on the bottleneck. Keep other stages stable, avoid optimizing all simultaneously. Change bottleneck only when clear improvement evidence exists.
Effort goes into referral programs while Activation or Retention remain weak.
Prioritize referral only when Activation and Retention are stable. Referral amplifies what is already working.
Revenue rises through first-time purchases, but cohort LTV falls and customers churn after 2-3 months.
Always evaluate revenue on cohort basis. Monthly revenue without cohort LTV is vanity. Retention stage is a prerequisite for sustainable revenue.
Marketing dashboard and product analytics show different numbers, and the discussion drifts into methodology disputes.
Set one canonical source per stage. Version definitions as code (SQL or tool configuration). Other tools must use the same definitions or be turned off.
Done signals checkable in under a minute
Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.