Method guideGrowth

Pirate Metrics AARRR

A practical method guide: clarify prerequisites and materials, understand the complete process, and look up outputs, pitfalls, and stop criteria.

ComplexityMedium

Time1-2 h Setup, laufend

Participants2-8

FormatWorkshop + async

MaturityEstablished

Plan Pirate Metrics (AARRR)

Prerequisite

What needs to be finished first

Complete firstClear scope

Agree the subject, boundary and decision before filling the model.

A broad or mixed scope produces a misleading model.

Complete firstAccess to relevant evidence

Bring source material and people who know the work; distinguish facts from assumptions.

Unverified interpretations become false certainty.

Preparation

What needs to be ready before start

Materials

Shared visual workspace, source evidence, and an agreed decision question.

People / roles

Growth/product, analytics, marketing, engineering, customer support.

Pre-read

Collect existing data, examples, constraints and relevant stakeholder perspectives.

Time needed

90–150 min to define and diagnose; instrumentation and experiments need separate plans.

Setup

Choose product, audience, acquisition channels and cohort window. Define observable events for Acquisition, Activation, Retention, Referral and Revenue. Calculate rates with explicit numerators, denominators, time windows and cohort rules. Compare stages and segments to locate a meaningful drop-off; verify instrumentation. Select one experiment addressing a diagnosed constraint and define its outcome measure.

Core question

The one question this method answers

Where does the customer funnel lose people, and which stage should improve?

Flow

Phases with goal, actions, output and pitfall

Diagnoses a product growth path through Acquisition, Activation, Retention, Referral and Revenue.

AARRR event and measure map

Fictional worked example

Fictional monthly cohort funnel for an appointment product: Acquisition 1,000 eligible landing visits; Activation 180 complete a first booking; Retention 90 book again within the defined return window; Referral: 12 invited people book; Revenue: 75 bookings generate revenue. Invented figures. Stages may overlap; document denominators, event definitions and cohort windows before comparing rates.

AARRR event and measure map
AARRR-Stufe	Event	Metric	Cohort / window
AcquisitionEvent, numerator, denominator and window for this stage	Eligible landing-page visit	Visits / eligible audience	Monthly entry cohort
ActivationEvent, numerator, denominator and window for this stage	First booking confirmed	First bookings / visits	Same entry cohort
RetentionEvent, numerator, denominator and window for this stage	Another booking in window	Repeat bookers / first bookers	Predefined return window
ReferralEvent, numerator, denominator and window for this stage	Invitee books	Invitees booking / invitations	Invited cohort
RevenueEvent, numerator, denominator and window for this stage	Booking generates revenue	Revenue / paying bookings	Same period

1. Event
Acquisition
Eligible landing-page visit
Activation
First booking confirmed
Retention
Another booking in window
Referral
Invitee books
Revenue
Booking generates revenue
2. Metric
Acquisition
Visits / eligible audience
Activation
First bookings / visits
Retention
Repeat bookers / first bookers
Referral
Invitees booking / invitations
Revenue
Revenue / paying bookings
3. Cohort / window
Acquisition
Monthly entry cohort
Activation
Same entry cohort
Retention
Predefined return window
Referral
Invited cohort
Revenue
Same period

AARRR Metrics · Dave McClure

1
Cohort
Define product path and cohort
Goal
Measurement frame
Inputs for this step
- Audience, business model and analytics events
Actions
- Choose audience, entry cohort, observation window and the relevant AARRR relationships for this product.
Output
Measurement frame
Measurement frame
Quality check
Is the funnel fitted to the product and time window?
Pitfall
AARRR is forced into one universal sequence. Adapt stages to behaviour and revenue model; state overlaps.
Timing: Allow time according to scope and evidence; no universal duration.
2
Events
Operationalize events and stages
Goal
AARRR event dictionary
Inputs for this step
- Tracking plan, business rules and analytics
Actions
- Define Acquisition, Activation, Retention, Referral and Revenue with events; specify numerator and denominator for each transition.
Output
AARRR event dictionary
AARRR event dictionary
Quality check
Are events observable and unambiguous?
Pitfall
Vague labels such as “active” vary by team. Define event, timing, deduplication and exclusion.
Timing: Allow time according to scope and evidence; no universal duration.
3
Conversion
Analyze cohorts and conversion
Goal
Funnel with sources and rates
Inputs for this step
- Instrumented events, data quality and segments
Actions
- Calculate stage conversion with consistent cohort rules and inspect missing data and segment differences.
Output
Funnel with sources and rates
Funnel with sources and rates
Quality check
Are window and denominator comparable for each rate?
Pitfall
Aggregate totals conceal drop-offs and different exposure. Use cohorts and relevant segments rather than cumulative totals.
Timing: Allow time according to scope and evidence; no universal duration.
4
Experiment
Investigate constraint and experiment
Goal
Prioritized growth hypothesis
Inputs for this step
- Funnel analysis, qualitative evidence and team capacity
Actions
- Choose a meaningful drop-off, validate its plausible cause and run one experiment with a guardrail and evaluation window.
Output
Prioritized growth hypothesis
Prioritized growth hypothesis
Quality check
Does the experiment test the suspected cause, not activity?
Pitfall
A campaign is launched without checking the cause. Combine behaviour data with interviews or observation.
Timing: Allow time according to scope and evidence; no universal duration.

Artifact

What comes out at the end

Sketch: what the artifact looks like

Pirate Metrics AARRR · Growth FunnelAcquisition, Activation, Retention, Referral und Revenue als messbaren Wachstumspfad prüfen

Form

Diagnoses a product growth path through Acquisition, Activation, Retention, Referral and Revenue. Record scope, evidence, assumptions, owner and next review in the versioned artifact.

Versioning / ownership

Record scope, date, participants, source version and owner. Note changed assumptions and the reason for each revision.

Tool alternatives

Shared whiteboard or paper template
Source links, records or measurement data

pirate-metrics-working-template.md

Prepare the decision, source evidence and next review.

Pirate Metrics AARRR: Worksheet

Question

To be confirmed

Desired outcome

To be confirmed

Scope

Run the method Time: See catalog planning range; no universal duration.

Complete this worksheet on paper or in your own document during the work.

Preparation for this scope

Method setup

Materials: Shared visual workspace, source evidence, and an agreed decision question.
Roles: Growth/product, analytics, marketing, engineering, customer support.
Advance information: Collect existing data, examples, constraints and relevant stakeholder perspectives.
Overall time needed: 90–150 min to define and diagnose; instrumentation and experiments need separate plans.
Setup: Choose product, audience, acquisition channels and cohort window. Define observable events for Acquisition, Activation, Retention, Referral and Revenue. Calculate rates with explicit numerators, denominators, time windows and cohort rules. Compare stages and segments to locate a meaningful drop-off; verify instrumentation. Select one experiment addressing a diagnosed constraint and define its outcome measure.

Prepare for the work steps

Cohort: Define product path and cohort

Audience, business model and analytics events

Events: Operationalize events and stages

Tracking plan, business rules and analytics

Conversion: Analyze cohorts and conversion

Instrumented events, data quality and segments

Experiment: Investigate constraint and experiment

Funnel analysis, qualitative evidence and team capacity

AARRR event and measure map

AARRR-Stufe	Event	Numerator / denominator	Cohort / window
Acquisition
Activation
Retention
Referral
Revenue

Acquisition: Event, numerator, denominator and window for this stage
Activation: Event, numerator, denominator and window for this stage
Retention: Event, numerator, denominator and window for this stage
Referral: Event, numerator, denominator and window for this stage
Revenue: Event, numerator, denominator and window for this stage

Define event, numerator, denominator, cohort and time window per stage.

AARRR Metrics · Dave McClure

Cohort: Define product path and cohort

Expected artifact: Measurement frame

Measurement frame

Entry:

...

Is the funnel fitted to the product and time window?

Events: Operationalize events and stages

Expected artifact: AARRR event dictionary

AARRR event dictionary

Entry:

...

Are events observable and unambiguous?

Conversion: Analyze cohorts and conversion

Expected artifact: Funnel with sources and rates

Funnel with sources and rates

Entry:

...

Are window and denominator comparable for each rate?

Experiment: Investigate constraint and experiment

Expected artifact: Prioritized growth hypothesis

Prioritized growth hypothesis

Entry:

...

Does the experiment test the suspected cause, not activity?

Open questions and next steps

...

Method guide: https://methodatlas.meierhoff-systems.de/en/methods/pirate-metrics/run-sheet

Example output

Concrete filled scenario, fictional example

Fictional example for orientation. Counts and timings describe this case and are not universal requirements.

pirate-metrics-beispiel.md

Concrete filled scenario, fictional example

Pitfalls

Recognize symptoms and steer against them

Quick overview, the full text is in the matching phase in the flow

Define product path and cohortAARRR is forced into one universal sequence.View in phase
Operationalize events and stagesVague labels such as “active” vary by team.View in phase
Analyze cohorts and conversionAggregate totals conceal drop-offs and different exposure.View in phase
Investigate constraint and experimentA campaign is launched without checking the cause.View in phase