Pirate Metrics (AARRR)

AARRR

Pirate Metrics, coined by Dave McClure as AARRR, is a five-stage funnel framework for measuring SaaS growth: Acquisition (how users find you), Activation (their first positive experience), Retention (returning to use it again), Revenue (converting to paid), and Referral (bringing others). Each stage has distinct metrics and failure modes. The framework makes it possible to diagnose exactly which stage is the constraint in your growth engine.

formula.sh

Growth = f(Acquisition × Activation rate × Retention rate × Revenue conversion × Referral coefficient)

> Acquisition: signups, trials, or free registrations per week
> Activation: % of new users who complete the aha moment event
> Retention: % of activated users still active at 30 / 60 / 90 days
> Revenue: % of active users who convert to paid (or ARPU growth)
> Referral: % of users who generate at least one new signup (viral coefficient)

example

example.sh

Monthly: 500 signups → 150 activated (30%) → 90 retained at day 30 (60% of activated) → 27 converted to paid (30% of retained) → 5 referrals (18% of paid)

500 × 30% × 60% × 30% = 27 paying customers per month. Referral coefficient: 5/27 = 0.19 — sub-viral but contributory

→ Constraint identified: Activation at 30% is the primary leak. Improving to 50% yields 45 paid customers/month — a 67% improvement from one change.

why it matters

AARRR is valuable not as a measurement framework but as a diagnostic tool. Most founders know their Acquisition numbers and their Revenue numbers but have no visibility into Activation and Retention. The stages in between — where product experience determines whether an acquisition becomes a customer — are where most SaaS products quietly destroy value every week.

The framework enforces stage sequencing. Pouring more money into Acquisition when Activation is 10% is burning cash to fill a bucket that has no bottom. AARRR makes the priority order obvious: fix Activation before scaling Acquisition, fix Retention before celebrating Revenue.

common mistakes

✗Treating AARRR as a sequential checklist to complete once rather than a continuous measurement loop to run weekly

✗Defining "Activation" as account setup completion instead of the first value event — this inflates Activation rate and hides the real problem

✗Optimizing the "easy" stages (Acquisition, Revenue conversion) while ignoring the hard stages (Retention) because the data is more uncomfortable

pro tips

✓Build a single AARRR funnel report that shows conversion rates between each stage, updated weekly — the stage with the worst conversion rate is your top priority

✓Run cohort analysis at the Retention stage, not just a single rate — a retention curve that flatlines above 20% indicates strong PMF; one that trends toward zero does not

✓Add a sixth stage if relevant: "Revenue Expansion" (NRR) between Revenue and Referral — expansion revenue is often more controllable than referral

the mrrsucks take

Your AARRR funnel is technically an AARRR, but in practice it is just a very expensive A. You acquire users magnificently, activate them occasionally, retain them briefly, convert them rarely, and they refer nobody. You have built a world-class top-of-funnel for a product that currently lacks a bottom.

faq

Should the "R" stages be measured in order: Retention before Revenue?+

In the framework, Retention comes before Revenue, which reflects the logic that retained users are far more likely to convert to paid than users who are about to churn. In practice, for SaaS with paid trials, Revenue and Retention are measured in parallel — but fix Retention problems first regardless.

What is a viral coefficient and how does it relate to Referral?+

The viral coefficient is the average number of new signups each existing user generates. A coefficient above 1.0 means the product grows without any paid acquisition. Below 1.0 is typical for most B2B SaaS (0.1–0.3 is common). Referral in AARRR tracks the rate at which users generate referrals, not necessarily a viral coefficient.

Is AARRR still relevant or has it been replaced?+

The framework is 15 years old but its diagnostic logic is timeless. The stages map cleanly to modern PLG funnels. The main evolution is that "Referral" is now often reframed as "Loop" — the mechanism by which the product generates its own next cohort of Acquisition, whether through virality, network effects, or user-generated content.

$10K MRR milestone›Zero revenue roasts›

related metrics

./install-the-daemon

$9. 365 roasts. one public endpoint of pure shame.