Pirate Metrics (AARRR) — five numbers you should track starting today

Dave McClure (500 Startups, now 500 Global) coined a simple framework for early-stage startup metrics in 2007: AARRR. The acronym is fun; the logic is the most serious thing an early-stage founder should measure.

Five stages of a user's journey:

1. Acquisition — How does the user find you?
2. Activation — Does the first session give them an "aha"?
3. Retention — Do they come back?
4. Referral — Do they tell others?
5. Revenue — Do they pay?

The key: these aren't ordered for fun, each one is a prerequisite for the next. If users don't activate, retention is zero. If retention is zero, referral and revenue won't happen.

Why most startups measure this wrong

Mistake 1: Acquisition only — growing user count without tracking activation. The rest is a leaky bucket.

Mistake 2: All metrics in one dashboard, none of them actionable. The number is there but no decision flows from it.

Mistake 3: Measuring absolute numbers, not ratios. "We have 500 users" tells you nothing. "500 acquisitions, 30% activate (150), 40% return in a week (60)" — that tells you where it breaks.

Practical guide per stage

Acquisition — where do they come from?

What to measure: visitors per channel, costs (CAC), conversion to signup.

AI's role: describe your product to AI and ask "Give me 10 acquisition channels for my target segment, ranked likely-most-effective to least, with reasoning."

This is a hypothesis list, not truth. Testing reveals reality.

Activation — does first-timer get the "aha"?

The most important stage, most often forgotten. Activation = user experiences product value for the first time.

For Innovaidor it might be "user gets the first critical insight in Core Chat".

How to measure: define a concrete event (3+ messages in Core Chat, document download, etc.). Fire it through Matomo as an event. Calculate: % of signups completing it.

Good activation: above 40%. Weak: below 15%.

Retention — do they return?

Retention is the most predictive metric for your company's long-term success. If retention is weak, everything else is futile.

How to measure: use cohort analysis. "Of January signups, 40% returned in week 2, 30% in week 4, 25% in week 8." If the curve flattens at some level → you have a product. If it keeps falling toward zero → you don't yet.

AI's role: analyze retention data. Prompt: "This is 30-day retention data. Is the trend line flattening? Where do you predict it lands?"

Referral — do they tell others?

NPS is the classic. Simpler: ask active users "How likely are you to recommend, 1–10?"

Tip: add a channel where the user can actually refer (invite code, share link). Measure how many actually use it.

Revenue — do they pay?

Free → Paid conversion. The metric to understand if you're building a sustainable business.

Good early-stage conversion: 2–5%. Second most important: MRR per user and churn (monthly cancellation rate).

AI's limited role in growth measurement

AI can't measure your company for you. Its strengths are:

1. Hypothesis generation — what to test next
2. Data analysis — once you have numbers, AI finds patterns and anomalies
3. Industry benchmarks — "Is 25% activation rate good for SaaS?"

The actual measurement happens through your production analytics (Matomo, PostHog, Stripe, DB queries).

Innovaidor and Pirate Metrics

In Innovaidor, Pirate Metrics is one of the methods. You activate it when you want to organize your product's metric stack. AI walks through five stages, asks what you measure now, suggests what you should measure, and generates a concrete "metrics list" you can implement.

Closing

Pirate Metrics is simple but powerful. Most startups don't even track Activation — focused only on Acquisition and wondering why users disappear.

Start with one: define this week what your activation event is. Measure it. If under 30% of signups reach it, you have an activation problem — and increasing Acquisition budget won't help.