Pirate metrics

One total revenue figure cannot say where growth is being lost. Spending more on acquisition when new users never reach a first good experience only fills a leaky bucket faster. Measuring each stage shows the single weakest point, and fixing that one moves the whole result more than polishing a stage that already works.

The framework also gives a team a shared language. Marketing owns the top, product owns activation and retention, and sales or customer success owns revenue and referral. When a number slips, it is clear whose stage it is and what to ask.

It is a diagnostic, not a growth plan. It tells you where the leak is. It does not tell you why. For that, use customer interviews and session recordings.

• Define one measurable event for each stage, for example "signed up", "sent first invoice", "active in week four".
• Count how many people move from each stage to the next.
• Find the stage with the biggest drop. That is the constraint, so work on it first.
• Watch referral closely, since each referred customer lowers the cost of acquiring the next.
• Review the five numbers on a fixed rhythm, weekly if volume allows.

The name comes from the sound of the acronym, AARRR, said like a pirate. Dave McClure introduced it in 2007 as a way to give a young company five numbers to watch instead of one.

• Acquisition: how people first find you, for example a visit or a signup.
• Activation: whether a new user reaches a first good experience of the product, such as finishing setup or booking a call. This is the product sense of the word. It is not the same as the sales sense, where a lead is "activated" by showing interest. See activation rate for that meaning.
• Retention: whether they come back or keep paying.
• Referral: whether they bring others.
• Revenue: whether they pay, and how much.

Some teams place revenue earlier in the order. The point is the five questions, not the sequence.

• Defining a stage loosely, such as "engaged", so nobody can count it. Use one measurable event.
• Treating all five stages as equally urgent. Fix the weakest first.
• Measuring only totals and not the conversion between stages.
• Judging stages over too short a window. Retention needs weeks of data.
• Chasing referral before activation and retention work. Few people recommend a product they do not use.