# Pirate Metrics (AARRR): The 5 Stages and Where It Breaks

> Pirate metrics (AARRR) explained: what each of the 5 stages measures, a worked dollar example with real benchmarks, and when to switch from AARRR to RARRA.

- URL: https://missiongrowth.io/blog/pirate-metrics
- Published: 2026-08-24 · Updated: 2026-09-24
- Author: Ömer Furkan Aktaş, Founder, Mission Growth
- Publisher: Mission Growth. Company facts: https://missiongrowth.io/llms.txt

Pirate metrics is Dave McClure's five-stage system for valuing a customer's journey in dollars: Acquisition, Activation, Retention, Referral, Revenue. He introduced it in 2007, and the acronym, AARRR, stuck.

Most explanations stop at the definition. They list the five stages, give one metric per stage, and never say when following them in that exact order stops working. It does stop working, in conditions you can check against your own numbers: rising acquisition cost, flat retention, a market that's gotten crowded.

In this guide:

- What each of the five AARRR stages actually measures
- The real free-to-paid conversion benchmarks behind the Revenue stage
- The checkable trigger for switching to RARRA's retention-first order
- A worked example that turns signups into dollars per signup
- Where AARRR breaks down for B2B SaaS

## What is the pirate metrics (AARRR) framework?

Pirate metrics, or the AARRR framework, values each part of a customer's journey in dollars instead of vanity numbers.

Dave McClure introduced the AARRR pirate metrics framework in 2007. Its five checkpoints, one line each:

- **Acquisition** – how someone finds and tries your product.
- **Activation** – the point where they experience real value.
- **Retention** – whether they come back.
- **Referral** – whether they bring others.
- **Revenue** – whether any of it makes you money.

McClure gave the original talk, "Startup Metrics for Pirates," at a Seattle Ignite event and posted about it on September 6, 2007. Before that, he'd spent three years as Director of Marketing at PayPal, then worked as an angel investor. He co-founded 500 Startups (now 500 Global) with Christine Tsai in 2010, three years after the talk that made his name.

AARRR has exactly five stages. One guide adds "Awareness" as a stage ahead of Acquisition; that's not part of McClure's original framework, and it restates what Acquisition already measures: how people first find you.

Some call it the pirate funnel, but that name undersells it. Treat the five stages as a checklist of checkpoints to value, not a pipeline every user has to complete start to finish. That distinction is the one the rest of this guide is built on.

## The five AARRR stages, checkpoint by checkpoint

Each of AARRR's five stages has one or two checkpoints worth tracking, and two of them rarely get a real number attached.

Acquisition, Activation and Retention are well covered everywhere. Referral and Revenue are the two stages that usually stop at a definition.

### Acquisition

Acquisition is how someone finds and tries your product.

Track two checkpoints: new signups or qualified leads, and customer acquisition cost (CAC), the amount you spent to turn a stranger into that signup.

Segment both numbers by channel and, where you can, by persona; a channel that looks cheap on average can be expensive for the one persona you actually want.

For example, once you know which channel is worth testing further, our guide to [growth experiment cadence](https://missiongrowth.io/blog/growth-experiment-cadence) covers building and prioritizing that testing backlog instead of running experiments at random.

### Activation

Activation is the point where a new user experiences your product's real value, the "aha" moment.

Track activation rate, the share of signups who reach it, and time to activate, how long that takes them.

### Retention

Retention is whether people come back. Track retention rate and its inverse, churn rate.

Mixpanel's 2024 Benchmarks Report, drawn from more than 7,700 customers, found the average retention rate after the first week fell from 50% to 28% in 2023. That drop is a reminder that "good retention" depends on your category and cohort.

::figure{src="/blog/figures/pirate-metrics-5.svg" alt="Average week-one product retention drops from 50% to 28% during 2023, showing how quickly a retention benchmark can shift." caption="By Mixpanel's count, week-one retention fell from 50% to 28% in 2023." width="720" height="182"}

### Referral

Referral is whether existing users bring in new ones. Track invites sent per active user and the conversion rate on those invites.

A channel only counts as a real growth lever once its viral coefficient clears 1: each active user needs to bring in more than one new user through invites for referral to compound on its own. See our guide to [viral coefficient](https://missiongrowth.io/blog/viral-coefficient) for the formula and how to measure it.

Designing and running a [referral program](https://missiongrowth.io/blog/referral-program) end to end, the invite mechanics, the reward structure, the launch sequence, is worth its own guide, and it covers building one from scratch.

### Revenue

Revenue is whether any of this makes you money.

The checkpoint that carries this stage is free to paid conversion, and the published benchmarks disagree sharply depending on how you ask someone to pay.

ChartMogul's SaaS Conversion Report puts the median free to paid conversion rate at 8% across 200 B2B software products. Trials that require a credit card convert at 30%, more than 5x higher than trials that don't.

First Page Sage's 2025 benchmark study of 86 SaaS companies found the same pattern from a different angle: 18.2% (organic) and 17.4% (paid) for opt-in trials, against 48.8% (organic) and 51% (paid) when the trial requires a card upfront.

The two studies disagree on exact numbers because they aren't measuring the same trial design. Asking for a card upfront filters out casual signups before they ever reach the conversion step, so the rate that's left reflects who signed up, and it can look identical whether the marketing behind it was good or mediocre.

Which number to trust for your own product depends on your trial type, not on which study you read first.


::figure{src="/blog/figures/pirate-metrics-2.svg" alt="A dot-range chart shows free to paid conversion from 8% to 51% across two studies, higher when trials require a card upfront." caption="Free to paid conversion ranges from 8% to 51% across two 2025-2026 studies, and the spread tracks trial design, not marketing quality." width="720" height="414"}

## Where AARRR breaks: when to flip to retention-first

AARRR's order, acquisition first, breaks once three conditions show up together, each one checkable against your own numbers instead of a guess.

The three triggers:

- Customer acquisition cost has been trending up for two straight periods.
- Activation to retention conversion has been flat or falling for two straight periods.
- Three or more direct competitors are fighting for the same buyer.

Hit all three, and pushing more people through the top of AARRR's order just feeds a leakier bucket. At that point, RARRA, the order that starts with retention, is the better fit. This is the AARRR vs RARRA decision in practice: same five checkpoints, different starting priority.


::figure{src="/blog/figures/pirate-metrics-3.svg" alt="A quadrant chart plots CAC trend against competitive density, marking the trigger for switching AARRR to an order that starts with retention." caption="A rising CAC and three or more direct competitors is the trigger for switching AARRR's order to retention-first." width="720" height="440"}

RARRA, Retention, Activation, Referral, Revenue, Acquisition, comes from Thomas Petit's "Mobile growth #RARRA" presentation at TheFamily's Growth meetup #5 in Berlin, June 2017. One of his slides credits the line to @gillianim: "If you invest in growth before you have retention, you're renting users."

AARRR also scores Referral as a one-time checkpoint rather than something that compounds on its own. Our guide to [loops versus funnels](https://missiongrowth.io/blog/growth-loops) covers the fuller case for thinking in loops as a model, beyond this one framework.

## Valuing each AARRR stage in dollars: a worked example

AARRR's stages earn their real value once you divide the revenue your paying customers generate back up through every earlier stage.

That turns five checkpoints into one dollar figure per stage you can defend in a roadmap review.

Here's the arithmetic, built from two real benchmarks and two stated illustrative assumptions. Say a SaaS product with a 14-day trial that requires a card gets 5,000 signups in a month. That signup count and the $49 average contract value below are illustrative starting numbers; the retention and conversion rates applied to them are real.

::dataset{key="pirate-metrics-worked-example" name="Worked example: valuing AARRR stages in dollars (illustrative signups, 2024-2025 benchmarks)"}

| Step | Calculation | Result |
|---|---|---|
| Retained after week one | 5,000 signups × Mixpanel's 2024 rate of 28% | 1,400 users |
| Paying customers | 5,000 signups × First Page Sage's 2025 rate of 48.8% | 2,440 customers |
| Monthly recurring revenue | 2,440 customers × $49 illustrative average contract value | $119,560 MRR |
| Value per signup | $119,560 MRR ÷ 5,000 signups | $23.91 |


::figure{src="/blog/figures/pirate-metrics-1.svg" alt="A five-stage AARRR funnel chart shows 5,000 signups narrowing to 1,400 retained users and 2,440 paying customers, ending at $23.91 of value per signup." caption="The worked example turns 5,000 signups into $23.91 of value per signup once retention and revenue benchmarks replace guessed round numbers." width="720" height="332"}

Rerun the same calculation with your own signup count, retention rate and conversion rate, and you'll have a dollar figure for each stage instead of a checkpoint you can only describe.


::figure{src="/blog/figures/pirate-metrics-4.svg" alt="A blank pirate-metrics scorecard checklist lists the five AARRR stages with empty fields for the reader's own signup, retention and conversion numbers." caption="Copy this scorecard and drop in your own signup count, retention rate and conversion rate to value each stage." width="720" height="531"}

Once you have a value per signup, check it against your unit economics with our [LTV:CAC calculator](https://missiongrowth.io/tools/ltv-cac-calculator).

Getting this per-stage valuation right matters most exactly when the numbers are smallest, which describes most young companies. Our guide to [startup SEO](https://missiongrowth.io/blog/startup-seo) covers how those same young teams should be earning that first batch of signups organically instead of buying all of them.

## Common mistakes when teams apply pirate metrics

Pirate metrics most often breaks down in a B2B SaaS motion: a long sales cycle and a handful of valuable accounts don't match its consumer defaults, self-serve signup, easy referral.

Rename the checkpoints for that motion instead of dropping the framework entirely. Four mistakes show up again and again:

- **Treating Activation as a self-serve event.** A B2B buyer rarely has an "aha" moment inside a trial the way a consumer app user does. Map Activation to reaching a qualified opportunity instead.
- **Measuring Retention like a consumer app.** Daily or weekly logins mean little for software a customer's team deploys once and relies on for months. Map Retention to active deployment: is the account still using what it bought.
- **Forcing a Referral checkpoint that doesn't exist.** With a small, known set of accounts, there usually isn't enough volume for a referral rate to mean anything. Drop the checkpoint rather than manufacture a metric for it.
- **Optimizing every checkpoint at once.** Find the single weakest stage in your own numbers first and fix that one before touching the other four. A team has limited engineering time; spreading it evenly across all five stages produces smaller gains than fixing the one actual bottleneck.

AARRR was built in 2007 for a world of self-serve signups and cheap experiments, and it still works well there. Treat it as five checkpoints to value; that gets you further than the plain definition most pages settle for.

Start with the checkpoint your own numbers say is weakest. Put a dollar figure on it with the scorecard above, then check it against CAC, retention trend and competitive density to see whether it's time to flip the order.

Buyer journeys that don't move in a straight line show up outside AARRR too. The same problem shapes [saas SEO](https://missiongrowth.io/blog/saas-seo) content planning, where a prospect's path from first search to signed contract rarely follows the tidy order a framework implies.

## FAQ

### Does pirate metrics work for B2B SaaS?

Yes, but the checkpoints need renaming. Map Activation to reaching a qualified opportunity instead of a self-serve signup event, and Retention to active deployment instead of daily app opens. Skip forcing a Referral checkpoint if you're selling to a small, known set of accounts; there usually isn't enough volume for it to mean anything.

### What is a product metrics framework, and why does it matter?

A product metrics framework is a fixed set of checkpoints, like AARRR's five stages, so a team measures the same things the same way over time. That consistency is what makes stage-over-stage comparison possible: without it, you're picking a different metric every quarter and can never tell if you're actually improving.

### What's the difference between AARRR and RARRA?

Same five checkpoints, different starting priority. AARRR runs Acquisition through Revenue in that order; RARRA reorders them to Retention, Activation, Referral, Revenue, Acquisition, on the logic that acquiring more users before you've proven you can keep them just feeds a leakier bucket.

### How many stages does the AARRR framework have?

Five: Acquisition, Activation, Retention, Referral and Revenue. Some guides list six by adding "Awareness" as a stage ahead of Acquisition, but that's not part of Dave McClure's original 2007 framework, and it duplicates what Acquisition already measures.

### What's a good referral rate to aim for?

There's no universal target. The rough rule of thumb is a viral coefficient above 1, meaning each active user brings in more than one new user through invites; below that, referral still helps, but it needs other channels to keep feeding it. See the Referral checkpoint above for where that threshold fits.