AARRR pirate metrics: the five-letter funnel every growth team still runs.

What AARRR pirate metrics actually is

AARRR pirate metrics is a five-stage funnel for tracking how customers move through your product. The five stages are Acquisition, Activation, Retention, Referral, and Revenue. Dave McClure named them this way at 500 Startups in 2007. Each stage has its own metric. Each metric has its own diagnostic. The whole point is to make funnel-stage problems impossible to miss.

The framework exists because growth teams overspend on the top of the funnel. Acquisition is the easiest stage to measure. Cost-per-click. Cost-per-install. Click-through rate. Every ad platform reports it. Activation, retention, referral, revenue: those numbers live deeper inside the product, and they require analytics work to surface. So teams measure what is easy and ignore what is hard. AARRR is the explicit decision to look at all five stages with equal seriousness.

The five-letter shape is what makes the framework sticky. AARRR is the sound a pirate makes. It is easier to remember than a list of acronyms or a process diagram. Almost two decades later, a senior growth lead can walk into a new company and say "show me your AARRR dashboard" and the team knows exactly what is being requested.

Claim: Dave McClure's 2007 SlideShare deck "Startup Metrics 4 Pirates" has been viewed millions of times and is widely cited as the original source of the AARRR framework. Source: SlideShare archive, Dave McClure / 500 Startups (2007). Context: The deck spread through the 500 Startups community and the broader Silicon Valley accelerator world. Every subsequent essay on funnel optimization points back to it.

Where Dave McClure invented it

Dave McClure published "Startup Metrics 4 Pirates" as a SlideShare deck in 2007. The deck spread through the early startup community. Brian Balfour, Andrew Chen, Casey Winters, and Lenny Rachitsky have all extended the framework in essays since. The original deck has been viewed millions of times and is still the reference cited in every subsequent funnel essay.

McClure was running 500 Startups at the time. The accelerator was working with dozens of seed-stage companies a year. Every founder was building a dashboard, and every dashboard looked different. McClure wrote AARRR to give the portfolio a shared vocabulary. Five letters. Five stages. One conversation about which stage was broken.

The framework had two structural moves that mattered. First, it broke the funnel into stages that were measurable inside the product, not just at the ad platform. Acquisition was at the platform. Activation, retention, referral, and revenue were inside the product. Second, it put each stage in the order most teams should think about it. Acquisition first because that is where new users enter. Revenue last because that is the result, not the lever.

The deck closed with a position McClure made loudly. Most founders, he argued, spend too much time on the top of the funnel and too little on the four stages below. Activation is where most products fail. Retention is where unit economics actually live. Referral is the cheapest acquisition channel any company has. Revenue is the result, not the lever. The order of the letters is also the order most teams optimize them in reverse.

"Most startup founders spend too much time obsessing over acquisition. The four stages after acquisition are where the leverage actually lives." Dave McClure — Startup Metrics 4 Pirates, SlideShare (2007)

The five stages, in order

The five stages of AARRR are Acquisition (how users find you), Activation (when they first get value), Retention (when they come back), Referral (when they invite others), and Revenue (when the business gets paid). Each stage has a headline metric. Each metric has a typical conversion rate to the next stage. Each stage has a different team or owner.

Acquisition

How users find your product. Search, ads, content, referral, partnerships, word of mouth. The headline metric is usually cost per acquired user, ranked by channel. The diagnostic question is which channels deliver users who go on to activate, not just which channels deliver the cheapest signup.

Activation

The moment a new user gets value. For Slack it was 2,000 messages sent by a team. For Dropbox it was the first file uploaded and shared. For Airbnb it was the first completed booking. The headline metric is the percent of acquired users who hit the activation event in their first session or first week. This is the single most leverage-rich stage in the funnel.

Retention

When users come back. The headline metric is some shape of D7 or D30 retention: percent of activated users who return in the seven days or thirty days after activating. Cohort curves matter here more than blended averages. A flat retention curve is the holy grail. A retention curve that falls to zero is a one-and-done business.

Referral

When users bring other users. The headline metric is the viral coefficient or k-factor: the number of new users each existing user creates. Dropbox's give-and-get referral program is the most-cited example. The metric most teams ignore. Referral is the cheapest acquisition any company has, and ignoring it means leaving free growth on the table.

Revenue

When the business gets paid. The headline metric depends on the model. For DTC it is contribution margin per order. For SaaS it is ARR per account. For a marketplace it is take rate times volume. Revenue is the result of the first four stages working. If revenue is broken and the first four stages are healthy, the problem is usually pricing, not the funnel.

The five AARRR stages with example metrics across four industry shapes

Stage	DTC ecommerce	B2B SaaS	Consumer app	Marketplace
Acquisition	Site visits, signups	Demo requests, MQLs	App installs	Guest signups
Activation	First purchase	First workflow run	First session > X minutes	First completed booking
Retention	Repeat purchase in 90 days	Weekly active in workflow	DAU / MAU > 20%	Repeat booking in 6 months
Referral	Referral code uses	Team-invite sends	Friend invites	Share rate
Revenue	Contribution margin per order	ARR per account	Subscription, ad revenue	Take rate × volume

Claim: Slack identified that teams sending 2,000 or more messages had dramatically higher long-term retention than teams that sent fewer. The 2,000-message threshold became Slack's activation metric inside its AARRR dashboard. Source: First Round Review interview with Stewart Butterfield (2017). Context: Every product change between 2014 and 2017 at Slack was evaluated against how quickly it moved new teams past the 2,000-message mark. The activation event drove product roadmap decisions for years.

Why the order matters (and the RARRA counter)

McClure put Acquisition first because that is where new users enter the product. But the order most teams should optimize their AARRR funnel is almost the reverse. Casey Winters and others at Reforge have argued for a re-ordering they call RARRA: Retention, Activation, Referral, Revenue, Acquisition. The case is simple. Acquiring users into a leaky product is a tax on the ad budget. Solve the leak first.

The RARRA case has three parts. First, retention is the hardest stage to move and the longest to measure. Building a product that retains takes quarters of work, not weeks. Get the work started early. Second, activation feeds retention. Until the activation event is right, retention will always look broken. Third, referral compounds retention. Users who retain are users who refer; users who churn never get to the referral stage at all.

The original AARRR letter order still has a defense. New companies have to acquire some users before they can measure activation or retention at all. You cannot run a cohort analysis with no cohorts. McClure's order is the bootstrap order. RARRA is the optimization order. Both are useful at different stages of a company's life.

Claim: Casey Winters at Reforge has argued in published essays that for many product-led-growth businesses, the order of AARRR should be reversed to RARRA because acquisition without retention is destructive to long-term growth. Source: Reforge published essays (2019-2024). Context: The counter-framing is not a replacement but a re-prioritization. RARRA argues that retention should be solved before acquisition is scaled, because acquisition without retention just churns through ad budget. The two letter orders coexist in modern growth practice.

How to instrument AARRR (5 steps)

Here is the 5-step framework I use to put AARRR into a working dashboard. The work takes about a week, plus a month of validating the activation event against real cohort data. The most common skipped step is step 2 — teams pick an activation metric that feels right instead of running the cohort analysis.

1. Define each stage's headline metric.Write one specific metric per stage, customized to your product. For Spotify the A-1 is account signup; the A-2 is first listening session over 30 seconds. For Airbnb the A-1 is guest signup; the A-2 is first completed booking. The verb matters more than the noun.
2. Find the activation moment that predicts retention.Run a cohort analysis. Look at every in-product behavior that happens before retention. The behavior with the highest tie to month-2 retention is your activation event. Mixpanel and Amplitude both ship cohort tools for this. In a warehouse it is a window function over user events grouped by signup cohort.
3. Build the five-stage dashboard.One number per stage. One owner per stage. No vanity metrics. Every number must be a leading indicator of the next stage. If a metric on the dashboard can move up while the next-stage metric falls, it does not belong on the dashboard.
4. Set conversion rates and benchmarks.Pick the conversion rate between every two adjacent stages. Acquisition to Activation. Activation to Retention. And so on down the funnel. Compare your rates to industry benchmarks. The stage with the worst gap-to-benchmark is your weakest link.
5. Run the diagnostic every week.The stage with the lowest stage-to-stage conversion rate is the binding constraint. Ship work against that stage. When the constraint moves to a different stage, ship work against the new constraint. Repeat. That is the whole operating model.

Claim: Across the roughly 50 growth audits we run per year at Real Growth Matters, the most common AARRR diagnosis is broken Activation, not broken Acquisition. Teams overspend on the top of the funnel because the bottom of the funnel feels harder. Source: Real Growth Matters Inc., internal audit data, 2024-2026. Context: Acquisition is the easiest stage to measure and the easiest one for spending money. Activation requires product work, cohort analysis, and changes inside the application. Teams reach for the easier lever, the funnel breaks downstream, and they buy more ads to compensate. The framework's whole value is forcing the team to look at the stage where the leverage actually lives.

The three failure modes most teams hit

Three failures show up in almost every AARRR audit I run. None of them are about the framework. All three are about how teams choose to use the framework. Catching them before the team commits to a metric is the difference between a working funnel and a working slide deck.

Failure mode 1: vanity acquisition

The team measures total signups and calls it Acquisition. Then it spends more on ads to drive signups. Signups climb. Retention goes flat. Revenue stays the same. The dashboard looks great. The business does not grow. The fix is to define Acquisition not by signup count but by signup count weighted by activation rate. A signup that does not activate is not really a customer; it is a name in the database.

Failure mode 2: missing or wrong activation event

The team either has no activation metric on the dashboard, or has one that does not predict retention. The fix is the cohort-correlation analysis from step 2 above. Look at every in-product behavior. Find the one that predicts retention. Make that the activation event, full stop. Slack ran this analysis and landed on 2,000 messages sent per team. Dropbox ran it and landed on first file shared. Your number will be different.

Failure mode 3: ignoring referral entirely

Most teams treat Referral as a nice-to-have. They do not measure it. They do not build the referral mechanic into the product. They leave the cheapest possible acquisition channel on the table for years. Dropbox's referral program is the most-cited example of what referral can do when the team takes it seriously. Most B2B SaaS companies could be running team-invite mechanics that drive 20-40% of new account creation. They are not, because they have not measured.

Claim: Dropbox's give-and-get referral program reportedly drove 2.8 million direct referral invitations in a 30-day window in 2008, materially contributing to the company's growth. Source: Industry reporting and Dropbox S-1, SEC EDGAR (February 2018). Context: The program became the most-cited example of the R-for-Referral stage of AARRR. It worked because the reward was the product itself (free storage), the invite mechanic had low friction, and both sides of the transaction received value. Modern teams rarely match that design quality, but the model is well documented.

AARRR vs north-star vs OKRs vs OMTM

AARRR sits in a neighborhood of growth frameworks that overlap and complement each other. The north-star metric is the single number above the funnel. OKRs are the quarterly objectives that hang off the north star. AARRR is the decomposition lens for diagnosing where the funnel breaks. OMTM (One Metric That Matters) is AARRR's nearest cousin. Below is how to decide which one your team should run.

How AARRR compares to north-star metric, OKRs, OMTM, and growth loops

Framework	Scope	Time horizon	Best fit
AARRR pirate metrics	Five-stage funnel decomposition	Continuous, diagnostic	Almost every growth team at almost every stage
North-star metric	One company-level outcome that predicts value	12-18 months	Scale-stage companies aligning teams on a shared target
OKRs	Quarterly objectives with 3-5 measurable key results	One quarter	Larger orgs coordinating across many teams
OMTM (One Metric That Matters)	The single metric for the company's current stage	3-9 months	Early-stage startups whose metric will change as they cross PMF
Growth loops	Recursive flywheel structure across the funnel	Continuous	Product-led growth companies with viral or content-driven acquisition

These frameworks are not alternatives. A mature growth team runs all of them at once. The north star sits above the AARRR dashboard. The OKRs are how each AARRR stage gets its quarterly target. The OMTM is the dial the team is moving this quarter, usually one specific input under one specific AARRR stage. Growth loops are the structural model for why the funnel compounds in the first place.

When AARRR breaks down

AARRR works for most consumer and B2B product businesses. It breaks down in a few specific cases where the five-stage funnel is not the right shape for the business. Knowing where it does not apply is as important as knowing where it does. Below are the three shapes of business where the framework fits poorly.

Marketplace businesses. Two-sided marketplaces have two AARRR funnels running in parallel, one for each side. The supply side and the demand side have to be measured separately, and one side often gates the other. Airbnb cannot drive guest activation if there are no hosts. A single five-stage funnel does not capture this.

Enterprise sales businesses. Long sales cycles, multi-stakeholder decisions, and contract-based revenue do not fit the AARRR funnel cleanly. Acquisition is qualified-lead generation, not signup. Activation is the post-sale onboarding period. Retention is renewal. The shape is right but the metrics are different, and the time horizon is years, not weeks.

Network-effect products in cold-start. A product whose value depends on other users (social, marketplaces, collaboration tools) cannot show activation to early users until the network is built. Andrew Chen has written extensively on this in The Cold Start Problem. The AARRR funnel becomes meaningful only after the network crosses a critical mass.

How it fits with adjacent concepts

AARRR is the funnel decomposition. The frameworks around it do the work AARRR enables. North-star metric is the one number you optimize. CAC payback and LTV are the unit-economics check. Cohort analysis is the diagnostic tool you reach for inside the Retention stage. Jobs-to-be-done is the framing that tells you what Activation should be. Growth loops describe the recursive shape of the funnel when it is working.

For commerce companies, AARRR pairs with CAC payback and the LTV ratio to test whether the funnel actually makes money. A funnel that converts well but loses money on every customer is not a working funnel. For B2B SaaS, AARRR pairs with the north-star metric as the single target the whole funnel is driving. For network-effect products, AARRR pairs with network effects and the cold-start problem.

The intellectual heritage runs through Alistair Croll and Benjamin Yoskovitz's 2013 book Lean Analytics, which extended the AARRR vocabulary with their One Metric That Matters concept. Clayton Christensen's jobs-to-be-done framing sits adjacent and asks the question AARRR cannot: what value is the user actually trying to receive? AARRR is the structural decomposition. JTBD is the why behind it.

Quick answers about AARRR pirate metrics

What does AARRR stand for?

Acquisition, Activation, Retention, Referral, Revenue. Five stages of a customer funnel, named by Dave McClure at 500 Startups in 2007.

Why is it called pirate metrics?

Because the five letters read aloud as AARRR, the sound a pirate makes. The name made the framework sticky and easy to remember.

Who created AARRR?

Dave McClure. He introduced the framework in a SlideShare deck called Startup Metrics 4 Pirates in 2007 while running 500 Startups.

Is AARRR or RARRA the right order?

Both are useful. AARRR is the bootstrap order, useful when you are still acquiring your first cohort. RARRA is the optimization order, useful once you have enough users to measure retention.

Which stage matters most?

Almost always Activation. The user behavior that predicts retention is where the leverage lives. Most teams underinvest here because measuring it requires product work, not ad spend.

Is AARRR still relevant in 2026?

Yes. The vocabulary has evolved with north-star metrics, growth loops, and jobs-to-be-done, but AARRR is still the default decomposition of a customer funnel. Almost every growth team runs some version of it.

Frequently asked

What is AARRR pirate metrics?

AARRR is a five-stage funnel framework for growth teams. The stages are Acquisition, Activation, Retention, Referral, and Revenue. Dave McClure named it pirate metrics at 500 Startups in 2007 because the letters read aloud as AARRR. The framework is still the standard funnel diagnostic almost two decades later.

Who invented AARRR?

Dave McClure introduced AARRR pirate metrics in a 2007 SlideShare deck titled Startup Metrics 4 Pirates while he was running early-stage investing at 500 Startups. The deck spread through the Silicon Valley accelerator world and is still the founding reference for the framework.

What does each letter in AARRR stand for?

Acquisition is how users find you. Activation is when they first experience the value of your product. Retention is when they come back and keep using it. Referral is when they invite other users. Revenue is when the business gets paid. Each letter is its own stage with its own metric.

What is RARRA and how is it different from AARRR?

RARRA reorders AARRR to put Retention first: Retention, Activation, Referral, Revenue, Acquisition. The argument, advanced by Casey Winters and others at Reforge, is that acquisition without retention burns money. Solve retention first and the rest of the funnel works.

What is an activation metric in AARRR?

The activation metric is the behavior a new user completes that predicts they will retain. For Slack it was 2,000 messages sent by a team. For Dropbox it was the first file shared with another person. The activation metric is the trip wire that separates a user who will stick from a user who will churn.

Is AARRR still relevant in 2026?

Yes. The framework is almost twenty years old and still the diagnostic lens most growth teams reach for first. The vocabulary has evolved (north-star metric, jobs-to-be-done, growth loops), but AARRR remains the default decomposition of a customer funnel.

How do I instrument AARRR for my product?

Five steps. Define one specific metric per stage. Find your activation event through cohort analysis. Build a five-stage dashboard with one owner per stage. Set conversion rates between every two stages. Run the diagnostic weekly and ship against your weakest stage.

What are the most common AARRR failures?

Three failures show up in almost every audit. Vanity acquisition metrics that look big but do not predict retention. Missing or wrong activation events. And ignoring referral entirely. Teams overspend at the top of the funnel because the bottom of the funnel feels harder to solve.