AARRR Model: Find the Leak Before You Spend More on Acquisition
The AARRR Model, also called pirate metrics, is Dave McClure’s 2007 funnel of Acquisition, Activation, Retention, Referral, and Revenue. It helps you measure each stage, find where people drop off, and work that stage before you buy more traffic.
Framework Card
- Name:
- AARRR Model
- Goal:
- Measure Acquisition, Activation, Retention, Referral, and Revenue, then work the stage that is leaking before buying more traffic.
- Flow:
- Acquisition → Activation → Retention → Referral → Revenue
- Best For:
- Growth that is flat while acquisition spend keeps rising; A signup spike that does not become a second visit; Separating a referral problem from a revenue problem
Why it matters
When growth is flat, the default move is more ads. If people arrive and never see value, or see value once and never return, more Acquisition makes the leak more expensive.
AARRR forces a walk of the five stages with measures. You can then say “Activation is the leak” instead of “we need more top of funnel.” It does not prove product-market fit, and it does not replace a full unit-economics model.
What it is
AARRR stands for:
| Stage | Job |
|---|---|
| Acquisition | How people arrive |
| Activation | First experience that shows value (not merely a signup) |
| Retention | They come back |
| Referral | They bring others |
| Revenue | They pay or otherwise monetize |
The useful output is a leak: which stage is weakest for this product and this goal. Referral is a measured stage. It is related to advocacy talk, but it is not 5A’s Advocate letter.
Commonly cited company stories (referral extra storage, freemium onboarding) are illustrations of possible stage work. They are not measurements this page should treat as yours.
AARRR is not 5A (Aware, Appeal, Ask, Act, Advocate). It is not Hook (trigger, action, reward, investment).
How it works
A useful pass is measured stages plus a bottleneck, not a pirate-metrics poster.
1. Name the product and the conversion
One product, one paid action (or the monetization you actually use). Mixing several products blurs the funnel.
2. Put a measure on each stage
Counts, rates, or even rough ranges from the case. If a stage has no measure, label it unknown.
3. Find the leak
Where do people drop relative to the goal? A strong Acquisition number with a weak Activation rate is a different problem from strong Retention and no Referral.
4. Work that stage before more Acquisition
State the implication. Buying more traffic into a broken Activation is a method error for AARRR.
How it compares
When another lens fits better, these frameworks do different jobs.
| Framework | What it helps you see | How it differs from AARRR |
|---|---|---|
| 5A Marketing Model | Aware, Appeal, Ask, Act, Advocate | Path letters. AARRR is measured funnel stages. |
| Hook Model | Habit loop | Repeat-use design. AARRR measures Retention as a leak. |
| 6 Types of Marketing Campaigns | Campaign jobs | Classification. AARRR diagnoses the funnel. |
The AARRR Model is the lens for a five-stage leak. Other methods help when the question is a path, a habit, or a campaign type.
When to Use This Framework
- Spend up, revenue flat. Acquisition may not be the leak.
- Signups without a second session. Activation or Retention is the likely leak.
- Users love it but do not pay or refer. Revenue or Referral is the stage to name.
Example
A concrete example makes the structure easier to reuse when you are under uncertainty.
Example: Ads fill the top, first-run fails
A SaaS team doubles paid search. Signups rise. People never complete the setup that shows the product working. Week-two return is low.
Implication: The leak is Activation (and then Retention). More Acquisition spend would scale a broken first run. 5A might still be useful later for Ask; AARRR’s job here is the measured first-value gap.
Example: Retention without revenue
A consumer app is used weekly. Almost nobody pays. Referral is also thin.
Implication: Acquisition is not the first job. Name Revenue (and whether Referral is even in the model). Inventing a Dropbox-style referral rate would be a method error.
Takeaway
What AARRR can help with
- Putting measures on five funnel stages
- Finding the leak
- Stopping more Acquisition into a later-stage break
- Separating Referral from Revenue
What AARRR cannot replace
- 5A. Path letters including Ask and Advocate.
- Hook Model. Habit-loop design.
- Six campaign types. Classification of campaign jobs.
- Full unit economics. LTV and CAC may sit beside AARRR; they are not the five letters.
- A growth guarantee. Diagnosis is not a lift.
Honest scope: AARRR structures a leak reading. It should not be sold as a data-driven growth engine.
Frequently asked questions
Not necessarily. Activation is the first experience that shows value. A signup with no value moment is still an Activation leak.
A named product, a measure or labeled gap for each of the five stages, a stated leak, and an implication that does not default to more ads. If the output only lists AARRR, you do not have a diagnosis yet.
Advocate in 5A is a path stage of post-purchase talk. Referral in AARRR is a measured funnel stage of bringing others in. Related, not the same model.
No. Those ratios can sit beside the funnel. AARRR still works as five stages and a leak. Missing unit economics is a labeled gap, not a reason to invent numbers.
When Retention or Activation is the leak, more Acquisition usually wastes spend. If Acquisition itself is empty, that can be the leak. Name the bottleneck from measures, not from a slogan.