AARRR pirate metrics for data analysts

What AARRR is and why interviewers love it

AARRR (Pirate Metrics) is Dave McClure's framework that describes the user lifecycle through five stages: Acquisition, Activation, Retention, Revenue, Referral. The reason it survives is practical: it forces you to think about a product as a system with five distinct failure modes, not as one big funnel.

On a data analyst loop at Stripe, Airbnb, or DoorDash, the question is rarely "do you know what DAU means". It's a system-design question dressed up as a product question: "Imagine we're launching a new checkout flow — what metrics would you track?" The candidate who lists eight random KPIs loses. The candidate who opens with "let me walk you through the lifecycle in five stages" wins, because they've shown they think in structure first, metrics second.

If you already know what a conversion funnel looks like, AARRR is its extended version — the funnel covers the path to a single goal, while AARRR covers the entire lifecycle including retention, monetization, and virality.

Acquisition

Question: where do users come from, and what does each one cost?

This is the top of the lifecycle. A user discovers the product through a channel — paid search, an influencer post, organic search, an App Store listing — and takes a first action. The analyst's job at this stage is to attribute users to channels and price them.

Core metrics: new users by channel, CAC, landing-page conversion rate, and the cost-side metrics CPC, CPM, CPI. The trap that catches juniors is reporting blended CAC across all channels — a single number that hides the fact that influencer traffic costs $8 while paid search costs $42. Always break CAC down by channel.

A worked example: a food-delivery app runs Google Ads, influencer partnerships, and ASO. You want to compare CAC by channel before reallocating budget.

-- CAC by acquisition channel for March 2026
SELECT
    u.utm_source,
    COUNT(DISTINCT u.user_id)                        AS new_users,
    SUM(mc.ad_spend)                                 AS total_spend,
    SUM(mc.ad_spend)
        / NULLIF(COUNT(DISTINCT u.user_id), 0)       AS cac
FROM users u
LEFT JOIN marketing_costs mc
  ON mc.utm_source = u.utm_source
 AND DATE_TRUNC('month', mc.spend_date) = '2026-03-01'
WHERE u.created_at >= '2026-03-01'
  AND u.created_at <  '2026-04-01'
GROUP BY u.utm_source
ORDER BY cac;

Note that NULLIF guards against division by zero when a channel had spend but zero attributed users — a real edge case during paused campaigns.

Activation

Question: did the user reach first value?

Signup is not activation. Activation is the moment the user understands why the product exists — the so-called aha moment. The point of separating Acquisition from Activation is that you can have 80% signup conversion and 12% activation, and the second number predicts retention.

To find an aha moment: take a cohort of new signups, split into retained (came back on day 7) and churned. Compare the actions each group took in the first session. The action with the highest correlation with day-7 retention is your candidate. Then validate with a controlled experiment.

Famous examples: Facebook's "7 friends in 10 days", Slack's "2,000 messages sent in a team", Dropbox's "1 file in 1 folder on 1 device". The pattern: a specific action × a specific window × a specific count. Vague aha moments like "engaged with the product" are useless because you can't operationalize them.

Retention

Question: does the user come back?

Retention is the foundation. Without it, every dollar spent on acquisition is wasted, monetization is a one-shot lottery, and referrals are impossible. Retention is the metric that gates everything else.

Core metrics: Day 1 / Day 7 / Day 30 retention, DAU / WAU / MAU, Stickiness (DAU/MAU), and Churn Rate. The retention curve usually drops steeply in the first days, then flattens to a plateau — that plateau height is the long-run retention. If the curve never flattens and keeps falling toward zero, the product has no product-market fit. See how to calculate D1/D7/D30 retention in SQL.

A reference query for day-N retention by signup cohort:

-- Day-N retention for the March 2026 signup cohort
WITH cohort AS (
    SELECT user_id, DATE(created_at) AS signup_date
    FROM users
    WHERE created_at >= '2026-03-01'
      AND created_at <  '2026-04-01'
),
activity AS (
    SELECT DISTINCT user_id, DATE(event_time) AS active_date
    FROM product_events
)
SELECT
    (a.active_date - c.signup_date)                  AS day_n,
    COUNT(DISTINCT a.user_id)                        AS retained_users,
    COUNT(DISTINCT a.user_id)::FLOAT
      / (SELECT COUNT(DISTINCT user_id) FROM cohort) AS retention_rate
FROM cohort   c
JOIN activity a
  ON a.user_id    = c.user_id
 AND a.active_date >= c.signup_date
WHERE (a.active_date - c.signup_date) BETWEEN 0 AND 30
GROUP BY day_n
ORDER BY day_n;

Revenue

Question: do users pay?

Monetization turns engagement into money. The model dictates which metrics matter — subscription products care about MRR, e-commerce cares about AOV and GMV, gaming cares about ARPDAU.

Core metrics: ARPU / ARPPU, LTV, Conversion to Paid, MRR / ARR, AOV, ARPDAU. The most-quoted rule in growth circles is LTV / CAC ≥ 3. Healthy SaaS targets LTV / CAC of 3-5 with CAC payback under 12 months; consumer apps run thinner.

A trap that derails revenue dashboards: confusing ARPU (per all users) with ARPPU (per paying users). If a freemium product has 2% conversion to paid and $50 ARPPU, ARPU is $1 — reporting only ARPPU hides the real revenue picture. Always show both with conversion rate linking them.

Referral

Question: does the user bring others?

Virality is the cheapest channel because marginal cost is near zero. The metric is the viral coefficient (K-factor) = (invites sent per user) × (invite acceptance rate). When K > 1, the product grows exponentially. K > 1 is rare outside social products — but even a K of 0.3-0.5 materially lowers blended CAC.

Core metrics: K-factor, Referral Rate, Referral Conversion, and NPS as a proxy for willingness to recommend. K-factor is a flow metric — K = 0.4 over 30 days is very different from K = 0.4 over 6 months. Always state the window.

How to build an AARRR dashboard

An AARRR dashboard is five blocks, one per stage, each with two or three metrics that actually drive decisions. The minimum viable layout:

A dashboard is not a number wall — every metric must point to an action. If a metric doesn't trigger a decision, take it off the dashboard.

AARRR by product type

The framework is universal; the operational definitions are not. The fastest way to look junior is to use a generic AARRR template for a domain it doesn't fit. A short cheat sheet:

For e-commerce: Acquisition is site visits and installs; Activation is first product view or first add-to-cart; Retention is repeat purchase within 30 or 60 days; Revenue is AOV, GMV, and gross margin; Referral is share-to-friend and promo-code referrals.

For B2B SaaS: Acquisition is signup or demo request; Activation is first integration or first workspace populated; Retention is weekly active accounts and key-feature adoption; Revenue is MRR and trial-to-paid conversion; Referral is invite-a-teammate flow and customer-led expansion.

For mobile games: Acquisition is install attribution; Activation is tutorial completion and reaching level 5; Retention is D1 and D7 retention; Revenue is ARPDAU and IAP conversion; Referral is friend invites and achievement sharing.

Common pitfalls

The first pitfall is treating signup as activation. Teams celebrate signup conversion going from 60% to 75% while activation stays flat — meaning the extra signups never reach first value. The fix is to instrument a real aha moment and report signup and activation side by side. Activation is the leading indicator of retention; signup is not.

The second is blended CAC. Reporting a single CAC across channels averages a $6 organic-search user with an $80 paid-influencer user and obscures the channel mix decision. Always break CAC out by channel, and report paid-only CAC separately from blended. Marketing teams will hate this transparency at first and love it within a quarter.

The third is misusing DAU/MAU. The ratio is a decent stickiness proxy for daily-use products (messaging, social, news) but useless for weekly or monthly intent (taxes, payroll, real estate search). Forcing DAU/MAU on a weekly-use product produces a flat 5-8% that says nothing about product health.

The fourth is using cumulative retention instead of day-N retention. Cumulative retention (any return in days 1 through N) only goes up; classic day-N retention (active on day N specifically) is what reflects habit. If your retention curve never goes down, you're plotting the wrong curve.

The fifth is LTV without a window. Reporting "LTV = $340" without saying over what horizon is meaningless. For new products, default to 12-month historical LTV until you have enough history to extend the window credibly.

Interview questions

What is AARRR and why is it useful? The five stages of the user lifecycle — Acquisition, Activation, Retention, Revenue, Referral. The value isn't the acronym, it's the structure: instead of a random list of KPIs, you understand which stage is broken and which metrics monitor that stage. It also forces you to separate signup from activation, which is the single most common source of dashboard confusion.

How do you find the aha moment for a new product? Compare behavior between retained and churned users in the first session. Find the action with the strongest correlation to day-7 retention. Then test the hypothesis: if you push more users through that action, does retention move? Correlation is the discovery method; an A/B test is the validation method.

Which AARRR stage matters most? Retention. Without retention, acquisition is a leaky bucket, monetization is a one-shot lottery, and virality is impossible. Retention is the foundation that lets the other stages compound.

How do AARRR and the conversion funnel relate? The conversion funnel covers the path to a single goal — typically the Acquisition-to-Activation portion of AARRR. AARRR is broader: it adds retention, monetization, and virality. The funnel is a subset of AARRR.

Related reading

• How to calculate D1/D7/D30 retention in SQL
• How to calculate LTV in SQL
• How to calculate CAC in SQL
• Activation framework for product managers
• Aha moment, explained simply
• North star metric for PM

If you want to drill product-metric and SQL questions like this every day, NAILDD is launching with 500+ analyst problems across exactly this pattern.

FAQ

How is AARRR different from a North Star Metric?

AARRR is a five-stage framework that covers the entire user lifecycle. A North Star Metric is a single headline number that captures the core value the product delivers — for Airbnb it's "nights booked". They don't compete: the NSM usually sits inside one of the AARRR stages (most often Activation or Retention), while AARRR gives you the diagnostic metrics that explain why the NSM moves.

Should I build AARRR for an MVP?

Yes, but compressed. One metric per stage is enough — typically signup count, activation rate, day-7 retention, and a placeholder for monetization. Spend most analytical energy on Activation and Retention, since those determine whether the product has the right to exist. Revenue and Referral can wait until you have a population that returns.

Does AARRR work for B2B products?

Yes, with adaptation. B2B sales cycles are longer and the decision-maker is rarely a single user, so "Acquisition" is often two events — a marketing lead and a buyer signoff. Activation in B2B usually means a full integration, not a button click. Referral is conference recommendations and Slack-community word-of-mouth, not viral invite links.

How often should I review the AARRR dashboard?

Weekly for product and growth teams. Monthly is too slow — by the time you spot a D7 retention drop on a monthly review, you've already lost a month of users. The standard cadence is daily watch on Acquisition and Activation, weekly review of Retention and Revenue, monthly deep-dive on Referral and cohort health.

What's the relationship between AARRR and unit economics?

Unit economics live inside the Revenue stage but pull from Acquisition and Retention. LTV / CAC is the headline ratio: LTV comes from Revenue + Retention, CAC comes from Acquisition. A healthy target for subscription products is LTV / CAC ≥ 3 with CAC payback under 12 months. AARRR is the operational view; unit economics is the financial view of the same machine.