AARRR vs. RARRA: Which Growth Funnel Order Actually Fits a Small Business

Pirate Metrics made AARRR famous for a reason: Acquisition, Activation, Retention, Referral, Revenue is an easy framework to remember and it covers the real shape of a growth funnel end to end. It also has a structural flaw that has cost small businesses real money: it puts Acquisition first, which trains founders to treat “get more traffic” as the default answer to a slow quarter, even when traffic was never the problem.

RARRA exists because that mistake was common enough to name. Reorder the same five letters to Retention, Activation, Referral, Revenue, Acquisition, and the framework forces a different question before any ad spend gets approved: does what we already have actually hold onto people? If the honest answer is no, acquiring more customers does not grow the business. It funds a bigger version of the same leak.

---

The Leaky Bucket, in Numbers

A subscription box business runs $8,000 a month in acquisition spend and brings in 400 new customers. Leadership reads that as growth: customer count is up every month. The AARRR-ordered dashboard agrees, since Acquisition is the first number reviewed in the meeting.

Run the same funnel in RARRA order, Retention first, and the picture changes:

Stage	Metric	Result
Retention	% of new customers still active at day 60	22%
Activation	% who complete first box setup within 7 days	61%
Referral	% who refer a second customer	3%
Revenue	Average revenue per active customer	$34/month
Acquisition	New customers per month	400

78% of every new customer this business pays to acquire is gone within two months. The $8,000 acquisition spend is not building a customer base, it is renting one for eight weeks at a time. Total customer count keeps climbing every month for the same reason a bathtub with the drain open can still look full: the tap is running faster than the drain, right up until the tap slows down.

---

Why the Order Changes the Decision

The five stages are identical in both frameworks. What changes is which stage gets budget and attention first, and that changes the whole conversation:

• AARRR-first instinct: revenue is flat, so run more ads, run a promotion, widen the top of the funnel.
• RARRA-first instinct: revenue is flat, so check whether the customers already inside the funnel are sticking around before spending another rupee to bring in more of them.

Neither instinct is wrong in every situation. A genuinely under-marketed business with strong retention and no traffic problem is exactly where AARRR’s acquisition-first order is the correct call. The mistake is applying that order by default, without ever checking Retention first, which is what most small businesses do because Acquisition is the easiest stage to measure and the one every ad platform is built to report on.

---

Diagnosing Which Stage Is Actually Broken

Before deciding which framework fits, pull four numbers over a fixed window, typically the last 90 days:

1. Acquisition: visitors, leads, or trial sign-ups.
2. Activation: the percentage who complete one defined first action (first purchase, first login plus one core feature used, first booking).
3. Retention: the percentage still active at your business’s natural repurchase window (7 days for a daily-use app, 30-60 days for a subscription box, 12 months for an annual service).
4. Referral and Revenue: referral rate and revenue per active customer.

Lay the four stages side by side and look for the steepest percentage drop between adjacent stages. That drop, not the stage with the smallest raw number, is where the leak actually is. A business with 10,000 monthly visitors and 40 sales does not necessarily have an Acquisition problem, if Activation is where 9,200 of those visitors disappear, the fix lives on the landing page and onboarding flow, not the ad budget.

---

Building the Diagnostic Without a Growth Team

None of the four measurements above require a specialized analytics platform:

• Acquisition and Activation numbers usually already sit in your ad platform and site analytics, they just need to be pulled into one sheet instead of read separately in two dashboards.
• Retention requires one cohort table: group customers by the month they first purchased, then track what percentage of each cohort is still active in each following month. A pivot table with a COUNTIFS formula builds this in under an hour.
• Referral and Revenue per active customer are two columns added to the same sheet, sourced from your billing or CRM export.

That single sheet, rebuilt monthly, tells you which of the five stages to fund next quarter, and which framework order actually matches your business instead of the one that happens to spell a catchier acronym.

The manual version above is the diagnostic. Running it continuously, with the cohort table rebuilding itself against fresh data and an AI prompt library that turns “Retention dropped to 19% this cohort” into a specific three-move fix, is what the Growth Engine Tracker at Analytics Forge is built to do. It is currently in production, built on the same AARRR/RARRA stage logic above, for founders who want the retention-first diagnostic running on autopilot rather than rebuilt by hand every quarter. Tell us which stage looks broken and we will let you know when it is ready.

Growth is not a single number going up. It is five stages, and the order you check them in decides whether the business you are building actually holds the customers it pays to bring in.