RACE, AIDA, RFM, CLV-NPS, or AARRR: Which Framework Actually Solves Your Problem

Every marketing framework promises the same thing: a clearer picture of where the business is leaking. Five of them dominate small-business marketing analytics, and the mistake most founders make is not picking the wrong one, it is picking one before identifying the actual symptom it needs to diagnose. A conversion problem measured with a retention framework produces a technically correct report that answers a question nobody asked.

This is a decision guide, not a ranking. Each framework below solves a different problem, and the fastest route to a useful dashboard is matching the framework to the symptom already visible in the business, not building all five and hoping one turns out relevant.

---

Start From the Symptom

If the business symptom is…	The framework to run is…	Because it measures…
Traffic is fine, but people are not converting into buyers	AIDA	The single-buyer psychological journey: Attention, Interest, Desire, Action. Diagnoses exactly which step of one sales conversation loses people.
Ad spend keeps climbing, revenue is not keeping pace	RACE / SOSTAC	The full marketing calendar across channels: Reach, Act, Convert, Engage, planned against a structured SOSTAC strategy. Diagnoses whether the quarter’s plan is working, not just one page.
Total revenue looks stable, but you cannot say who is actually valuable	RFM	Recency, Frequency, Monetary value, scored per customer from transaction history alone. Diagnoses who to protect, who to win back, who to stop chasing.
Revenue is healthy, but cancellations or complaints are creeping up	CLV-NPS	Customer Lifetime Value crossed against Net Promoter sentiment. Diagnoses which high-value customers are emotionally checked out before the cancellation lands.
New customers sign up but do not come back	AARRR / RARRA	The five-stage growth funnel, Acquisition, Activation, Retention, Referral, Revenue, reordered around whichever stage is actually leaking. Diagnoses whether growth spend is building a customer base or renting one.

---

The Two Funnel Frameworks: AIDA and RACE

AIDA and RACE are often confused because both track a buyer moving toward a purchase, but they operate at different zoom levels.

AIDA zooms into one interaction: a single landing page, one ad creative, one sales call. It asks whether a visitor noticed the offer (Attention), engaged with it (Interest), wanted it (Desire), and completed the purchase (Action). AIDA is the right tool when a specific page or campaign is underperforming and you need to know exactly which of the four psychological steps is losing people.

RACE (paired with SOSTAC for the strategic layer underneath it) zooms out to the whole marketing operation across a quarter: every channel, every campaign, planned and measured against Reach, Act, Convert, Engage. RACE is the right tool when the question is not “why did this one page underperform” but “is the overall plan working.”

Run AIDA on a single underperforming asset. Run RACE on the calendar that asset lives inside.

---

The Two Customer Frameworks: RFM and CLV-NPS

RFM and CLV-NPS both look at customers who already bought, not new visitors, and both get confused for the same reason AIDA and RACE do: overlapping subject, different question.

RFM is built entirely from behavior already sitting in the transaction log: how recently, how often, and how much a customer bought. It requires no survey and produces a segment map (Champions, At Risk, Lost, and others) purely from purchase history. It is the faster of the two to stand up, since the data already exists.

CLV-NPS adds a layer RFM cannot see: sentiment. A customer can score well on RFM, buying recently, buying often, spending well, and still be a Detractor quietly building a case to leave the moment a competitor removes the switching friction. CLV-NPS requires running an NPS survey and crossing it against financial value, which takes more setup than RFM but catches the silent, high-value churn risk that pure purchase data misses entirely.

Most businesses should build RFM first, since it needs no new data collection, and add CLV-NPS once a lightweight NPS survey process exists.

---

The Growth Framework: AARRR / RARRA

AARRR and RARRA cover the same five stages, Acquisition, Activation, Retention, Referral, Revenue, in a different order. AARRR runs Acquisition first, which fits a business with proven retention that genuinely needs more top-of-funnel volume. RARRA runs Retention first, which fits the more common small-business situation: acquisition spend climbing while nobody has confirmed that new customers actually stick around long enough to justify it.

This is the framework to reach for when the symptom is specifically about new customers not returning, distinct from RFM’s focus on segmenting an existing base and CLV-NPS’s focus on sentiment-driven churn.

---

Building One Framework Instead of All Five

The failure mode with frameworks is not picking the wrong one, it is trying to stand up all five before any single one is fully built. That produces five half-finished spreadsheets and zero decisions. Pick the framework matching today’s loudest symptom from the table above, build it properly in a spreadsheet using data you already have, and only add the next framework once the first one is actually changing a weekly decision.

Each of the five frameworks above has a pre-built, low-code engine at Analytics Forge: the spreadsheet, the data intake form, and an AI prompt library that turns the numbers into a specific action, one framework per bundle, priced separately so you buy only the one that matches your actual symptom. Tell us what the business is doing wrong and we will point you at the right one.