RFM analysis guide — customer segmentation that makes promotions hit the mark
How RFM analysis works, how to score it, and how to turn customer segments into promotions that actually land.
RFM is one of the simplest customer segmentation methods that is still relevant today: three numbers, one very practical business question — who should I promote what to?
TL;DR
- RFM scores customers on three dimensions: Recency (when they last bought), Frequency (how often), and Monetary (how much they are worth).
- Combining the three scores sorts customers into actionable segments — Champions, Loyal, At Risk, and so on.
- The value is not the label but different treatment per segment: reactivate the at-risk, keep the loyal, and steer first-time buyers toward their second purchase.
- You can start from a simple spreadsheet — see the RFM Analytics Template for a ready-to-use version.
Three numbers that answer many questions
Recency answers: how fresh is the customer’s relationship with your business. Someone who bought a month ago is far more likely to buy again than someone whose last purchase was a year ago.
Frequency answers: how strong the habit is. A customer who bought five times in six months is in a different place from one who bought once — even at the same value.
Monetary answers: how much value they bring. Usually this is total spend over a chosen period, say the last 12 months.
Each customer gets a 1–5 score on all three dimensions. The combination, such as 5-5-5, is the customer’s position on the map: the Champions segment. One whose recency used to be good but has gone cold, say 1-4-4, is At Risk — a valuable customer starting to drift.
From segments to treatment
Segmentation becomes useful when each segment gets a different treatment:
| Segment | Signal | Treatment that usually fits |
|---|---|---|
| Champions | 5-5-5 | Early access to new products, referral programs |
| Loyal | high frequency | Appreciation vouchers, upgrade bundling |
| At Risk | used to buy often, now quiet | Re-activation campaigns with a limited-time incentive |
| New customer | one purchase so far | A nudge toward the second purchase, product onboarding |
| Hibernating | long silent | A “why did you leave” survey, low-cost exploration promos |
The most common mistake is treating all customers the same — one promo blast for everybody. RFM breaks that assumption using data you already have: your transaction history.
Start with the smallest thing
You don’t need a data team to begin. Three columns — transaction date, customer ID, transaction value — are enough. From there, RFM scores and segments can be computed in a spreadsheet.
If you’d rather skip ahead, our RFM Analytics Template already computes scores and segments from your transaction data automatically. And if your business sells on marketplaces, start where margin leaks the most: pricing per product.