ModeesMost small senders do not have a segmentation problem. They have a spreadsheet with a last-purchase date, an order count and a total spend, and no decision about what those numbers mean. RFM — recency, frequency, monetary value — is the smallest useful answer to that question.
It is not new, and it is not clever. It is useful because it turns three raw columns into four groups with different messages, and it fits in one export.
Score each column
For every contact who has bought:
- Recency — days since the last purchase. Lower is better.
- Frequency — number of purchases in the last 12 months.
- Monetary — total spend in the last 12 months.
Score each from 1 to 5 using quintiles of your own list, not fixed thresholds. A list where the best customer spends £200 needs different buckets from one where they spend £20,000. Quintiles keep the groups comparable as the list grows.
A simple score of 4–5 on each gives you a 3–15 range. Do not over-engineer the arithmetic; the value is in the ordering, not in a false precision.
Four groups, four jobs
- Champions (high R, high F, high M): recent, frequent, valuable. Do not discount. Ask for a review, offer early access, or invite them into a referral programme. Their job is advocacy.
- Loyal (good R and F, lower M): regular but modest spend. Grow the basket with a relevant add-on or a bundle, not a blanket percentage off.
- At risk (low R, historically high F or M): they used to buy and stopped. This is the group where a personal, specific message earns its keep: "you bought X in spring; here is what changed since". Do not open with a coupon.
- Lapsed (low R, low F, low M): a reactivation campaign with a clear reason to return, or suppression. Sending the same newsletter forever costs money and drags engagement down.
Two more groups are worth naming if you have the data: new (one purchase, recent) and promising (recent, low frequency, high value). They need onboarding and a second-purchase nudge respectively, not a loyalty offer.