Strategy & economics
RFM segmentation
Also known as: RFM analysis · recency-frequency-monetary
A customer-segmentation framework that scores each customer on Recency, Frequency, and Monetary value — three ordinal dimensions that together classify the customer base into behavioural bands.
RFM (Recency, Frequency, Monetary) segmentation is the oldest serious customer-scoring model in lifecycle marketing. Recency = days since last purchase or engagement. Frequency = count of transactions in a fixed window. Monetary = total spend in the same window. Each dimension is binned into ordinal scores (typically 1-5), and the combination produces bands like "champions" (555), "loyalists" (544), "at-risk" (155). RFM is simple to compute, works without any ML, and maps cleanly to lifecycle triggers. Its weakness: it's backward-looking — it describes what a customer did, not what they'll do. Modern programs pair RFM with propensity models for forward-looking scoring. For most teams, RFM is the right starting model — build it first, prove the programs work off the bands, then layer in predictive signals.