RFM analysis

RFM analysis is a simple, transaction-based way to score and segment customers using three dimensions: Recency (how long since their last purchase), Frequency (how often they buy), and Monetary value (how much they have spent). It needs nothing more than an order history to compute, which is why it has stayed popular since long before modern customer segmentation tools existed.

Each customer is typically ranked on each dimension, often split into quintiles scored 1 to 5, and the three scores are combined into a cell such as "555" for a customer who is recent, frequent and high-spending. Businesses then name resulting groups, for example "champions," "at risk," or "lapsed," and target each with a different message or offer.

RFM matters because it is fast to build, easy to explain to non-technical stakeholders, and effective at flagging valuable customers who have gone quiet before they fully churn, feeding directly into cohort analysis and retention work. It works best for businesses with repeat, trackable purchases, and is a weaker fit for one-time or subscription products. Its main limitation is that it describes past behavior without explaining why it changed, and it says less about a customer's future customer lifetime value than a proper predictive model.