Data Analytics: Foundations to Practice · Segmentation and Cohort Analysis
RFM Segmentation
RFM is one of the most widely used customer segmentation frameworks precisely because it requires minimal data and produces genuinely actionable groups. This chapter covers how it works.
RFM segmentation groups customers along three behavioral dimensions calculated from their own transaction history: Recency (how recently a customer last made a purchase or engaged), Frequency (how often a customer purchases or engages within a given time period), and Monetary value (how much a customer has spent in total), each dimension reflecting a genuinely distinct aspect of customer behavior that, together, paint a fuller picture than any single dimension alone could [3].
Key Takeaways
- RFM segments customers by Recency (how recently), Frequency (how often), and Monetary value (how much spent), each capturing a distinct behavioral aspect.
- Each dimension carries distinct predictive value: recent purchasers are more likely to buy again, frequent purchasers are more loyal, higher spenders represent more realized value.
- Customers are commonly scored on each dimension (often via quintiles) and combined into named segments like Champions, At Risk, and Lost.
- RFM's limitations include ignoring non-transactional information, bias toward recently acquired customers, and a poor fit for long-replacement-cycle categories.