What is RFM analysis? Recency, frequency, monetary explained
RFM analysis (recency, frequency, monetary) is a customer segmentation method that ranks and groups customers by how recently they last bought, how often they buy and how much they spend. It uses order history you already have to find your best customers and those drifting away.
The three dimensions
- Recency: how long since the customer's last purchase, usually measured in days. Recent buyers are more likely to buy again.
- Frequency: how many purchases the customer made in a given period.
- Monetary: how much the customer spent in that period.
How scoring works
A common approach gives each customer a score from 1 to 5 on each dimension, with 5 the best (most recent, most frequent, highest spend). Scores are often assigned by ranking customers and splitting them into five equal groups. The three digits together, such as 5-5-5 or 1-2-1, form the customer's RFM cell. Implementations vary: some average the three scores, others weight them for the business. A retailer selling items people rarely repurchase might put less weight on frequency.
Hypothetical example:
| Customer | Last order | Orders (12 months) | Spend (12 months) | RFM |
|---|---|---|---|---|
| A | 5 days ago | 8 | $800 | 5-5-5 |
| B | 200 days ago | 6 | $600 | 1-4-4 |
| C | 10 days ago | 1 | $40 | 5-1-1 |
A is a top customer worth rewarding. B used to buy often but has gone quiet, a classic at-risk customer for a win-back campaign. C is new and might respond to a welcome series nudging a second order.
Why it matters for an online store
RFM turns a customer list into segments you can act on without complex modeling: loyalty offers for champions, win-back emails for lapsed buyers, and cross-sell suggestions for one-time buyers. It also helps explain what drives customer lifetime value.
Limitations
- It looks backward and says nothing about why customers behave as they do.
- Thresholds depend on your purchase cycle; "recent" for groceries is not "recent" for furniture.
- Scores go stale, so recalculate regularly and compare segments over time with cohort analysis.