Understanding RFM

Overview

RFM stands for Recency, Frequency, Monetary. It is a way of grouping and segmenting customers and is often used in direct marketing and database marketing. The three dimensions — Recency, Frequency and Monetary — are easy ways to segment customers into groups based on their measured behavior.

• Recency: How recent was the last purchase

• Frequency: How many purchases are made over the customers’ lifetime

• Monetary: How much is spent (on average, per order) over the customer’s lifetime

There is no cut and dry rule for how exactly to segment customers. Many believe that 3-5 ranges for each dimension is sufficient and that more than 5 would be too many. The goal for setting your ranges is to have a fairly even (but not exact) distribution of customers across each segment.

Example

Let’s use Recency as an example. Many companies will consider a customer who purchased in the last 12 months to be active and outside of that, they are lapsed. A standard way to segment Recency would be to determine if their last purchase was:

• 0-12 months ago

• 13-24 months ago

• 25-36 months ago

• 37+ months ago

This gives 4 possible segments that a customer could fall into. However, if your business has a much lower purchase frequency or much higher purchase frequency, you may wish to adjust the ranges to reflect this. Perhaps a high frequency business would choose ranges such as the following:

• 0-3 months

• 4-6 months

• 6-12 months

• 13-24 months

Each dimension (i.e., R, F, or M) should have options that some of your customers fall into. The goal is that each customer receives a score based on their activity, within each dimension. In the example above, every customer would have a score of 1-4 because there are 4 ranges. If 80% of customers fall into only one of the ranges, you may want to reconsider the grouping so that you can more clearly separate your best customers from your worst customers.

RFM score using segments

Many companies will simply take the score from each of the three segments and put them together. For example, if Customer A was...

• segment 1 for Recency

• segment 3 for Frequency

• segment 5 for Monetary

...that customer’s score would be 135.

RFM score calculation

There is also a singular score that was developed that combines the effect of each of the three segments. The formula is:

(1/R) * F * √M

The numeric values in this equation are:

R = the number of months since the last purchase

F = total purchases made

M = total monetary value

Each customer is scored using this equation and then ranked in order from the best customers to the worst customers.

Customers are then divided into deciles: 10 equal groups.

Daasity uses both of these approaches outlined above. This information can also be found in the Customers Explore and Dashboards.

The RFM Decile Grouping illustrates the gross margin value per customer, on average, for each group of customers. The top group is your best (i.e, most valuable customers), and the bottom group is your worst (least valuable, or most costly customers).