RFM analysis: a simple way to split your customers into groups

RFM analysis is a simple method for sorting customers by their buying behaviour. It uses three numbers that almost every business already has: Recency (how recently the customer bought), Frequency (how often) and Monetary value (how much they spent). You do not need special software. A spreadsheet is enough.

Why use it

• Stop sending the same offer to everyone.
• Find your best customers and protect them.
• Spot customers who are drifting away while you can still win them back.
• Spend less on messages that do not work.

How the three scores work

• R, recency: days since the last purchase. Recent is better.
• F, frequency: number of orders in a period. More is better.
• M, monetary value: total spent in the same period. More is better.

Each customer gets a score, for example from 1 to 3 or from 1 to 5, for each of the three. A customer with 3-3-3 is your top customer. A customer with 1-1-1 has been silent for long, bought once and spent little.

Step by step in a spreadsheet

1. Export the data: customer, date of each order and order amount, for the last 12 months.
2. Calculate R: days between today and the customer's last order.
3. Calculate F: the number of orders per customer.
4. Calculate M: the total amount per customer.
5. Set the limits. Sort each column and split it into equal thirds (or fifths). The most recent third gets R=3, the least recent gets R=1. The same for F and M.
6. Combine the three scores into a code such as 3-2-3.
7. Name the groups and decide what to do with each.

Limits depend on your business. A shop that sells groceries will think in days, a company that sells machinery will think in years. Use your own numbers, not someone else's.

Groups and what to do with them

• Champions (high R, F and M): thank them, give early access to news, ask for reviews and recommendations. Do not discount, they buy anyway.
• Loyal (high R and F, medium M): offer larger bundles or upgrades to raise the average order.
• Promising newcomers (high R, low F): help them make a second purchase with a useful follow-up or a small incentive.
• At risk (used to buy often, now quiet): a personal message, a survey or a reason to return. This is the most valuable group to act on.
• Sleeping (low R, low F): one reactivation email with a clear offer.
• Lost (very low R, F and M): a last attempt, then remove them from regular mailings to protect your sender reputation.

Example. An online shop with 4,000 customers found that 300 "at risk" customers had once bought three or more times, but not in the past five months. A personal message with a short questionnaire and a small bonus brought back about a fifth of them, which paid more than the whole month's advertising.

Tips

• Recalculate every month or quarter. Customers move between groups.
• Keep the number of groups small, six to eight is enough.
• Test each message on a part of the group before sending to everyone.
• Remember consent: send marketing only to people who agreed to receive it.

RFM is not perfect, because it only looks at the past. But it is quick, free and almost always shows you money you are leaving on the table.