OM Enterprises Group
Digital MarketingNov 20, 2026

RFM Customer Segmentation: How to Sort Your Customers and Send Messages That Actually Work

RFM Customer Segmentation: How to Sort Your Customers and Send Messages That Actually Work

Most businesses treat their customer list as one big group. The same SMS goes to someone who bought yesterday and someone who last purchased two years ago. The same offer goes to a customer who spends heavily every month and one who bought once with a coupon. The result is wasted messages, rising opt-outs and discounts given to people who would have paid full price. A simple method called RFM analysis fixes this by sorting customers according to three behaviours that predict future value. You need no special software, only your sales records. This guide shows how to use RFM customer segmentation to send better messages.

What RFM stands for

Recency. How recently did the customer last buy? A customer who purchased last week is more likely to buy again than one who disappeared a year ago.

Frequency. How often does the customer buy? Repeat buyers are more loyal and more profitable than one-time purchasers.

Monetary value. How much has the customer spent in total or on average? High spenders deserve extra attention.

Together, these three numbers describe the relationship in a way that raw contact lists cannot.

Step one: gather your data

Export customer purchase data from your store, billing software, POS or CRM. For each customer, you need a unique identifier such as phone number or email, the date of each order and the order value. Clean the data first: merge duplicates, remove test orders, handle refunds and returns correctly and exclude internal accounts. Choose a time window that fits your business, such as the last twelve or twenty-four months. A grocery shop may care about weeks, while a furniture store may look at years.

Step two: calculate the three numbers

For recency, calculate the days since each customer's last purchase. For frequency, count the number of orders in the time window. For monetary value, add up total spend, or use average order value if that suits your business better. A spreadsheet is enough. Sort and compare the figures to see how your customers are spread out.

Step three: score each dimension

Rank customers on each measure and divide them into groups, commonly five tiers from 1 to 5, where 5 is best. For recency, customers with the most recent purchases get 5. For frequency and monetary value, the highest get 5. Quintiles, which split the customers into five equal-sized groups, are a common approach, but you can also set thresholds that make sense for your business, such as scoring frequency 1 for one purchase and 5 for six or more. Each customer ends up with a three-digit code, such as 555 for the best customers or 211 for those who bought once long ago and spent little.

Step four: group them into practical segments

Hundreds of possible codes are too many to manage, so combine them into a handful of segments with clear actions.

Champions. Bought recently, buy often and spend the most. Reward and involve them.

Loyal customers. Buy regularly with good spend. Keep them engaged and encourage reviews and referrals.

Potential loyalists. Recent customers with moderate frequency. Nurture them toward repeat purchases.

New customers. Bought recently, only once. Focus on a great first experience and a second purchase.

Promising. Recent but low spend. Introduce higher-value products or bundles.

Need attention. Above-average scores that are starting to slip. Re-engage before they drift away.

At risk. Used to buy often and spend well, but have not purchased for some time. Contact them with personal win-back efforts.

Cannot lose them. Your biggest spenders who have gone quiet. Prioritise direct outreach, even a call.

Hibernating or lost. Low scores on all three. Try a final, low-cost reactivation message, then reduce effort.

Step five: plan actions for each segment

Champions. Early access to new products, exclusive previews, loyalty rewards, requests for reviews and referral invitations. Avoid heavy discounts, since they already buy.

Loyal customers. Cross-sell complementary products, personalised recommendations and membership benefits.

New customers. Welcome series, usage tips, delivery updates and a gentle prompt for a second order with a small incentive if margins allow.

At risk. A friendly "we miss you" message with a useful reason to return, such as new arrivals or a restock, and a reminder of what they liked before.

Cannot lose them. Personal phone call or direct message from the owner or account manager, asking for feedback and offering a solution.

Hibernating. Run a low-cost reactivation campaign by email or WhatsApp where consent exists. If there is no response after a few attempts, pause and keep the contact for rare, high-value announcements.

Choose channels by segment

Your most valuable segments justify richer channels such as personalised WhatsApp messages, calls or tailored email. Mid-value segments can receive automated sequences. Low-value segments should get low-cost, low-frequency messages. Respect consent for each channel, and never contact customers on WhatsApp for marketing without opt-in. Use SMS for time-sensitive reminders within DLT rules.

Adapt RFM to your business

A subscription business may replace frequency with tenure or renewals. A B2B company may use order value and contract length. A restaurant may measure visits per month. A service business may look at appointments and spend. The principle is the same: use behaviour, not guesswork. Revisit scoring every quarter, since customer patterns change.

Measure the outcome

Track response rate, repeat purchase rate, revenue per message, opt-out rate and movement between segments, such as the share of new customers who become loyal or at-risk customers who return. Compare targeted campaigns with your old blanket messages. If revenue per message rises while opt-outs fall, your segmentation is working.

Common mistakes

Common errors include dirty data, ignoring returns, giving discounts to champions, treating all lost customers equally, building too many segments, never refreshing scores and forgetting to connect segments to actual campaigns.

Frequently asked questions

Do I need software for RFM?

No. A spreadsheet works for small customer bases, and many CRMs offer RFM tools for larger ones.

How many customers do I need?

Even a few hundred customers can be segmented, though larger datasets give clearer patterns.

How often should I update scores?

Monthly or quarterly, depending on purchase cycles.

Send the right message to the right customer

OM Enterprises helps businesses build customer segments and run targeted WhatsApp, SMS and email campaigns. Talk to our CRM and messaging team.