RFM analysis: how it works and what to do with each segment
By Santiago Bazarra · February 16, 2023 · Updated November 7, 2026 · 7 min read
If you're looking for a way to better understand your customers and their purchasing behavior, RFM analysis is an excellent place to start. This powerful tool allows you to segment your customer base based on recency, frequency, and monetary value, giving you valuable insights into the behavior of different groups of customers.
What is RFM analysis?
RFM analysis is a data-driven approach to customer segmentation. By analyzing customer data in terms of Recency (how recently they made a purchase), Frequency (how often they make purchases), and Monetary value (how much they spend), you can gain insights into the behavior of different groups of customers.
How does RFM analysis work?
To perform RFM analysis, you need to collect data on your customers' purchase history, including the date of their last purchase, the number of purchases they've made, and the total amount they've spent. You can then rank customers on each of these metrics and assign them a score between 1 and 5 for each metric.
For example, a customer who made a purchase in the last week would be assigned a high score for recency, while a customer who hasn't made a purchase in several months would be assigned a low score. Similarly, a customer who has made many purchases and spent a lot of money would be assigned a high score for frequency and monetary value.
Once you've assigned scores to each customer, you can use these scores to segment your customers into different groups based on their behavior. For example, you might have a group of high-value customers who make frequent purchases, a group of low-value customers who make infrequent purchases, and a group of recent customers who haven't yet made a lot of purchases.
How can you use RFM analysis to improve your business?
Once you've segmented your customers using RFM analysis, you can use this information to tailor your marketing and sales strategies to each group. For example, you might create different types of content or promotions to target each group.
If you have a group of high-value customers who make frequent purchases, you might focus on providing them with personalized offers and promotions, and consider creating a loyalty program to reward them for their continued business. If you have a group of recent customers who haven't yet made a lot of purchases, you might focus on providing them with educational content that helps them understand the value of your products or services.
By using RFM analysis to segment your customers and tailor your marketing and sales strategies to each group, you can improve your overall customer retention and satisfaction, which can ultimately lead to increased revenue and growth for your business.
The RFM segments and what to do with each
Scores from 1 to 5 on three variables give 125 combinations, which in practice collapse into eleven named segments. The names vary by tool; the logic is always recency crossed with frequency, with monetary value used to prioritise within each group.
| Segment | Typical scores | Who they are | What to do |
|---|---|---|---|
| Champions | R 5 · F 4–5 | Bought recently, buy often, spend the most | Reward them, ask for reviews, early access. No discount needed |
| Loyal customers | R 3–5 · F 3–5 | Buy regularly and respond to promotions | Loyalty programme, cross-sell into categories they don't buy yet |
| Potential loyalists | R 4–5 · F 2–3 | Recent, more than one purchase | Strengthen the relationship: recommendations, loyalty invitation |
| New customers | R 5 · F 1 | First purchase very recently | Onboarding: secure the second purchase, the one that predicts retention |
| Promising | R 4 · F 1 | Bought recently, once | Give them reasons to return before recency drops |
| Need attention | R 3 · F 3 | Used to buy well, starting to space out | Timely reminder, limited-time offer, relevant content |
| About to sleep | R 2–3 · F 1–2 | Below average on recency and frequency | Reactivate now; it is the last cheap moment |
| At risk | R 1–2 · F 3–5 | Spent a lot, haven't bought in a while | Personalised campaign around what they used to buy; the most expensive loss |
| Can't lose them | R 1–2 · F 5 | Your best historical customers, now inactive | Direct contact, special treatment, find out what happened |
| Hibernating | R 1–2 · F 1–2 | Last purchase long ago, few purchases, low spend | Low-cost reactivation offer; if no response, stop |
| Lost | R 1 · F 1 | One purchase, long ago | Exclude from acquisition spend and stop paying to reach them |
Two things usually surprise a company the first time it sees its own table: champions are rarely more than 5–10% of the base yet account for a third to a half of revenue, and the largest segment is almost always lost plus hibernating. In retail it is common for half the base to have gone more than a year without buying.
How to run RFM analysis: spreadsheet or tool?
A first RFM analysis fits in a spreadsheet. Export twelve months of sales, group by customer, compute days since last purchase, number of orders and total spend, then score each variable 1 to 5 using quintiles (the top 20% get a 5, the bottom 20% a 1). Scoring relative to your own base beats fixed thresholds, because it adapts to your buying cycle.
The spreadsheet stops working when you want to repeat it monthly, when sales come from more than one system, or when segments have to reach your email tool or Meta without exporting files. That is when you need software that computes RFM continuously on unified customer data.
When comparing RFM analysis tools, three things matter more than the rest:
- It resolves identities before scoring. If the same person buys in-store and online and the two records are not merged, they appear twice, each half with a worse score than reality. A champion can show up as two promising customers. RFM is only reliable on unified data.
- Segments refresh on their own. A monthly export is a report, not a segmentation. The tool should recalculate scores as new orders arrive and move customers between segments automatically.
- Segments go straight to your channels. The output of RFM is an audience, and it should land in your email platform, ad accounts or CRM as a live sync, not a CSV.
A Customer Data Platform does exactly these three things, which is why RFM is usually the first analysis a company activates once it has one. Our own CDP computes RFM segments out of the box on unified profiles and syncs them to the channels where campaigns run; the same profiles also feed customer lifetime value and churn models, which add the forward-looking dimension RFM lacks.
Which martech tool would be needed to carry out RFM analysis properly?
By using a Customer Data Platform (CDP) you could effectively perform RFM analysis. A CDP is a software tool that allows you to collect, analyze, and segment customer data from multiple sources, including your website, social media channels, customer support interactions, or any other that you might have.
To perform RFM analysis using a CDP, you first need to integrate your customer data into the platform by connecting your website analytics tool, email marketing platform, CRM system, and any other database with customer information available. Once integrated, the CDP will automatically assign scores to each customer based on their behavior in terms of Recency, Frequency, and Monetary value.
Using the insights gained from RFM analysis, you can segment your customers into different groups and tailor your marketing and sales strategies to each group. By using a CDP to perform RFM analysis, you can gain a more comprehensive understanding of your customers' behavior, make informed business decisions, and improve your overall performance.
Frequently asked questions about RFM analysis
What does RFM stand for? Recency, Frequency and Monetary value: when the customer last bought, how often they buy and how much they spend. It is a method for segmenting customers by purchasing behaviour.
How many RFM segments are there? With scores from 1 to 5 there are 125 combinations, usually grouped into ten or eleven named segments: champions, loyal, potential loyalists, new, promising, need attention, about to sleep, at risk, can't lose them, hibernating and lost.
How often should RFM be recalculated? At least monthly; weekly in high-frequency businesses. What matters is using the same method each time, so that movement between segments is comparable.
What is the best tool for RFM analysis? For a one-off, a spreadsheet. For ongoing use, any platform that unifies customers across channels before scoring, refreshes segments automatically and syncs them to your marketing channels. Customer Data Platforms are built for this; general analytics or BI tools can compute the scores but rarely activate them.
Does RFM work for B2B? Yes, and it works well: business customers usually have clear repurchase patterns. You need to adapt the period (buying cycles are longer) and treat the account, not the contact, as the customer.
What is the difference between RFM and CLV? RFM looks at the past and classifies; CLV estimates how much a customer will be worth in the future. They complement each other: RFM tells you which segments you have today and CLV how much it is worth investing in each one.