EN
Marketing

How to Do Customer Segmentation with RFM Analysis: Scoring, Segments, Segment Revenue, and Action Plan Training

How do you do customer segmentation with RFM analysis? Scoring, five segments, segment revenue, and an action plan for each segment: an 8-lesson e-commerce guide with examples.

Contents 16 sections
  1. What will you be able to do by the end of this training?
  2. Lesson 1: What is RFM analysis?
  3. Lesson 2: How do you prepare the data?
  4. Lesson 3: How do you score customers?
  5. Lesson 4: How do you form segments?
  6. Lesson 5: How much revenue do the segments generate?
  7. Lesson 6: What action for each segment?
  8. Lesson 7: How big is the monetary opportunity of segmentation?
  9. Lesson 8: How do you keep segments updated regularly?
  10. How do you calculate RFM step by step in a spreadsheet?
  11. Which wrong assumptions break RFM analysis?
  12. What are the most common mistakes in RFM segmentation?
  13. How does NetKasam make customer segmentation easier?
  14. Mini quiz: Test what you have learned
  15. Conclusion: Move customers from numbers to segments
  16. Frequently Asked Questions About RFM Analysis

Your customer list has a thousand people. If you send everyone the same campaign message, a customer who has shopped every month for years and a customer who bought once and never came back get the same treatment. The first gets an unnecessary discount, and the second never hears about anything made for them. Customer segmentation is a way of grouping customers by behavior and treating each group differently, and its simplest, most practical method is RFM analysis.

This article is not a lecture; it is an 8-lesson hands-on training. We will answer the question how to do customer segmentation with RFM analysis using an example e-commerce store with 12 months of data and 1,000 customers: the meaning of RFM, scoring, forming segments, segment revenue, action for each segment, the opportunity calculation, and regular updates. At the end there is a mini quiz with an answer key. All figures are example values chosen to demonstrate the method; use your own order data.

What will you be able to do by the end of this training?

In short: You will score your customers on three criteria, split them into five segments, calculate each segment's revenue, and define one action per segment.

Training mapSix steps of RFM segmentation
  1. 01
    Extract the dataGet the last order, number of orders, and total amount per customer.
  2. 02
    ScoreScore each criterion from 1 to 3.
  3. 03
    GroupReduce the score combinations to five segments.
  4. 04
    ValueCalculate each segment's revenue.
  5. 05
    Turn into actionSet a single goal for each segment.
  6. 06
    UpdateRe-run the segments every three months.

The 8 lessons of the training: (1) what RFM is, (2) data preparation, (3) scoring, (4) segments, (5) segment revenue, (6) action per segment, (7) opportunity calculation, (8) regular monitoring.

Recommended ReadingHow to Prepare a Wholesale and Retail Price List: Wholesale Discount, Quantity Tiers, Minimum Order, and Customer Group Pricing Training

Lesson 1: What is RFM analysis?

In short: RFM defines a customer by three behaviors: when they last purchased, how often they purchase, and how much they spend.

  • Recency. Days since the last purchase. A customer who purchased recently is more likely to buy again.
  • Frequency. The number of orders in a given period. A customer who buys often is attached to the brand.
  • Monetary (amount). Total spend in the period. A high-spending customer contributes more to revenue.

These three criteria alone answer the question “who is valuable and who is about to disappear” and require no extra data: order history is enough. To calculate a customer's lifetime value separately, see how to calculate customer lifetime value, and for customer-level profitability, see how to calculate profitability by customer.

Lesson 2: How do you prepare the data?

In short: From the last 12 months of orders, extract three numbers for each customer: days since the last order, number of orders, and total amount.

CustomerLast order (days ago)Number of ordersTotal amount
A2057.500 ₺
B14034.200 ₺
C151450 ₺
D2601300 ₺

When preparing the data, pay attention to three points: deduct canceled and returned orders, merge multiple records of the same person (different email, different channel), and evaluate test orders and corporate bulk orders separately. Use a consistent key such as phone number or email to identify the customer. For keeping account records in order, see how to track customer and supplier accounts.

Lesson 3: How do you score customers?

In short: Set thresholds for each criterion and give a score from 1 to 3; define the thresholds according to your own sales rhythm.

Criterion3 points2 points1 point
Recency (last purchase)0–30 days31–90 days91 days and above
Frequency (number of orders)4 and above2–31
Amount (total spend)₺3,000 and above1.000–2.999 ₺Below ₺1,000

These thresholds are an example; for frequently bought low-value products, move the thresholds down, and for rarely bought expensive products, move them up. For example, for customer A, recency is 3 (20 days), frequency is 3 (5 orders), and amount is 3 (₺7,500), so the score is 3-3-3. For customer D it is 1-1-1. Customer B (140 days, 3 orders, ₺4,200) is 1-2-3: once valuable but silent for a long time.

FormulaCore rules of RFM segmentation
  1. Recency = Today − Last order date (days)
  2. Frequency = Number of orders in the period
  3. Amount = Total spend in the period, net of returns
  4. RFM score = Recency score + Frequency score + Amount score (3–9) or a three-digit code (e.g., 3-3-3)

Lesson 4: How do you form segments?

In short: Reduce the 27 possible combinations to five understandable segments; give each segment a simple definition and name.

SegmentDefinitionCustomer
ChampionRecent, frequent, and high-value80
LoyalBuys regularly, medium or high value170
PromisingRecent new customer with a single purchase220
At riskPreviously frequent and valuable, silent for a long time150
LostSilent for a long time, low frequency and amount380
Total1.000
DistributionDistribution of 1,000 customers by segment
  1. Champion80 customers
  2. Loyal170 customers
  3. Promising220 customers
  4. At risk150 customers
  5. Lost380 customers

38% of customers (380) are in the lost segment: those who placed one or two orders and never returned. This is a normal picture; the share of one-time customers is high in e-commerce. What matters is noticing the segment before they go silent, that is, the at-risk customers. To measure what share of customers buy again, see how to calculate and increase repeat purchase rate.

Lesson 5: How much revenue do the segments generate?

In short: Multiply the average spend per segment by the number of customers; you will see that a small group of customers carries a large share of revenue.

SegmentCustomerAvg. annual spendRevenue
Champion804.500 ₺360.000 ₺
Loyal1702.000 ₺340.000 ₺
Promising220450 ₺99.000 ₺
At risk1501.500 ₺225.000 ₺
Lost380300 ₺114.000 ₺
Total1.0001.138.000 ₺
DistributionDistribution of 12-month revenue by segment
  • Champion₺360,00031.6%
  • Loyal₺340,00029.9%
  • Promising₺99,0008.7%
  • At risk₺225,00019.8%
  • Lost₺114,00010%

The champion and loyal segments, although they make up 25% of customers (250), carry 61.5% of revenue (700,000 ÷ 1,138,000). Retaining these two segments is generally cheaper than finding new customers; for the cost of a new customer, see how to calculate customer acquisition cost.

Lesson 6: What action for each segment?

In short: Give each segment one goal and one message; instead of sending the same campaign to everyone, act according to need.

SegmentTargetAction
ChampionRetainThank-you, early access, news about new products; no discount needed
LoyalGrowComplementary product suggestions, basket growth
PromisingSecond purchaseA follow-up after the first order and a related product suggestion
At riskWin backA “We miss you” message and a time-limited offer
LostTest at low costA bulk, inexpensive reminder; if there is no response, remove from the list

Giving a discount to the champion segment is mostly unnecessary: these customers are already buying. Direct the discount budget to the at-risk segment. For campaign design, see how to set up a store campaign, and to measure a campaign's effect on sales, see how to measure the sales drop after a campaign. Messages are sent only to customers who have given contact permission.

Lesson 7: How big is the monetary opportunity of segmentation?

In short: Assume reasonable conversion rates for the targeted actions and calculate the extra revenue; the result is an estimate, not a guarantee.

If 10% of the 150 at-risk customers (15 people) return and their average annual spend is ₺1,500, the extra revenue is 15 × 1,500 = ₺22,500. If 20% of the 220 promising customers (44 people) make a second purchase and spend ₺450 per purchase, the extra revenue is 44 × 450 = ₺19,800. Total extra revenue is 22,500 + 19,800 = ₺42,300.

ChartEffect of segment actions on 12-month revenue
  1. Current 12-month revenue₺1,138,000
  2. At risk: 15 customers return₺22,500
  3. Promising: 44 customers make a second purchase₺19,800
  4. Revenue after actions₺1,180,300

Prepared with example values. Calculation: 1,138,000 + 22,500 + 19,800 = ₺1,180,300. The conversion rates are assumptions; campaign costs must be deducted from these amounts.

Extra revenue is not profit: deduct the cost of discounts, messages, and labor. To see profitability by segment, see how to analyze profitability by product. To increase the average basket at the same time, the article how to calculate and increase average basket value is a useful complement.

Lesson 8: How do you keep segments updated regularly?

In short: Customers move between segments; refresh the analysis every three months and track the transitions.

FrequencyWhat happens
Every monthSend actions to the at-risk and promising segments and record the result
Every three monthsRecalculate RFM and compare segment counts and revenue
Every yearReview the score thresholds and segment definitions
IndicatorHow it is calculatedSample value
Top two segments' share(Champion + Loyal) ÷ Total customers%25
Top two segments' revenue share700.000 ÷ 1.138.000%61,5
Win-back rateReturned ÷ At risk10% (target)
Second-purchase rateMade a second purchase ÷ Promising20% (target)
Takeaways from the exampleFour numbers from the RFM segmentation example
  • 25%share of champion and loyal customers
  • 61.5%revenue share of these two segments
  • 150number of at-risk customers
  • ₺42,300estimated extra revenue from the two actions

To add the results to a management report, you can use the layout in how to prepare a weekly management report. To handle this together with customers who abandon their carts, see how to calculate and reduce cart abandonment rate.

How do you calculate RFM step by step in a spreadsheet?

In short: Summarize the order list by customer, apply a score formula to each criterion, and map the scores to segments; for a thousand customers it takes a few minutes.

  1. Order list. Export the last 12 months of orders with customer key, date, and amount columns; remove returns and cancellations.
  2. Customer summary. For each customer, find the last order date, the number of orders, and the total amount (a pivot table does this). The difference between the last order date and today is the recency in days.
  3. Score columns. In three new columns, enter 1, 2, or 3 according to the thresholds. A conditional formula does this in a single line.
  4. Segment column. Convert the score combination into a segment name: for example, if recency is 3 and frequency is 3, it is a champion; if recency is 1 and frequency is 2–3, it is at risk.
  5. Summary. Pull out the number of customers and the total amount per segment; your table will look like the example in this article.

Once you set this up, next quarter you only paste in the new order list and refresh the result. Do not forget to filter the segment list to customers with contact permission before sending messages. To verify the result, look at which segment a few customers you know landed in: if a loyal customer you know appears as at risk, the thresholds are wrong.

Which wrong assumptions break RFM analysis?

In short: Product cycles, one-off bulk orders, and data merging errors mislead the segments; set thresholds according to your product's purchase rhythm.

  • Purchase rhythm. If you sell a product bought once a year, 90 days of silence does not mean lost. Set the recency threshold according to your product's normal repurchase cycle.
  • Bulk orders. If a corporate customer leaves a large amount in a single order, they may look like a champion. Track these customers separately.
  • Campaign effect. Customers who buy during a heavy campaign period may go quiet in the next period; read the results together with the period effect.
  • Duplicate records. If the same person appears with two different records, frequency comes out low and the lost rate comes out high. Merge the customer under a single key.
  • Channel difference. A customer coming from a marketplace may have limited contact information; check in advance which customers you can reach for messaging and follow the permission rules.

For this reason, RFM is not a final decision but a prioritization tool that shows which customers to look at first. Test the results not just as numbers but against customer examples you know: look at which segment a few customers you know landed in, and if a segment definition makes no sense, fix the thresholds.

Recommended readingWhat Is Dead Stock, How Do You Detect It and How Do You Clear It? Training on Inventory Aging, Holding Cost and a Liquidation Plan

What are the most common mistakes in RFM segmentation?

  1. Sending the same campaign to every customer. Segment differences are ignored.
  2. Giving champions a discount. It creates unnecessary cost for a customer who is already buying.
  3. Not deducting returns and cancellations. Amount and frequency come out higher than they are.
  4. Creating too many segments. With more than five segments, running a separate action for each is hard.
  5. Never updating the thresholds. The sales rhythm changes and segments go stale.
  6. Sending messages without permission. It creates legal risk and loses customers.
  7. Not measuring the result. If the win-back and second-purchase rates are not tracked, you cannot know whether the action worked.

How does NetKasam make customer segmentation easier?

  • Customer and supplier account tracking. You keep customer records and transactions in one place (customer and supplier account tracking).
  • Single-screen order management. You see orders from different channels by customer (order management).
  • e-Commerce integration. Orders from your own site enter the same flow (e-commerce integration).
  • Online sales. You manage your own sales channel (online sales).
  • Saved reports. You save order and account reports with their criteria and export them to Excel; you can prepare your RFM table from this data (reports).
  • Bookkeeping is free. e-Invoice, customer and supplier accounts, inventory, cash, bank and expense management are free in every plan; see the pricing page for details.

Mini quiz: Test what you have learned

In short: Answer the 5 questions below, then compare with the answer key. 4 or 5 correct shows you have completed the training.

  1. If there are 80 champions and 170 loyal customers, what percentage of the 1,000 customers is that?
  2. If the 80 customers in the champion segment spend an average of ₺4,500, what is the segment's revenue?
  3. If the champion and loyal segments' revenue is ₺700,000 and total revenue is ₺1,138,000, approximately what percentage is their revenue share?
  4. If 10% of the 150 at-risk customers return and spend ₺1,500 a year, what is the extra revenue?
  5. If 20% of the 220 promising customers spend ₺450 on a second purchase, what is the extra revenue?

Answer key:

  1. (80 + 170) ÷ 1.000 = %25.
  2. 80 × 4.500 = 360.000 ₺.
  3. 700.000 ÷ 1.138.000 ≈ %61,5.
  4. 15 × 1.500 = 22.500 ₺.
  5. 44 × 450 = 19.800 ₺.

Frequently Asked Questions About RFM Analysis

What is RFM analysis?

It is a segmentation method that scores customers by when they last purchased, how often they purchase, and how much they spend, and then groups them.

What data do you need for RFM?

For each customer, the last order date, the number of orders, and the total amount are enough. Returns and cancellations should be deducted.

How many segments should I create?

Five segments are enough for most businesses: champions, loyal, promising, at risk, and lost. More makes management harder.

How are score thresholds determined?

By your product's normal repurchase cycle and average order value. If the product is bought once a year, set the recency threshold longer.

Should I give a discount to a champion customer?

Generally no; this customer is already buying. Attention without a discount, such as a thank-you and early access, is more appropriate.

How do you win back an at-risk customer?

With a personal message based on their previous purchases and a time-limited offer. Measure the result by the share who return.

How often should the analysis be refreshed?

Every three months. Because customers move between segments, old segments go stale quickly.

Can I message all customers?

No. Send messages only to customers who have given contact permission; check your own messaging consent rules and the applicable regulations.

Conclusion: Move customers from numbers to segments

The most useful thing you can do today is extract three numbers per customer from the last 12 months of orders: days since the last order, number of orders, and total amount. Give three scores, split them into five segments, and look at the number of at-risk customers. The result of a single message you send to this group will show you the value of segmentation with your own data.

Bookkeeping Is Now Completely Free

Manage your bookkeeping and e-Invoice operations with NetKasam without paying a thing. Customer and supplier accounts, inventory, collections and more in one panel, completely free.

Start FreeOnline Sales

Read similar articles

19 free tools

Commission, shipping, VAT and profit calculations in your browser, no sign-up needed.

All tools