Using RFM Analysis for Effective Customer Segmentation in Marketing

Discover:

1. What is RFM Analysis?
2. How to calculate RFM Scores
3. How to Use RFM Scores
4. A Case Study: RFM Analysis in a Marketing Agency and its application in Fraud Detection

💡I write about Machine Learning on Medium || Github || Kaggle || Linkedin. 🔔 Follow “Nhi Yen” for future updates!

What is RFM Analysis?

In fraud detection in marketing analysis, segmenting customers into homogeneous groups can help detect fraudulent activities. Understanding the traits of each group and engaging them with relevant campaigns is more effective than segmentation based on customer age or geography.

RFM stands for Recency, Frequency, and Monetary

One of the popular and effective methods used to analyze customer behavior is RFM analysis. RFM stands for Recency, Frequency, and Monetary value, and each factor corresponds to a key customer trait. These RFM metrics can indicate fraudulent activities because frequent purchases and high monetary value could suggest suspicious transactions.

For businesses that lack monetary data, engagement parameters such as bounce rate, visit duration, and time spent per page can help identify fraudulent activities. This leads to a variation of RFM called RFE, where the Engagement parameter is a composite value based on these metrics.

The following points illustrate how RFM analysis can be used to identify customer behavior patterns that can help detect fraud:

• Recent purchases: customers who make recent purchases are more likely to respond to promotions.
• Frequency of purchases: customers who frequently buy products are more engaged and satisfied.
• Monetary value: the monetary value of purchases can differentiate between heavy spenders and low-value purchasers.

Implementing Customer Segmentation with RFM Analysis

Marketers use RFM analysis to answer these questions:

• Who are the top customers?
• Which customers may leave?
• Who could be valuable customers?
• Which customers can be kept?
• Who is most likely to respond to campaigns?

To demonstrate how RFM analysis works, let’s consider a sample dataset of customer transactions.

Table 1: Example Customer transactions dataset

Based on their transactions, Table 1 displays the recency, frequency, and monetary values for 15 customers.

RFM Scores Calculation

1. Rank each R, F, M attribute separately

Let’s score the customers in this example by ranking them based on each RFM attribute separately. Assuming a ranking system from 1 to 5 using RFM values, we’ll start with recency and display the resulting rankings in the table below.

The table above shows customers sorted by recency, with the most recent purchasers at the top. Customers are assigned scores from 1 to 5, with the top 20% receiving a recency score of 5 (customers 12, 11, 1), the next 20% (customers 15, 2, 7) receiving a score of 4, and so on.

We can also sort customers by frequency, from most to least frequent, assigning the top 20% a frequency score of 5, and so on. For the monetary factor, the top 20% of customers (big spenders) will receive a score of 5 and the lowest 20% a score of 1. The resulting frequency and monetary scores are summarized below:

2. Calculate RFM Score

To arrive at an aggregated RFM score, we can combine the individual rankings of R, F, and M for each customer. The resulting RFM score, shown in the table below, is the average of the individual R, F, and M scores, with equal weights given to each RFM attribute.

How to Use RFM Scores

When analyzing customer purchase behavior, RFM (Recency, Frequency, Monetary) scores are often used to segment customers into different categories based on their spending habits. However, the question arises: “Is it fair to average out individual R, F, and M scores for each customer and assign them to an RFM segment based on their purchase or engagement behavior?”

Depending on the nature of your business, you may need to increase or decrease the importance of each RFM variable to arrive at the final score. For example:

• In a consumer durables business, monetary value per transaction is normally high but frequency and recency are low. In this case, marketers could give more weight to monetary and recency aspects rather than the frequency aspect.
• In a retail business selling fashion/cosmetics, a customer who searches and purchases products every month will have a higher recency and frequency score than monetary score. Accordingly, the RFM score could be calculated by giving more weight to R and F scores than M.
• For content apps like Hotstar or Netflix, a binge watcher will have a longer session length than a mainstream consumer watching at regular intervals. For bingers, engagement and frequency could be given more importance than recency, and for mainstreamers, recency and frequency can be given higher weights than engagement to arrive at the RFE score.

This simple approach of scaling customers from 1–5 will result in at most 125 different RFM scores (5x5x5), ranging from 111 (lowest) to 555 (highest). However, it’s overwhelming for marketers to analyze all 125 segments individually, and it’s difficult to visualize this imaginary 3D cube. Therefore, the monetary aspect of RFM is viewed as an aggregation metric for summarizing transactions or aggregate visit length, and these 125 RFM segments are reduced to 25 segments by using just R and F scores.

RFM Model — A case study

A marketing agency uses recency and frequency scores to analyze customer behavior. The scores are then displayed on a 2D graph to make it easier for users to understand. To simplify the results, the agency combined some segments instead of creating 25.

Implement RFM Model in Marketing — Source: Clever Tap

The RFM grid above provides the following information for each segment:

• A brief description of the segment
• Recency (last activity)
• Frequency (activity count)
• Average monetary value
• Reachability of users across different channels

To understand the behavior of users in each segment, and recommend effective marketing strategies, we need to interpret the RFM segments.

Let’s examine a few interesting segments and learn more about them:

• Champions: These are your best customers who have made recent, frequent, and high-value purchases. It’s important to reward them for their loyalty as they can become early adopters for new products and help promote your brand.
• Potential Loyalists: These are your recent customers with an average frequency of purchases and a good amount spent. To encourage them to become your loyalists or champions, consider offering them membership or loyalty programs or recommending related products to upsell.
• New Customers: These are customers with a high overall RFM score but not frequent shoppers. To build a relationship with them, provide onboarding support and special offers to increase their visits.
• At Risk Customers: These are customers who have made frequent, high-value purchases in the past but haven’t purchased recently. To reconnect with them, send personalized reactivation campaigns and offer renewals and helpful products to encourage another purchase.
• Can’t Lose Them: These are customers who used to visit and purchase frequently but haven’t been visiting recently. To bring them back, offer relevant promotions and run surveys to find out what went wrong and avoid losing them to a competitor.

In addition, here are some case studies on how to use RFM analysis to detect frauds in marketing:

Case Study 1: Detecting fraudulent orders with RFM analysis

• A company used RFM analysis to identify fraudulent orders on their e-commerce platform.
• They found customers with high scores in all three RFM categories: recent purchase, high frequency, and high monetary value.
• They discovered some customers making multiple large purchases in a short time frame, indicating unusual behavior.
• Further analysis revealed that some customers were using stolen credit cards for purchases.
• The company used RFM analysis to prevent fraudulent orders and protect their revenue.

Case Study 2: Identifying suspicious behavior with RFM metrics

• A company used RFM analysis to identify suspicious behavior in their customer database.
• They identified customers with high scores in all three RFM metrics but made purchases at odd hours or locations.
• They found that some customers were using VPNs or proxies to conceal their true identity or location.
• Analysis of these customers’ order histories and IP addresses revealed fraudulent activities, such as creating fake accounts to game the system.
• The company took action against these suspicious accounts, preventing further fraudulent activities.

Case Study 3: Preventing coupon fraud with RFM segmentation

• A retail company used RFM segmentation to prevent coupon fraud in their marketing campaigns.
• They segmented customers based on their RFM scores and offered targeted discounts to each group.
• They discovered some customers using fake or duplicate coupons to get discounts.
• Analysis of these customers’ RFM scores revealed low scores in all three categories.
• The company limited discounts to customers with high RFM scores, rewarding their loyal and high-value customers while preventing coupon fraud.

In conclusion, RFM analysis is a customer segmentation technique that uses data to help marketers make informed decisions. By identifying homogeneous groups of customers, marketers can create personalized marketing strategies to improve engagement and retention.

If you found this article interesting, your support by following steps will help me spread the knowledge to others:

👏 Give the article 50 claps

💻 Follow me

📚 Read more articles on Medium

🔗 Connect on social media Github| Linkedin| Kaggle

❗Get UNLIMITED access to every story on Medium with just $1/week — HERE

☕Buy Me a Coffee — HERE

#RFMAnalysis #CustomerSegmentation #MarketingAnalytics #FraudDetection #DataDrivenMarketing