Customer Segmentation using RFM Analysis

Customer Segmentation using RFM Analysis

How well do you know your customers? Whether you operate in a B2B or B2C space, chances are that 80% of your business comes from just 20% of your customers (Pareto's Principle). According to a study by Forbes, acquiring new customers costs five times more than retaining the existing ones. Identifying high-value customers is crucial to increase revenue and building brand loyalty.

Customer Segmentation is a critical strategy for businesses to understand and engage with customers effectively. Understanding the behavioral patterns of customers can help personalize the purchases they make and cater to their needs better. 

Questions like who are the customers who contribute more to sales, who are the customers about to churn, will help digital marketers understand the behavioral patterns of their customers. 

While there are many criteria based on which the customer base can be segmented, this solution focuses on segmenting customers based on the RFM analysis.

What is RFM Analysis?

RFM analysis (Recency, Frequency and Monetary) is a method used to identify and segment existing customers based on their purchasing behavior.  The key metrics of RFM analysis include

Recency

Recency refers to how recently a customer has made their purchase and this is the most important metric of the other metrics. This metric is a strong indicator of customer loyalty and interest.

Frequency

Frequency refers to how often a customer makes purchases or interacts with a business within a specific period. It measures the level of engagement and loyalty of a customer.

Monetary

Monetary value refers to the total amount of money a customer has spent with a business during a specific period. 




Industry-Specific Applications of RFM Analysis

  • SaaS and subscription services: RFM analysis can be adapted for SaaS and subscription services to segment users based on engagement, renewals, and revenue contribution.
  • Financial Sector: RFM analysis can improve credit scoring and risk assessment by evaluating customer transaction patterns, helping financial institutions make more informed loan approval decisions.

Data Requirements

For RFM analysis, you'll need a transactional dataset with the following equivalent columns (details) 
  • A product (Product ID)
  • A related transaction (Transaction ID)
  • Number of products purchased in a transaction (Product Quantity)
  • The product purchase price (Product Price)
  • Transaction date (Date)
  • Customer who made the purchase (Customer ID, Customer Name)
We have used a sample table of e-commerce data for illustration.

Steps for Implementing RFM analysis

1. Gather and Prepare Transaction Data : 

Gather all transaction data, ensuring it includes customer identifiers, transaction dates, and monetary amounts, and address missing or inconsistent values, ensuring data integrity before analysis.

2. Compute RFM Metrics: 

To segment customers based on their behavior, we compute three key metrics: Recency (R), Frequency (F), and Monetary Value (M). Below are SQL queries for each, along with detailed explanations.

Recency (R)

Recency measures how recently a customer made a purchase. It is calculated as the number of days since their last transaction. Customers with recent purchases are more engaged, while those who haven’t bought in a long time may be at risk of churn.

The time frame for RFM analysis should be tailored to your business model and industry. Choosing the right period for RFM analysis is essential, as it directly influences the accuracy of customer segmentation and the quality of insights derived.

Frequency (F)

Frequency tracks how often a customer makes purchases within a specific period. A higher frequency indicates a loyal customer who regularly shops, while a lower frequency suggests occasional or one-time buyers.

Monetary Value (M)

The total amount spent by the customer in the same period

RFM Query Table

SELECT
"Customer ID",
"Customer Name",
DAYS_BETWEEN(MAX("Transaction Date"), CURRENT_DATE()) AS "Recency",
COUNT ("Order ID") AS "Frequency",
SUM("Transaction Amount") AS "Monetary Value"
FROM  "Customer Data" 
GROUP BY "Customer ID",
  "Customer Name" 
ORDER BY "Recency" ASC,
  "Frequency" DESC,
"Monetary Value" DESC 



3. Segment Customers using Cluster Analysis

Manual scoring can skew the results and may not be practical for handling large volumes of data. In contrast, using machine learning algorithms like cluster analysis ensures unbiased, efficient, and data-driven segmentation. Unlike traditional scoring methods such as the quantile or percentile-based approach, cluster analysis recognizes inherent relationships and patterns in the data. With cluster analysis, business can obtain accurate segmentation and devise targeted strategies to improve sales and customer retention.

Follow the below steps to apply cluster analysis,
  1. Click the Create New icon and choose New chart from the drop-down menu.
  2. Add the columns to the chart shelf as shown below,
    1. X-axis: Customer Name
    2. Y-axis: Monetary Value with Sum function.
  3. Click Generate Graph and change the chart type to bar chart.

  4. Click the Analysis icon and select Cluster Analysis > Add Clusters.
  5. The Model is chosen automatically based on the columns dropped in the shelves.
  6. By default, the columns dropped in the shelves (Monetary Value) are selected as factors. Click the drop-down icon to include Recency and Frequency columns as factors.

  7. The number of clusters is determined automatically but can be adjusted based on business needs and specific customer segmentation goals to ensure optimal categorization.
  8. Choose the Normalization method to prevent values of high ranges from dominating the results. For instance, Recency (measured in days) and Monetary Value (measured in currency) have different scales, and normalization ensures a balanced contribution from each metric.
  9. Click Apply.

4. Export Clusters Data

Once the customer profiles have been segmented using cluster analysis, Export the Current view in the preferred table format to build more data visualizations to understand the clusters.


Build an RFM Analysis Dashboard

The RFM analysis dashboard provides a comprehensive view of customer behavior. Let's look at the steps involved in building this dashboard.

1. Import the Clusters Data

Import the downloaded clusters table back into Zoho Analytics using the files option as given below.
  1. Click the New icon on the side navigation panel and choose New Table/ Import Data.
  2. Select files and choose the Clusters Table to import and click Next.
  3. A data preview will be displayed; verify the data types of columns and click Create.

2. Create Reports to Understand the Characteristics of the Clusters

While the data is clustered, understanding the characteristics of each cluster is what enables businesses to take strategic actions. This includes identifying which customers need targeted marketing, personalized engagement, or retention efforts. Recognizing patterns within clusters provides insights into customer behavior, which is essential for optimizing marketing campaigns, improving retention strategies, and enhancing customer experience.
The below reports help understand the distribution of customers across different monetary value, recency and frequency segments within each cluster.

Clusters vs Monetary Value 

  1. Access the cluster table (imported data) and click the new icon > chart view.
  2. Drag and drop the columns as given below:
    1. X-axis - Clusters
    2. Y-axis - Monetary Value with the Count function.
    3. Color - Monetary Value with the Actual Range function.

Analyzing the chart, we can infer that,
  • Cluster 1 consists of a diverse group of customers spanning all spending levels.
  • Cluster 2 includes moderate to high spenders who contribute significantly to revenue.
  • Cluster 3 comprises low to mid-range spenders, often occasional buyers.
  • Cluster 4 represents high-value customers with premium spending habits.
  • Cluster 5 consists primarily of low spenders with minimal purchasing activity.
You can similarly create reports to know about the distribution of customers for the Recency and Frequency metrics.

The below table lists the characteristics of clusters

Cluster
Cluster Classification
Recency
Frequency
Monetary
Recommended actions
Cluster 1
Needs Attention
100 to 150 days
Low to Moderate 
 Diverse spending
Re-engagement campaigns, discounts, or reminders to encourage repeat purchases.
Cluster 2
Loyalist
0-50 (Highly Active)
High
Consistent moderate-to-high spenders
Loyalty programs, exclusive deals, early access to new products to maintain engagement.
Cluster 3
Potential Loyalist
0-50 (Active)
 Low to Moderate  Budget-conscious, occasional buyers Cross-selling, personalized recommendations, and value-based promotions.
Cluster 4
Champions
100-150 (Inactive)
Moderate to High (Frequent buyers)
High spenders
VIP experiences, personalized services, and premium offers to retain and enhance their spending.
Cluster 5
Hibernating
Mostly inactive or infrequent
Low
Minimal spending
Win-back campaigns, incentives, special discounts, and targeted ads to regain interest.


Based on the above table, you can give specific labels to the clusters using the bucket columns option.

RFM Dashboard


Limitations & Considerations of RFM Analysis

While RFM analysis is a powerful customer segmentation tool, businesses should be aware of certain limitations and factors that can influence results:

  • Data Freshness and Relevance: RFM analysis relies on transactional data, making the freshness and relevance of this data crucial for accurate customer segmentation. Setting up automated data imports ensures real-time updates, reducing the risk of working with stale data.
  • Seasonal Variations: Customer purchasing behavior often fluctuates due to seasonal trends, holidays, and industry-specific cycles, which can impact RFM scores and lead to misleading segmentation if not accounted for properly. Instead of analyzing only recent months, compare customer behavior for the same period in previous years to detect true engagement patterns.


      • Sticky Posts

      • What's New in Zoho Analytics - November 2025

        We're thrilled to announce a significant update focused on expanding your data connectivity, enhancing visualization capabilities, and delivering a more powerful, intuitive, and performant analytics experience. Here’s a look at what’s new. Explore What's

      • What's New in Zoho Analytics - October 2025

        Hello Users! We're are back with a fresh set of updates and enhancements to make data analysis faster and more insightful. Take a quick look at what’s new and see how these updates can power up your reports and dashboards. Explore What's New! Extreme

      • What’s New in Zoho Analytics – September 2025

        Hello Users!! In this month’s update, we’re raising the bar across multiple touchpoints, from how you bring in data, plan and track projects to how you design and brand your dashboards. We’ve added the all-new Gantt chart for project visualization, expanded

      • Announcing Agentic AI - Ask Zia!

        We are delighted to roll out the new agentic AI capabilities in Ask Zia, where every stage of the BI workflow is assisted by AI. With a human-in-the-loop approach, Ask Zia ensures that you’re in command of the decision, while AI handles the complexity.

      • Invitation-Based User Access in Zoho Analytics

        Hello everyone, We’re rolling out an important update on how users are added to your Zoho Analytics Organization and Workspaces. Previously, when admins added users, they were automatically added to the organization. Moving forward, to improve security