RFM ANALYSIS AND SEGMENTATION

RFM analysis is an evaluation of the customer base according to three criteria: recency (Recency), frequency of purchases (Frequency), and the sum of consumer expenditures (Monetary). The RFM model allows for segmenting the audience according to key business parameters based on each buyer’s transaction history. This method is convenient and effective for B2C but can also be used in B2B.

What we’ll discuss

• What is RFM analysis?
• The algorithm and essence of RFM analysis of the customer base
  • Step #1: Data collection and scoring
  • Step #2: Building the RFM model
  • Step #3: Segmentation and development of a marketing strategy
• Example of how to conduct RFM analysis in Excel
• Objectives and benefits of RFM analysis in marketing
• Limitations and disadvantages of RFM segmentation
• Conclusions

RFM analysis was first used in email marketing, as discussed in the article by Tom Wansbeek and Jan Roelf Bult, “The Optimal Choice for Direct Mail Marketing”, published in 1995 in the journal Marketing Science. This work confirmed the Pareto principle, according to which 80% of sales come from 20% of a brand’s clients.

In the 1990s, the RFM model was widely used to optimize direct marketing campaigns. Thanks to this method, companies understood whom to send commercial offers to and significantly saved on creating printed advertising materials.

What is RFM Analysis?

RFM analysis requires data on transaction dates and amounts spent, linked to customer records. These three behavioral attributes enable businesses to segment their audience, which can then be leveraged to refine marketing strategies.

1. Recency refers to how recently a customer made a purchase at a store or company.
2. Frequency represents the number of purchases a customer has made over their entire relationship or within a specific timeframe.
3. Monetary Value indicates the average amount spent by a customer per transaction.

The optimal threshold for each of these criteria varies depending on the business sector, its seasonality, product range, average product cost, and other characteristics of the goods or services offered. RFM analysis is widely used in retail and professional services. Owners of online stores and agencies that track customer behavior and record transactions can impact their company’s profitability by effectively interacting with their target audience.

Example: A real estate seller does not expect high purchase frequency from the same client since the important factor is the amount spent by the client in a single transaction. A client might not return but can generate significant revenue by purchasing a villa by the sea or a commercial property downtown. Conversely, a client who buys a budget apartment might seem less valuable but could prove more beneficial if they buy similar properties for rental purposes over the years.

In retail, consider an online children’s clothing store. By converting a new buyer into a brand enthusiast, a business can ensure a consistently high frequency and recentness of purchases. Even if the items are inexpensive, the child’s rapid growth means parents will regularly update the wardrobe, especially with seasonal changes.

In some business areas, the indicators of recency and frequency might be uniformly low across all customers. This can be the case for a travel agency, a hotel chain, or a beachside café. Some owners might rely on chance, while others collect contact information and invite customers back for future holidays or vacations. If a large family vacations annually in the same mountain town and frequents a favorite restaurant, they might spend fifty times more there during their holiday than a random tourist who might never revisit because they won’t even remember the place.

On the other hand, high indicators of recency and frequency are typical for sectors like psychotherapy, medicine, beauty services, sports services, marketing, and advertising. Grocery stores and daily wear retailers, with effective customer engagement, successfully boost all three RFM metrics by ensuring customers are regularly supplied with necessary and, more often, numerous non-essential promotional items.

Customers who made a purchase last week and were satisfied are likely to return and make another purchase if reminded timely and made an attractive offer. Those who made infrequent but high-value purchases are likely ready to spend more, making them ideal targets for premium products. Those who have spent the most over a long period are likely to continue doing so, making them candidates for exclusive community memberships or loyalty programs.

“RFM analysis allows marketers to increase revenue by focusing on specific groups of existing customers (i.e., customer segmentation) through messages and offers likely to be relevant based on a set of behavioral responses. This leads to increased response rates, customer retention, satisfaction, and customer lifetime value (CLTV)”. “RFM Segments Based on RFM Analysis: An In-Depth Guide” — Florian Delval, director, technical PMM

The RFM model considers consumers’ buying habits and focuses business owners and marketers’ attention on key customers to invest in advertising for profitable buyers.

Algorithm and Essence of RFM Analysis of the Customer Base

The RFM analysis algorithm and its essence involve segmenting the customer base manually, one of the key advantages of this method since it does not necessarily require purchasing specialized software and learning how to use it.

To segment your customer base, you will need Microsoft Excel or Google Sheets along with transaction data associated with each customer. This information is recorded in a CRM system. Therefore, it’s beneficial to implement a CRM as early as possible in your business to avoid losing valuable data that can later be easily exported into a spreadsheet.

It’s possible to automate RFM analysis using tools such as the Reveal app, the Mailchimp platform, etc. These services handle all the work of assessing and segmenting customers; you only need to set the scale for the R, F, M values according to your business specifics (maximum customer lifecycle duration, cost of goods or services, frequency of purchases made by a consumer).

In most cases, the scoring range is from 1 to 5. The maximum score could be 10 points or even 3, depending on the size of the customer base and the diversity of the business’s product offerings.

Let’s consider the algorithm for calculating RFM. For example, take a business with a recency and frequency scale ranging from 1 to 3 points and an upper spending limit of $300 per year.

Step #1: Data Collection and Scoring

The analysis period depends on the specifics of the activity, but typically data from the past year is used. Essential parameters for analysis include the customer identifier (their name and surname or phone number, email, etc.), the date of the last purchase, the number of purchases or other conversion actions made by the client, and the total amount spent by the client.

Each customer entered into the database is assigned a score for each of the three categories of the RFM model.

• Recency. Subtract from three the number of months since the client’s last purchase (maximum — 3).
• Frequency. Subtract from three the number of purchases the client made in the last year (maximum — 3).
• Monetary Value. Customers who spent $300 or more receive 3 points, those who spent $350 get a score of 3.5, and so on.

Step #2: Building the RFM Model

This stage involves determining the value of customers and grouping them. Each client receives a score from 0 to 3 points for each of the three variables, which can be considered separately or added together to determine the total score of each customer, i.e., their value to the company. For example, a client with scores “3,1,3” is one who made a recent purchase (R = 3), one purchase in the last year (F = 1), and spent $300 or more (M = 3).

After distributing clients into groups, you will have 27 segments, from the worst 111 to the best 333. If the upper limit of the scale is higher, for example, 5 points or 10, there will be significantly more segments. In the final stage of segmentation, i.e., the development of a marketing strategy, some groups can be combined if the difference between their scores is small.

111 — old single, low check	121 — old occasional, low check	131 — old frequent, low check
112 — old single, medium check	122 — old occasional, medium check	132 — old frequent, medium check
113 — old single, high check	123 — old occasional, high check	133 — old frequent, high check
211 — dormant single, small check	221 — dormant occasional, small check	231 — dormant frequent, low check
212 — dormant single, medium check	222 — dormant occasional, medium check	232 — dormant frequent, medium check
213 — dormant single, high check	223 — dormant occasional, high check	233 — dormant frequent, high check
311 — recent single, low check	321 — recent occasional, low check	331 — recent frequent, low check
312 — recent single, medium check	322 — recent occasional, medium check	332 — recent frequent, medium check
313 — recent single, high check	323 — recent occasional, high check	333 — recent frequent, high check

Step #3: Segmentation and Marketing Strategy Development

The main goal of RFM analysis is to deepen the understanding of the target audience and improve communication with it. Ad messages will be more relevant to the queries of each specific target group if the final stage of segmentation is approached responsibly.

Certainly, your target audience will be customers who score the highest points when combining their scores. However, not only such customers are valuable to the company. Analysis based on Recency, Frequency, and Monetary indicators allows the creation of a large number of small groups. For each of them, you can create an individual sales funnel or, at least, different email series.

• Clients with high R (recency) and F (frequency) parameters and a low M (average check). They buy regularly but spend not too much money, choosing inexpensive products. Instead of promotions and discounts, it’s worth offering them participation in a loyalty program or launching cross-selling campaigns and upselling.
• High M, low F. These are clients who spend a lot of money but buy infrequently. In this case, it is necessary to interest them more often, remind them about the company, possibly offer a customer card with cumulative points or a discount on each subsequent purchase.
• Low R. Clients who have not bought from you for a long time can be reactivated if they see new products or sales. A deadline can revive dormant customers, accelerating their decision-making about the purchase.

Once RFM segments are created, they need to be named, for example, “Golden clients”, “Discount seekers”, “Rare spenders”, “Loyal customers”, “At-risk clients”, etc.

The final stage of RFM analysis is the most important and labor-intensive. For each obtained segment, it is necessary to create a unique marketing strategy tailored to specific behavioral models.

How often you need to review segments and change the marketing strategy based on customer behavior depends on three factors:

• The consumer lifecycle
• The product or service lifespan
• The period during which the client can make a repeat purchase

Remember that the customer groups created based on RFM analysis are not static. Consumers can move from one segment to another. Loyal customers can become “dormant”, and at-risk clients can become discount seekers.

It is advisable to repeat the RFM analysis at least once a year. For large stores and companies where a large number of people spend a lot of money daily, updating segmentation results should be done much more frequently, sometimes even monthly.

However, do not rush to transfer clients from one segment to another if your business is seasonal. A client might be “golden” for your company but only during the 3 months of the year when they have the highest purchase activity.

Example of How to Perform RFM Analysis in Excel or Google Sheets

o avoid manually systematizing data about the customer identifier, the date of the last purchase, the number of orders for a certain period, and the total amount spent during that same time, it is advisable to export them to Microsoft Excel. After this, add the following columns:

• current date;
• number of days since the customer’s last purchase;
• Recency;
• Frequency;
• Monetary Value.

In the screenshot, I highlighted in green the columns whose data can be loaded from the CRM, and in yellow those that need to be filled in manually or using formulas.

Let’s fill in the column “Number of days since the last purchase”. To do this, calculate the difference between the current date and the time of the client’s purchase using a formula and drag the cell down to the end of the table.

Example table at the link

For each parameter of the RFM model, you should have your own scoring scale. For example, let’s analyze the recency of purchases:

• Up to 50 days — 3 points;
• From 51 to 150 days — 2 points;
• From 151 days — 1 point.

Evaluating the frequency of purchases:

• 1 purchase — 1 point;
• From 2 to 5 — 2 points;
• 6 or more — 3 points.

Analyzing the monetary equivalent:

• Up to $2000 — 1 point;
• From $2001 to $8000 — 2 points;
• From $8001 — 3 points.

To consolidate the data and summarize, you need to create a new column named “RFM” and use the formula =E2+F2+G2 to fill it.

By sorting the last column in descending order, you will get the most profitable clients at the top. After this, the work approaches a logical conclusion. You can then organize the main action plans for each segment.

The segment group “Regular Customers”

Segment	Characteristic	Action
333	Strategically Important	20% of clients who generate 80% of the profit. Do not offer them discounts; instead, give a gift or assign a special status, card.
331	Regular Customers	Customers who frequently make small purchases. They can be moved to the strategically important segment by offering bonuses.
322-321	New Customers	Recently familiarized with the brand, interested, but have not made expensive or regular purchases. They should be invited on social media, informed about the products, and offered promotions.

Continue the table in this manner for each type of customer, separately preparing tables for “Dormant Customers,” “Customers at Risk of Leaving,” etc.

Goals and Benefits of RFM Analysis in Marketing

Any segmentation of the customer base allows for the personalization of trade offers and optimization of advertising costs by showing people only what they are more or less interested in.

While cohort analysis, geographic and demographic segmentation, and other types of target audience research only provide a superficial understanding of consumer behavior or characteristics, or the effectiveness of certain sales channels, RFM analysis provides comprehensive and accurate data. It’s the numbers and reasoned evaluation based on their summation that help business owners and marketers adapt to market changes and meet customer needs.

“While demographic indicators are static, the RFM model considers the dynamic behavior of consumers, giving you tools to adjust your marketing strategy over time. In your personal life, you likely wouldn’t treat a new acquaintance the same way as an old friend, and the same applies to your clients. You wouldn’t want to sell to absolutely new customers the same way you do to your most loyal customers who frequently make large purchases”. “RFM Analysis Is Your Key to Targeted Marketing Campaigns” — Gavin McLaughlin, director of Analytics at SkyPoint Cloud

RFM segmentation reflects the value of customers and identifies those who belong to the 20% generating 80% of a company’s profit. By considering its metrics, you can enhance conversion at every stage of the customer journey, not just during the first interaction or solely on the website.

Key Tasks of RFM Analysis

• Enhances email marketing campaign effectiveness. By dividing the email database into groups, you can create several series of emails and launch automated trigger mailings.
• Strengthens brand loyalty. The best way to maintain a connection with recent or new customers is to timely offer them relevant content. For one segment, this might be educational materials, while for others, it might be promotions and product selections.
• Reduces customer churn. To avoid losing customers who have not ordered from you in a long time, send personalized messages, offer repeat purchases with a discount, or conduct surveys. The latter not only helps retain customers but also improves service or products.
• Reduces marketing costs. Conversion into applications can be increased for each segment. However, the most important will be those customers who are true fans of the brand, provide native advertising, leave positive reviews, and regularly generate profit. Segmentation allows you to focus your attention precisely on such consumers, investing in their retention and increasing the average check.
• Helps create phone scripts for the sales department that will assist salespeople in finding approaches to different types of customers not typically written about in books or online, as they are unique to your business sector.
• Improves advertising campaigns on Google and social networks, refining ad displays, enhancing text, lead magnets, and CTAs.
• Expands the customer base and increases reach. Using RFM, you can identify your best customers and use them as a starting audience on an advertising platform. By modeling similarity to automatically identify potential customers with similar characteristics, ads are shown to targeted users.
• Even if the goal for your business or a specific project is not a purchase but another action, such as video viewing or applying for a consultation, RFM analysis helps assess results and enhance advertising effectiveness. There are various versions of such segmentation, for example, RFD.

RFD (Recency, Frequency, Duration) — a modified version of RFM analysis that can be used to analyze the behavior of consumers of business products aimed at an audience of viewers/readers/surfers”. “RFM (market research)”, — Wikipedia

RFM analysis offers many advantages that make it a universal and always effective method of target audience segmentation and personalization in business.

• Does not require complex tools or sophisticated analytical capabilities. A marketer without special training or even the business owner can perform RFM customer segmentation using a standard spreadsheet.
• Considered effective in direct marketing. Besides email campaigns, the results of the analysis are now widely used to improve communication with clients in popular messengers and social networks.
• Is a comprehensive tool that can be used throughout the entire customer lifecycle. This includes setting up trigger mailings when contacts move between RFM segments.
• Can be automated and provide an instant response of advertising scripts to user behavior.

If you plan to automate segmentation using the appropriate service, automatic scenarios for each RFM segment will save you time and avoid mistakes due to human factors.

When a client transitions from one segment to another, you receive a notification from the script on your analytical platform. At the same time, the consumer will automatically be sent a relevant offer, such as an individual promo code or a survey regarding service quality.

Limitations and Disadvantages of RFM Segmentation

The capabilities of RFM analysis have their limits. This method has proven effective in the retail and professional services industries, but for car dealerships or real estate agencies, the metrics can be confusing and unexpected. Certain parameters are insignificant in one niche and determine 99% of the outcome in another. No matter how automated your analytics system may be, a marketer, and sometimes a team of specialists, is needed to develop a strategy based on RFM segmentation.

The disadvantages of RFM analysis are few but significant.

• It is not very effective if the customer base is small or if there are few orders.
• It is almost unsuitable for companies that clients approach only once or twice.
• The method is retrospective, making it difficult to predict the future of the business. To develop a strategy, it needs to be combined with data obtained by other methods.
• Analyzing large customer bases requires special software.
• Scoring and grouping need to be updated periodically. In some cases, this takes so much time that a dedicated specialist should be assigned for RFM analysis.
• It requires a deep understanding of the business to interpret the results correctly.

The last point relates to common mistakes made by marketers who are conducting the analysis for the first time or are not sufficiently familiar with the company for which they are developing the strategy. They might write off inactive clients who could actually be re-engaged and converted into loyal customers. Another mistake is overly aggressive advertising targeted at “golden” customers. Even regular customers sometimes need a timeout.

A third common misconception involves neglecting seasonality and personal events in the client’s life. A client might make an expensive purchase on their birthday or during the winter holidays and then not interact with the company for a long time, yet they should not be considered part of the “dormant” segment.

Answering Common Questions

Conclusions

Many business owners believe they know their most profitable customers well. Marketers also often mistakenly value those customers most who are ready to buy the most expensive product or order a premium service. However, monetary value is just one of the three key parameters of a customer’s value to a company. It’s also important to consider the recency and frequency of their purchases.

Based on this data, you can determine how many target audience segments you have and what marketing measures should be used to increase profits. RFM analysis can be performed manually using Microsoft Excel or Google Sheets. For automation of such segmentation, there are special paid programs mostly used by high-turnover companies.

Demographic, social, and personal indicators of the target audience are mostly static. The RFM model considers the dynamic behavior of consumers. It allows for the personalization of advertising and communication with clients.

RFM segmentation also visualizes consumer responses to your marketing. A business owner or marketer can make strategic changes for future campaigns based on such analysis. There are many factors that affect RFM parameters, including the type of products, price, and trade format.

This method of analysis is considered most effective in the B2C sector, especially in retail, with a base of more than 10,000 contacts. When applied in the B2B sector, it is advisable to reduce the number of consumer groups, consider seasonality, and additional factors.