Segment Those Who Care Most: RFM Analysis for Retail

Divide and conquer is not only Caesar’s motto but also a great piece of advice for modern retailers. Once you’ve collected information on your customer base in your CRM system or loyalty program, you need to analyze it. You should reveal everything your data may know about your customers and use those insights to grow your business.

The Theory

RFM analysis is a model for segmenting customers based on three parameters that define their purchase habits: Recency (the date of the latest purchase), Frequency (how often the customer makes purchases), and Monetary value (the total value of all a customer’s purchases). With RFM segmentation, analysts can put your customers into some kind of Cartesian system where the X, Y, and Z axes mirror the RFM parameters.

Typically, RFM analysis starts with a data set where the main identifier is a customer’s name, surname or clientID (for instance, their ID in a loyalty program). The most important data you have to include in this data set is the date of the last purchase, the total number of purchases, and the total value of all purchases for each customer. This is needed to segment your customer base into groups based on the following:

• Recency: most recent customers
• Frequency: rarely (single orders), infrequently, often
• Monetary value: low, average, or high total value of a customer’s purchases.

At this point, you might be thinking, but “low” and “recent” are too abstract. And you’d be right. That’s why analysts assign numbers from one to five (or one to three) to describe these parameters according to their intensity or size. For example, a customer might be coded as “123” or “222” if you’re using levels one to three.

Example of an RFM analysis in Excel | Credit: A Complete Guide on RFM Analysis

Customers with the same codes will be grouped, so you’ll get a few customer segments. You can then break down these segments to find common behavioral traits inside each segment or across several segments with one or two identical parameters.

You can separate those who bring the most revenue and concentrate your efforts on them, or you can cultivate a lookalike audience. You can find those customers who may not realize their buying potential but who, consciously or unconsciously, want to be advertised to so they purchase even more. That might be a true treasure for your brand.

The way you choose these groups is up to you, of course. RFM analysis only provides a set of rules for organizing your customer base. You and your analysts are fully responsible for finding insights and applying those insights.

When is the Best Time to Try RFM Analysis?

• You’re selling fast-moving consumer goods in a market filled with competitors.
• You’ve launched your loyalty program and your customer base is growing fast.
• You know your audience is diverse, and you’re looking for an easy way to segment it.
• You’ve never personalized your offers in your emails/advertising/promo materials, but you want to.
• The effectiveness of your advertising campaigns is low because you don’t have proper targeting.
• You’re ready to build a remarketing strategy, but you don’t know where to start.

Best Tools for RFM Experiments

RFM analysis is undoubtedly helpful for any retailer. The reliability of RFM statistics and the simple mechanics of analysis make it affordable even for businesses without a separate analytics department. Let’s see what tools can help you with RFM analysis, starting from the simpler and moving to the more sophisticated:

• Excel or Google Spreadsheets are a great idea for beginners to experiment with.
• Specialized marketing analysis tools and built-in plugins for CRM systems will encourage you to get deeper insights from your customer base.
• Python, R, and hierarchical, k-means, and hybrid clustering are programming languages, libraries, and approaches that allow data scientists to perform complex RFM analysis with large databases and using complicated data manipulations.

Each level of analysis is affordable for a growing business. Consider which tools will give you the most profit with the least implementation troubles.

The Motivational Kick in the Pants for Doubtful Retailers

Once RFM analysis helps you answer a couple of serious questions like "Which customers are at the verge of churning?" and "Which customers must I retain?" based on data rather than gut feelings, you’ll be hooked.

Because it works.

Once you send your customers the personalized offers they want — and once you polish your remarketing campaigns — you’ll thank RFM analysis for being the first source to give you such deep knowledge of your customers. So leave your doubts behind, explore your business, and cultivate analytics.

Mariia Bocheva is a business development executive at OWOX BI, a leading technology partner for Google in EMEA.