Since I started to work as data-marketing consultant, one of the goal was to handle the heavy digital processing work, so that internal marketing teams have time to breathe, and focus on creating relevant strategies for each customer groups.

Photo by Nathan Dumlao / Unsplash
The question I asked myself was: "how to design a segmentation that saves time, is easy to use, and has instant access to the people hidden in the data?"

Not long ago, I realized the benefits of RFM analysis-one of the most powerful indicators of customer retention. This is why I integrated it on every consultancy program, driving great results for my cliens.

I got extraordinary results since I'm using it, x3, x4 and x5 increasing in sales and revenues used as base analysis to drive the marketing strategies.

What is RFM (Recency, Frequency, Monetary)?

RFM segmentation is a method to identify the most important types of customers grouping by scores based on their recency, frequency and monetary values. The purpose is to predict which customers are more likely to buy again in the near future.

Knowing this "customer features" allow companies to target specific customer groups based on their behaviors.

Thereby using wisely these insights you can generate an higher response rates, increased loyalty, and better customer lifetime value.  

How can RFM help you?

  • Delivering targeted communication that resonates better
  • Best channel that works for the segment
  • Enables you to deepen your relationship with your customers
  • You can try and test your pricing options
  • Focussing on most profittable customer segments
  • Upsell and cross-sell products for each specific segment

Over the past years, I have been building Customer Retention strategies for various types of e-commerce. In a nutshell, by knowing how recently a customer bought from you, how many orders he placed, and the total value of those orders.

I shoot many of my close up macro images on a small rustic wooden table I bought from a yardsale for $15. I place the table in my kitchen - beside the sliding glass door – for full window light – all natural night. The downside is that I have to shoot during the day - and my kids come by and eat the food I’m photographing.  :)
Photo by Jessica Lewis / Unsplash

Interesting data that you can include for each persona is information regarding:

  1. Persona X is interested mainly in product X but also buys product Y
  2. Persona X, is there a difference in behaviour between new and returning Persona X.
  3. Persona X visits mainly these pages but also visits these other pages.
  4. Use the Behaviour Flow report in Google Analytics to see where Persona X lands and
    where they go after to answer these questions:
    – Where are they leaving the site?
    – Are they finding what they are looking for?

Testing, Testing and More Testing

Now you have data that can help you create different personas, but the changes on the test site are always important. It may not be a good idea to completely change the structure and navigation of your website, because returning visitors may be confused and unable to find the products or information they have been looking for on your website.

Progressive changes may be better, and always use A/B or multivariate testing to test the hypotheses to ensure that these changes will be positive, and users will actually be better involved in new changes or making new conversions.

You may find that the design on the landing page for Persona X can help a large number of potential business customers participate and convert into potential customers. However, the same design does not apply to PersonaY.

Performing different tests for each role can help you optimize the path landing page and adjust it to a page more suitable for each role.

Building a website for multiple personas is challenging, which is why testing and evaluation is essential to ensure that the target audience participates and becomes a customer.


If you want to create a growth plan that generates a new continuous revenue stream, contact me for a free evaluation.