It's easier to sell to existing customers than to acquire new ones. After all, existing customers are already convinced of the value of your product or service and have already gone through the purchasing process. The exact value a customer represents can be found in transactional data. With a RFM analysis, you can find and utilize that value.
RFM stands for recency, frequency, and monetary value and is a way to segment buyers into different groups based on order history. RFM assigns values to how recently a customer made a purchase, how often they made a purchase, and how much money was involved.
You can calculate RFM in various ways, such as using Excel or a CRM. However, the most efficient way is to use an existing RFM model written in Python or R. For this, you'll need a data scientist because those calculations are quite complex without advanced tools. A data scientist can help you implement such a model technically and ensure it aligns with your organization's transactional data. You provide the data and can also focus on translating the RFM analysis into your work.
In most advanced Customer Data Platforms, such as BlueConic, you can automatically and in real-time link transactional data to customer profiles — which saves a lot of headaches.
In the RFM analysis, each variable scores, for example, between 1 - 5. The assumption is that a customer who scores a 5 for all variables is your best customer, while a customer who scores a 1 for all variables is your least good customer. You can also calculate an average RFM score.
This results in as many as 125 (5x5x5) possible combinations. That's too many to get a good overview. Therefore, I advise you to go for a manageable number of segments. If you use the average RFM score, for example, you can choose gold, silver, and bronze. In the example below, 11 segments have been chosen, and frequency and monetary value have been combined because they both indicate how much the customer buys. Each segment has also been given a name.
To better understand how these segments relate to each other, you can plot them in a two-dimensional figure. In the example below, recency is on the x-axis, and frequency-monetary value is on the y-axis. The percentages indicate how many customers are in that segment relative to the total number of customers.
Your best customers are in the top right, while the customers you want to extract more value from are in the bottom left. So, you want to push customers to the right and up with targeted marketing actions.
There, the first steps have been taken: your customers have been segmented based on their transactional data. In the next blog, we will work with these segments. Which approach will you choose for a 'Champion' and which for an 'About to Sleep' customer?