Predictive RFM Segmentation

Are you in need of quick, performance-oriented customer segmentation? We'll tell you which customers to focus on in your campaigns.

Segmentation with prediction and focus on performance

If you don't need to know the character of your customers in detail, you can use predictive RFM segmentation. It is a predictive model that can separate good customers from average customers and inactive ones based on transactional data. The RFM abbreviation stands for recency, frequency and monetary. The model can also predict the probability and monetary value of future customer purchases over the next 3 months based on past transactions (purchases).

Each model is first optimized based on correlations in your data, including the selection of input variables. Such a model is then applied to your data in order to calculate the probability of future purchase and future monetary value of the customer.
Models of this type are the most commonly used ones and also the most useful ones for direct marketing (either digital or traditional), CRM and customer service. The advantages include processing speed, minimal implementation costs, and automated processing.

Good news at the end. We have the model fully automated, and you can enjoy the results in just 5 minutes. In Keboola or on-premise, you only need to press the button.

Data Science - Data Mind - Predictive RFM Segmentation



RFM Segmentation (Recency, Frequency, Monetary) - Value view of segments in terms of number of purchases (frequency), monetary value and time period since the last purchase (recency). This is a simple segmentation, but it can be very effective.

Customer Segmentation - Finding internally consistent groups among similar customers.

Cluster analysis - a statistical technique designed to group elements based on proximity or similarity. It is used in both technical fields and in marketing to gain an overview of the population.