Data is a very important asset for companies. They only offer value if they are analyzed and resolve business problems. Let’s talk about customer segmentation, which is a very useful technique for improving your relationship with your customers.


Customer segmentation is the division of customers into groups with common characteristics, which allows for better organization and management of the relationship between companies and customers.

This technique is important for several departments, namely marketing, finance and sales:

  • For Marketing, knowing the various groups and their characteristics allows targeted and personalized communication according to the needs and preferences of each group (message, communication channel, number of interactions).;
  • For Finance, being able to assess the value and risk of each segment and develop financing options and attractive prices for each group is an asset;
  • For Sales, through the creation of a portfolio of products/services aimed at the preferences of each customer, customer acquisition techniques and proposals presented can be defined, making commercial contact targeted and more effective.


With efforts channeled into satisfying each group according to their preferences, the communication from all departments becomes personalized for each segment, which increases customer satisfaction, and in turn, the company’s sales and profit!


To perform customer segmentation, we use statistical tools and techniques.

The tool we are going to discuss below is cluster analysis.



Cluster analysis

Cluster analysis is a technique for categorizing and grouping data according to their similarity.

When carrying out a cluster analysis we must identify homogeneous groups of customers characterized by certain attributes. It is expected that there is similarity between individuals within a cluster and differences between individuals from different clusters.

With this technique, groups of different types of similarity can be defined, for example groups of people with the same age, professional category, interest in the same products, preferred contact time, preferred method of contact, preferred method of payment, among others.


In addition to customer segmentation, cluster analysis has other objectives, namely:

  • Error detection;
  • Fraud detection;
  • Pattern detection;
  • Others…


Some algorithms used for customer segmentation are:

  • K-means;
  • Kohonen networks (SOM);
  • TwoStep.


PSE uses these techniques for customer segmentation in day-to-day activities. If you want to communicate in more effective way with your customers and potential customers, PSE is the ideal partner!