How to set up the customer-based RFM analysis
Copy the link to the article

Let's take a closer look at how to properly configure the indicators.


In the RFM analysis module, the sales volume of the store is taken as 1. Customers who made orders in the store are sorted by the sum of orders in descending order: the most active customers are at the top, the least active are at the bottom. In the future, the system summarizes them by turns.

To make it clearer how the "Monetary" calculation process takes place, let's take the total sales of the store N as 100 euros. And the customer base will consist of 4 customers who bought for 60, 20, 15 and 5 euros.

We calculate the indicators:

1.60 / 100 = 0.6

  1. (60 + 20) / 100 = 0.8

  2. (60 + 20 + 15) / 100 = 0.95

  3. (60 + 20 + 15 + 5) / 100 = 1

It turns out that for store N we denote the borders of promixately 0.7 and 0.98 so that the first customer gets into the first segment, 2 and 3 - into the second segment, and the third - into the third. There are more customers in a real store, so that's why the borders are set closer to the beginning. The default is 0.2 and 0.35.


Here you set the border values ​​for the frequency of orders. To find out these numbers, go to the "Customers" section and sort by the number of orders.

For example, if the largest number of orders for customer A is 10, and the lowest for the customer B is 1, and customers C, D and E have 8, 6, and 3 orders, respectively.

In this case, the number of orders at the border between "Often" and "Average" will be 7, and between "Average" and "Rarely" - 4. In this situation, customers will be distributed as follows:

  • the number of orders is more than 7 - frequent;
  • the number of orders is less than 7, but more than or equal to 4 - average frequency;
  • less than 4 orders - not frequent.


The time since the first purchase in each business is individual. Therefore, set yourself the time interval since the last purchase, which will be considered appropriate for each of the metrics (average / recent).

After saving the settings, the background recalculation of indicators starts. Upon completion, the user is notified with a notification.


Segments do not include deleted and canceled orders / customers.

Thank you for your feedback.
Was this article helpful?
  • Рекомендации не помогли
  • Нет ответа на мой вопрос
  • Текст трудно понять
  • Не нравится описанный функционал
Previous article
"RFM analysis on customers" widget
RFM analysis is a prediction of customer behavior based on past activities.
Next article
Universal analytical widget on customers
The universal widget allows you to build your own reports based on specified parameters.