Aim:

To propose a multi-criterion methodology for customer classification, considering the utilization of KNN algorithms as the base of aggregate planning, from an adjusted RFM model (recency [or currentness], frequency, and monetary [momentary value]).

Methods and techniques:

A methodology was used to classify customers according to six variables: faithfulness to the company, purchasing frequency, customer’s assets, variety of purchased items, physical nearness, and temporary purchasing horizon.

Main results:

A hierarchical scale of variables to classify customers was designed. A general customer classification was added to their classification according to the six variables of the study, which allowed for a customized analysis of their performance.

Conclusions:

The utilization of this methodological tool in ACINOX Holguin sales company validated its effectiveness to solve customer classification issues. Upon the application of this methodology, the executives of the institution had access to several individual clusters of each variable, which supported aggregate planning, enabled decision-making, and optimized the sales process under a general vision.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )