Currently, they are usually found with companies that have one or more forms of customer segmentation.Among these, we find those who exploit mainly the demographic dimension, using variables such as age, sex, income or occupation. Others incorporate transactional elements such as time consumption, frequency of use and fidelity. Less frequently are those that include psychographic variables such as risk aversion, leadership, or the tendency to introversion early adoption of technologies.
However, there is always the question of whether one way or another directed to segment a portfolio or isthis form that customers can better classify them into categories.
In this context it is important to note that any form of segmentation must meet at least two criteria: first, to ensure the generation of groups in which members share a number of attributes and second, that while these are as different possible characteristic attributes of the other segments. A simple segmentation by age range, for example, meets both requirements, generating homogeneous groups in terms of age, which in turnare sufficiently different from other groups. However, it is evident here, that within each group are still presentdifferent profiles, for example, different sexes, so that we can not speak of a high homogeneity of the resulting groups, beyond what can explain the variable age range.
How do it?
The great virtue of data mining is to allow objects to several paradigms in an organization, however, is always vital to reconcile these results with expert knowledge of the business, so give them a sense of globalbusiness and to improve the results the actions arising from this.
Reduction of uncertainty based on information