A proposed algorithm for the derivation of consumer profile from minimal transactional data

Abstract

The 80/20 rule says that 20% of the customers produce 80% of the sales; this rule indicates the existence of hidden sales potentials that must be revealed. Those hidden sales potentials can only be discovered by building acustomer profile. A smart customer profile finds the high potential targets, creates CRM strategies, and starts programs to sell the hidden targets. Shifting just a small percentage of the customers whom are not generating profit to the top group of customers generating profit adds significantly sales growth increase profits. Customer profile puts the full picture together to build sales and profits and maximizes the marketing ROI (Return on Investment). Stating the importance of having a clear customer profile, the question of how to build a clear, relevant and comprehensive customer profile is raised. In this thesis, and algorithm is suggested, which if adapted ends up provinding the user with a demographic profile of his customers and clear distribution of the customersaccording to profitability. This distribution is based on the measurement of customer's LTV (Lifetime value) and Loyalty. In addition to the distribution which will result from the application of this algorithm, the resulting multidimensional valuable metric constitute consistent data to apply data mining techniques and get important result.