Firm-Customer Relationship Efficiency

Abstract

Firm-Customer Relationship Efficiency (FCRE) measures the performance of the retention function at the customer level from a productivity perspective. The efficiency measurement aids in the allocation of resources to improve the performance of the retention function. Existing methods to evaluate retention efforts exclude resources expressed in non-monetary terms like quality despite the growing importance for the maintenance of the firm-customer relationship. Retention operation is assumed to be a production process during which the firm and the customer exchange resources through a black box. The efficiency of this transformation is measured using benchmarking against a frontier of best or efficient practices. DEA is applied to construct it. A set of models is built to measure performance losses, which are attributed to the negative effect on revenue from overinvestment, waste of resources and unutilized sales potential. We built on the concepts of congestion, technical inefficiency and slacks to measure these respectively. The analysis also contributes to improved performance by investing on efficient customers, reducing waste gradually and setting minimum expected sales target by identifying worst performers. It is also shown how FCRE can be included in CRM tasks like identifying customer segments and churn modelling. The FCRE concept was applied on an online service with a subscription business model. The congestion model was modified to identify congestion in one input. The waste model was solved to identify total waste and sales increase potential. We applied further on the waste model layering technique to get more feasible targets for efficiency improvement and the worst practice frontier concept to identify worst performers. In order for the efficiency scores to be used for managerial purposes, we investigated the relationship between the efficiency scores and customer information. We find that FCRE can be used to identify customer segments and in signalling churn.