Predicting Customer Wallet Without Survey Data

Abstract

A single company provides only a part of the total volume of products or services required by a customer. From the company perspective, this total business volume conducted by a customer, the customer's size-of-wallet, is generally unobservable. The percentage of this business done with the company, the customer's share-of-wallet , is unobservable as well. This article focuses on the prediction of these values and on the derived concept of potential-of-wallet, which is the difference between the size-of-wallet and the actual business volume the customer does with the focal company. In the existing literature, the models predicting the customer wallet need survey data to estimate the model parameters. The authors propose an approach to predicting customer wallet without using survey data. In the empirical application, the authors show that a company can generate substantial gains by targeting customers with a large potential-of-wallet.