Evaluation of the Effectiveness of a Model Based Salesman's Planning System by Field Experimentation

Publisher Website
Google Scholar
Abstract
During 1975 twenty United Airlines' salesmen in New York and San Francisco participated in an experiment to evaluate the sales effects of supporting call frequency decisions with an interactive management science model. The CALLPLAN model [Lodish, L. M. 1971. CALLPLAN: An interactive salesman's call planning system. Management Sci. 18 (4, December, Part II) 25–40.] integrates judgemental sales response estimates to various call frequencies made by the salesman and his manager for each account into a mathematical program. Travel time, time per call, and account profitability are also considered. The model output is an account call frequency schedule which maximizes anticipated sales or profit subject to limited time available. Four passenger and one cargo sales representative utilized the model in each city. The users were initially skeptical of the system. However, after they used the interactive model, they realized that they were in control of it and viewed the experience as productive. The ten CALLPLAN participants were chosen randomly from ten pairs of salesmen who were Individually matched by local management using personal characteristics, compatability of territory size, revenue, and account mix. The remaining ten salesmen in the control group were told that they were participating in an experiment and each member manually estimated anticipated sales to compare with the CALLPLAN group. After six months the average CALLPLAN salesperson had 8.1% higher sales than his matched counterpart (t = 2.42, 9 degrees of freedom). The probability that such a large increase could occur by chance is less than 2.5%. Combining both judgemental and objective data with a mathematical program resulted in behavioral changes which significantly improved sales performance.

This publication has 0 references indexed in Scilit: