| Hi-index | 0.00 |
This paper describes and discusses a Plackett-Burman experiment aimed at increasing supermarket sales of a top-selling national magazine. The experiment involved 10 factors relating to in-store advertising and the location of the magazine within the store. We discuss issues including choice of factors, alternative designs, sample-size determination (number of test stores and the length of the test period), and the analysis of the resulting data. We show the large benefits that accrued from our approach of testing many factors simultaneously compared to the common industry practice of sequentially testing one factor at a time. We also discuss the potential opportunities of applying this approach to other service problems.