Abstract
Importance–Performance Analysis (IPA) is a simple and useful technique for identifying those attributes of a product or service that are most in need of improvement or that are candidates for possible cost-saving conditions without significant detriment to overall quality. To this end, a two-dimensional IPA grid displayed the results of the evaluation about importance and performance of each relevant attribute. This paper shows that ordinal preferences are better than metric measures of the importance dimension and proposes a formula to transform the ordinal measure into a new metric scale adapted to the IPA grid. This formula makes allowances for the total number of features considered, the number of rankings, and the reported orders of preference.