Session I of this seminar describes several experiences with product tests and mappings as development tools. One paper deals in particular with preference maps from paired comparisons. The present paper adds some of our experiences in drawing mappings on the basis of results from paired comparison tests. It shows that the order of presentation does have an influence on the perception and the rating of product characteristics. It also shows that the rank order effect cannot be neutralized properly by switching the order of product presentation in a random half of each sample. As a consequence, mappings based on paired comparison tests are - in our experience - difficult to interpret. Thus, if at all possible, we recommend running monadic tests for such purposes. The experiences we describe derive from more solid products, i.e. roasted coffees, but the same can certainly be found in tests with the more ephemeral characteristics of (fine) fragrances.