A preference map's predicted ideal is conditional on the studied samples, measurements and consumers and requires prototype confirmation.
A bounded preference map can efficiently guide prototype selection when uncertainty and validation are explicit.
Projection, scaling, selected products and respondent coverage influence the map. Whitespace is not demand; in-sample fit does not demonstrate future acceptance.
Prediction Before Confirmation
A preference map connects selected descriptive profiles and liking responses. Its apparent ideal is a model output, not an existing successful product. Build a feasible prototype and test it with an appropriate consumer sample before asserting improvement.
Academy Application
Use a clearly labeled fictional map to teach sample coverage, omitted dimensions and the distinction between preference and descriptive similarity. Avoid implying a universal whiskey hierarchy.
Provenance
Meilgaard et al.,2016,printed pp.434–452 and586–588.
