Frequently asked questions

Aren't surveys much less accurate than actual sales data, because people won't necessarily do what that say they will?

No. Stated Preference (survey) DCM results have proven to be exactly proportional to Revealed Preference (sales data) results. The first proof of this was published in 1989 by Taka Morikawa, and there have been many subsequent proofs. Stated Preference DCM's predicted optimal configuration will be the optimum configuration, and if it is predicted to be twice as good as the least optimal, this will be so. Revealed Preference can only consider the status quo, and the narrow range of configurations and pricing historically available. This limits the scope of Revealed Preference studies, and renders them useless for predicting uptake of totally new products or pricing.

Is what you're doing Conjoint Analysis?

No. DCM is much more advanced and accurate than Conjoint Analysis. Conjoint is a theory about the behaviour of numbers, DCM is a theory about the behaviour of people. While conjoint 'seems to fit' what is happening in real markets, DCM is based on a theory that explains the underlying phenomenon (Random Utility Theory). Results support the theory. Over time, further results both extend the theory and lead to more efficient practice.