Conjoint analysis (CA) and discrete choice models (DCM) are preference structured models that are
widely used in market research and analytics. They often cover common ground, yet important distinctions that exist
between the two make them better suited for different types of research programmes.
The respondents in a conjoint analysis, evaluate product profiles independently of each other
whereas in DCM, they simultaneously consider a set of profiles and select the one they are most likely to
purchase (if any). The latter approach is more akin to the decision-making process that consumers
use in real life, and this is important in pricing research.
DCM is also better suited for modelling the interaction effects between different
product characteristics such as brand and price, which is essential in pricing research that
measures the distinct price elasticities of demand of brands.
On the other hand, since respondents are rating all product profiles, conjoint analysis extracts
more information about the relative importance of the profiles, attributes and levels.
Like conjoint analysis, advanced hierarchical Bayes versions of DCM produce utilities at the individual
level, and permit what-if simulations, where respondents are assumed to maximize utility.
Due to the aforementioned differences, conjoint analysis is better suited for product development studies
where trade-offs are to be made. The analysis reveals consumer’s preferences of product features, and in the context of
pricing, it can tell us what price may be charged for certain features.
On the other hand, DCM is the current gold standard in pricing research (ad hoc). It more accurately reveals the relationship
between share and price.