Discrete Choice Modeling (DCM) relates product attributes to choice using an experimental design.  The design reflects multiple competitive products in a market (or even multiple product categories), which may have unique attributes.  The respondent evaluates a series of choice sets comprised of several products and makes a choice within each set.   

The discrete choice analysis has been traditionally done at the aggregate level.  However, due to computing systems' accelerated ability to efficiently compute iterative mathematical models, hierarchical Bayesian analysis can be  applied by SMART to yield DCM attribute utility value estimates at the individual respondent level, which allows us to segment survey respondents on the basis of their product/attribute preferences.

 The discrete choice/trade-off analysis task produces powerful quantitative output consisting of attribute utility values (a.k.a. part-worths).  The utility values become the ingredients of a customized PC-based market share simulator, which enables marketing managers to simulate probable market response to discrete attribute changes.  The interactive tool yields the share of choices (market share) each configuration (including competitive profiles) receives under interchangeable, "what if" scenarios.  Simulations can be run for as many segments as the sampling plan accommodates.