Conjoint Analysis Techniques (from Sawtooth)

CBC (Choice-Based Conjoint)
· Consumers choose sets of attributes (in the form of a complete product), rather than individually rating desired attributes
· Price can be included as an attribute to make conclusions about the demand for individual attributes
· Software packages can dynamically create “products” by combining attributes so that the marketer doesn’t have to
· Mathematical analysis can identify desirable attributes and market segments with different preferences
ACBC (Adaptive Choice)
· Asks a few simple questions first and then uses those to create a new “decision set” of products; then continues with regular CBC
· Works better with large numbers of attributes
· Works better with small sample sizes (of survey-takers)
· Model better reflects consumer decision-making process (creation of a decision set first, then decide on attribute-combinations)
ACA (Adaptive Conjoint Analysis)
· Similar to ACBC, but the software “learning” process occurs continuously
· Uses regular CBC questions to “learn” consumer’s preference and then hones in on certain value-areas like ACBC does
· Simple for the marketer to set up
CVA (Traditional Full-Profile Conjoint)
· Similar to CBC, but doesn’t combine complete sets of attributes every time (might only choose to combine six (varying) attributes, even though you’ve specified 12)
· Better for smaller sample sizes
· Not as good if you’re interested in interaction effects between attributes
· As CBC, can calculate part-worths of attributes