Sawtooth Software 2020 European Conference | Virtual Attendance  | September 22–25
Early-bird pricing ends August 22!

More Evidence CBC Is Most Popular Conjoint Method

In an E-News last year, we reported that we suspected the use of CBC had eclipsed that of ACA (ACA had been the most popular technique during the 1990s, according to two industry surveys). We came to that opinion by looking at our own sales data for conjoint analysis software systems. Over the period 2000-2002, the percent of new units sold by conjoint method was:

45%-- Choice-Based Conjoint (CBC)
37%-- Adaptive Conjoint Analysis (ACA)
18%-- Traditional Full-Profile Conjoint (CVA)

We now have more information suggesting CBC is more widely used than ACA. Based on a recent survey of Sawtooth Software users (April 2003), relative use of our three conjoint analysis packages over the last 12 months (weighted by number of projects done) was:

50%-- Choice-Based Conjoint (CBC)
36%-- Adaptive Conjoint Analysis (ACA)
14%-- Traditional Full-Profile Conjoint (CVA)

If all conjoint analysts were studied (Sawtooth Software and non-Sawtooth Software alike), the proportion of those using a discrete choice method would be expected to be even higher relative to adaptive conjoint analysis.

With partial-profile CBC further expanding the capacity of CBC to study more attributes, this is probably also responsible for some of the shift from ACA to CBC. Users have often asked our opinion regarding partial profile CBC and ACA. Partial-profile CBC requires larger sample sizes to stabilize results (relative to ACA), and individual-level estimation under HB doesn't always produce accurate individual-level part worths. The individual-level parameters have less stability than with ACA, but if the main goal is achieving accurate market simulations (and large enough samples are used), some researchers are willing to give up the individual-level stability. Partial-profile CBC results tend to reflect greater discrimination between most and least important attributes relative to ACA, though it is not a given that this means improved accuracy in predicting real world choices.