Have an idea?

Visit Sawtooth Software Feedback to share your ideas on how we can improve our products.

Identifying "target" groups through individual level conjoint utilities

Suppose after fielding and analyzing a conjoint study, I want to target a specific group of people.  My question: is it appropriate to use individual utilities and/or importances to do so?  This is similar to segmentation, but instead I have a specific goal or group of people I want to find.

If I have a brand attribute with 2 brands, A and B, I can easily flag those that "prefer brand A" by identifying all respondents with a positive utility for brand A.  And if I want to identify those "strongly preferring brand A", maybe I can also look to see if their importance for brand is relatively high compared to some value, such as the average.

Would the above be appropriate for targeting groups of respondents, and then continuing to analyze this group via conjoint or non-conjoint measures?  Is there a better way?

asked Jan 30, 2019 by anonymous

1 Answer

0 votes
Yes, I think you can do that...as long as you are working from "normalized" utilities where each person is placed on a similar scale.  That's available in our "zero-centered diffs" utility output tab in HB.
answered Jan 31, 2019 by Bryan Orme Platinum Sawtooth Software, Inc. (176,815 points)