A traditional MaxDiff exercise typically shows 3-5 options on a screen and asks the survey respondent to select the one option they like most and the one they like the least (other dependent measures like purchase interest can also be used).
In a multiple-response MaxDiff, we focus only on the most preferred options (versus asking both most & least preferred) across multiple questions within the same exercise.
This approach is similar in concept to a discrete choice exercise (DCM) that asks the survey respondent to select only their most preferred option without asking which one they least prefer. We have found there is minimal tradeoff in modeled utility scores, and we arrive at the same conclusions.
From the MaxDiff exercise, Utility Scores are derived on each measure so that they can be prioritized or ranked in terms of relative impact on motivation or eye-catching. The Utility Scores are then indexed, where a score of 100 represents an average score.
By including more than one MaxDiff measure within the survey, the output also lends itself to quadrant mapping to identify the options that may score highest on motivation and highest on being eye-catching.
To further help with prioritization efforts, MaxDiff can be combined with TURF analysis to model the combination of options that maximizes motivation across the broadest population.
TURF modeling is flexible in that we have the ability to force certain options into the final solution when there are known ‘must haves’ that need to appear in the final solution, even if they are is not mathematically part of a solution that maximizes reach.
In the following example, the top combination includes 6 options that maximizes reach to 83% of consumers before the incremental gains have minimal impact.
