MaxDiff: Maximum Difference or Best-Worst Scaling is a technique for measuring the preference and importance that respondents (consumers) place on a list of items. MaxDiff plays a critical role in understanding trade-offs that people would make and ultimately provides a rank-ordering of the list.
Attributes: List of items (e.g. flavors, claims, features, images) from which a member can choose their least and most preferred. Attributes can be either text or images.
Sets: Attributes are distributed into sets, which appear as questions to respondents. The number of sets is dependent on the number of attributes, type of attributes (images/text) being tested, and user preferences.
Utility Score: A derived score used to interpret MaxDiff results. Utility score is calculated by the below formula:
Utility score is used to show the relative preference of attributes being tested. The higher the utility score, the higher likelihood that the attribute will be chosen as the ‘best,’ among the set, indicating higher consumer preference. A zero (or near-zero) utility score indicates that the attribute was chosen ‘best’ and ‘worst’ an equal proportion of times, indicating the attribute is polarizing in some way.
Attribute Performance Chart: In-platform visualization that shows the count/percentage of best/worst selections per attribute
Utility Score Chart: In-platform visualization that shows the utility score for each attribute
“Most”/”Least” Values: The number of times an attribute was selected “Most” or “Least” in a MaxDiff action.
Total Exposures: The number of times an attribute was exposed in a MaxDiff action to respondents. As each attribute is exposed once per respondent, this number will always equal the number of respondents who have completed your MaxDiff survey.
Set Order: In the Horizontal Raw data file, sets and set composition are laid out in the order they were seen by each respondent.
