Creating a MaxDiff

MaxDiff (Best-Worst Scaling) estimates preference or importance scores for a set of items, such as brands, product features, or claims. Rather than asking respondents to rank a large list, MaxDiff breaks it into smaller tasks. In each task, respondents choose the best and worst items from a subset of typically 4 to 6 items. This makes the process more manageable and produces more accurate, reliable results.

MaxDiff outperforms traditional rating scales by offering clearer distinctions between preferences. Since respondents choose only the best and worst options without relying on a response scale, it eliminates common biases seen in cross-cultural research.

Enter options individually or paste them from Excel or Word. To tailor the list shown to each respondent, apply a Dynamic list. This lets you build logic that controls which items appear — for example, showing only the brands a respondent has purchased before. Once a dynamic list is applied, change the MaxDiff type to Relevant items in the Advanced tab. This is necessary because the type determines how the exercise design is configured, and the design settings for Relevant items account for the fact that not all respondents will see the same items.

If a dynamic list containing an exclusive item is applied to a Relevant items MaxDiff, that item is automatically excluded from analysis. Since it never appears in a respondent's design, including it would distort the utility calculation.

Three labels are available in the settings panel:

1. Best and worst labels: Label the scale respondents use to choose items.
2. Items label: Provides context for the items — for example, "Cookies" above a list of cookie flavors.
3. Show task counter & Task label: A counter showing the respondent's progress through the exercise. The counter can be disabled and the label is optional.

The Advanced tab contains additional settings for refining the exercise beyond the basic configuration.