MaxDiff analysis: how to rank feature priorities

MaxDiff asks people to choose the most and least important items from small sets. Repeated choices produce a relative ranking when ordinary ratings leave every feature near the top.

What is MaxDiff analysis?

MaxDiff, commonly called best-worst scaling, compares items through repeated choice tasks. A balanced study varies which items appear together so each gets fair exposure. Teams use it to rank features, benefits, or messages. Keep each study focused on one of those questions.

Why use MaxDiff instead of ratings?

A rating scale lets someone call every feature important. MaxDiff requires a trade-off in each set, making it easier to distinguish priorities. It cannot tell you whether respondents like the entire list; the highest-ranked item may simply be the least unappealing.

How do you design a MaxDiff survey?

1. Write specific items at the same level of detail. Avoid comparing “better reporting” with “weekly PDF exports.”
2. Remove duplicates and overlapping ideas.
3. Balance how often items appear and which alternatives they face.
4. Keep sets short enough to scan, and pilot the tasks with the intended audience.

Task count and sample size depend on the number of items, analysis model, precision, and groups you want to compare. There is no universal minimum that fits every study.

What do MaxDiff scores mean?

Basic scoring compares best and worst selections, adjusted for appearances. Statistical models can estimate relative utility. Either output describes preference within that study. Look at score gaps and uncertainty before treating neighboring ranks as different priorities.

A MaxDiff example

Here is an illustrative ranking of four reporting features:

These scores were rescaled to total 100. A score of 38 does not mean 38% of customers will buy scheduled reports. It shows the strongest relative preference among the items tested. Ask why it leads, and whether the ranking holds for the customers you serve.

MaxDiff, Kano, or conjoint?

Add engineering effort, risk, accessibility, and business goals when using any of these results to plan work. Preference alone does not decide a roadmap.

How does SenseFolks use MaxDiff?

FeaturePriority uses an adaptive sequence. It starts with Kano, uses MaxDiff when enough features advance, and uses pairwise comparisons to refine the shortlist. Creators enter the feature list; the respondent’s answers determine which stages follow.

Create a FeaturePriority survey, write concrete feature titles, and pilot it with relevant users before embedding it. Review which features advance as well as their final ranking.

Frequently asked questions