Skip to content
Open fromusers

MaxDiff

Pro plan

MaxDiff (best-worst scaling) shows participants small sets of items and asks which is best and which is worst in each set. Across sets, you learn relative importance without forcing a full ranking of every item.

Use MaxDiff when the list is too long for comfortable Ranking, or when you need sharper discrimination among many features, messages, or benefits. Use Conjoint analysis when preference depends on combinations of attributes and levels, not a flat item list.

  1. Open Build → Form and select a page.
  2. Choose Add content → MaxDiff (Best-Worse) (Pro).
  3. Enter the Options to compare and generate sets.
  4. Configure set size, interface, labels, and advanced validation options.

Available in surveys, usability tests, and in-product surveys on the Pro plan.

SettingOptionsNotes
Options per set3, 4, 5, or 6How many items appear in each trade-off
InterfaceStandard list, Modern cardsVisual presentation
Best / Worst labelsCustomizableClarify what “best” means in your study
Show descriptionsOn/offExtra detail under options
Randomize option orderOn/offOrder within a set
Allow no preferenceOn/offPermits ties when enabled
Options to compareEditable listThe full item inventory
Generated setsAuto-builtReview after generation
Min sets per participantAdvancedCoverage control
ValidationLenient, Standard, StrictHow strictly answers are checked
Show progress barOn/offWithin the MaxDiff exercise
Navigate between setsOn/offAllow moving among sets when enabled

Items: twelve onboarding benefits. Options per set: 4 · Interface: Modern cards · Best/Worst labels: Most important / Least important · Allow no preference: off.

MaxDiff analysis focuses on:

  • Utility score
  • Summary table

Higher utility generally indicates higher relative importance. Report scores with your sample size and exercise design noted.

  • Requires Pro.
  • Poor item wording reduces discrimination even with a solid design.
  • Very large inventories need enough sets per participant for stable scores.
  • Keep item wording parallel and concrete.
  • Pilot one or two sets to confirm best/worst labels are unambiguous.
  • Do not mix unrelated dimensions in one MaxDiff unless the question framing makes the trade-off fair.

Common questions

How many options appear in each set?

You choose 3, 4, 5, or 6 options per set. FromUsers generates the sets participants see.

Can participants skip a preference?

You can allow a no preference option so participants can indicate a tie when they cannot choose best and worst.

What does MaxDiff show in results?

Results Studio highlights utility scores and a summary table so you can compare item importance.