< Back to all solutions

Maximum Difference Scaling (MaxDiff)

Max license
& Messaging
Set up in minutes

Most appealing concept idea. Most compelling claims message. Most important factor in a decision. There are endless possibilities for wanting to know the superlative items. Max Diff gives you the ability to test many items in a way that isn't overwhelming to respondents, and confidently assess how much more appealing, compelling, important, etc. each item is.

What you'll get
  • Preference likelihood
    The probability that a given item would be selected within the survey.
  • Average-based preference likelihood
    The probability that a given item would be selected when paired with any other one item.
  • Utility scores
    Raw scores, centered around zero, that reflect directional preferences of items.
  • TURF Simulator
    Download a simulator to conduct your own TURF analyses using the data from the MaxDiff.
Common applications

Product claims

Ask respondents which message statements would be the most and least motivating to purchase a product.

Concept development

Ask respondents which concept ideas would be the most and least appealing in a product.

Decision making

Ask respondents which factors would be the most and least influential in making a decision.

Customer experience

Ask respondents which elements were the most and least enjoyable of their experience.

In a typical rank-order exercise, respondents are shown a list of items and asked to put them in order of some characteristic, such as preference or importance. But this method has two important limitations: respondents can only meaningfully engage in this task when there are just a few items, and there is no way to gauge the magnitude of difference between items—is #1 far better than #2, or only slightly better?

Max Diff uses advanced statistical methods to test a large number of items (up to 200) in a way that respondents don't feel overwhelmed. Additionally results are not presented as mere rank orders, but preference likelihoods that can be used to draw conclusions about how much more likely a person is to select one item over another.

Best practices
  • Use when you have more than 6 items in a list.
    Stick with other methods when you have a small list.
  • Review your list carefully.
    It could be tempting to dump a large number of items into a single MaxDiff, but if there are some items that are vastly different and others that are very similar the results are going to misrepresent preferences.

    It's better to sort your items into meaningful categories and run separate MaxDiffs. Use a follow-up study to then pit the winners against each other.

  • Select the best/worst method for the classic MaxDifff experience.
  • Select the image grid option when you want to focus on images over text.