MaxDiff is a research technique used to identify consumer preferences among a set of attributes (ex: flavors, product features, etc.). MaxDiff questions show respondents subsets of attributes, and for each set require respondents to select their best and worst options for the given set.  

1. To launch a MaxDiff survey, go to the create action menu and select MaxDiff survey. 
2. Enter an internal label, internal description, research type, respondent-facing title, and respondent-facing description for your survey, and click next.
3. Set up any desired targeting and quotas for your survey. You'll notice the response cap is greyed out on this action. The platform will automatically determine your ideal sample size based on the number of attributes you're testing. Due to the respondent minimums for MaxDiff surveys, use care and good judgment when setting up any targeting and quotas to ensure you can meet the minimum respondent requirements. Click next.
4. Add your MaxDiff question and any other survey questions. You must add at least 1 MaxDiff action in order to proceed. For your MaxDiff action, indicate the number of sets you'd like to use.
   You cannot use in-step targeting for your MaxDiff action, meaning you cannot target your MaxDiff to another question in your survey. This has to do with the strict respondent minimums around MaxDiff actions. Your MaxDiff action MUST be completed by every respondent in the survey. Click next.
5. In the review page, you can preview your survey, add tags and launch. 

We strongly recommend you allow your MaxDiff action to fill to the minimum number of respondents (noted above). If your MaxDiff action is closed early or does not fill to the specified number of respondents, there is a possibility that your study is not experimentally balanced, potentially causing exposure biases towards certain attributes. If you must close a MaxDiff action early, your results should be interpreted in a more directional sense.

Learn more about MaxDiff surveys and analysis.