Ranking

The Ranking question asks respondents to drag and drop a list of items into their preferred order — most to least important, most to least preferred, and so on. Because respondents must make trade-offs to order the full list, the data reveals true priorities rather than the "everything is important" inflation that plagues rating questions.

When to use it

Reach for Ranking when:

• you need to understand relative priority, not just absolute scores
• you are doing feature prioritisation and want stakeholders to be forced to choose
• you suspect a rating question will produce tied scores across all items

How to configure it

Open the question in the builder and expand Settings:

• Items — add, edit, and reorder the list of options to rank.
• Partial ranking — require respondents to rank only the top N items instead of the full list.
• Shuffle on load — randomise the initial order to prevent position bias.
• Required — force a ranking before the respondent can continue.

Tips

Tip — Keep the list to 5–7 items. Ranking 10 or more items is cognitively exhausting — respondents lose precision and care by the bottom of the list, which undermines the quality of the lower-ranked data.

Tip — If you have more items than can be ranked comfortably, use MaxDiff instead. MaxDiff shows small sets at a time and statistically aggregates the results — it is far more accurate for long lists.