Question types
When you add a question to a Static Survey, you pick an answer type. There are nine, plus an NPS preset.
The answer types
Section titled “The answer types”| Type | Participant sees | Use it for |
|---|---|---|
| Short text | A single-line box | Names, roles, one-line answers |
| Long text | A multi-line box | Open feedback, “what would you change” |
| Choice | Radio buttons, pick one | Mutually exclusive options |
| Multi-select | Checkboxes, pick several | ”Select all that apply” |
| Scale | A row of numbers | Agreement, satisfaction, likelihood |
| Rating | Stars | Quick sentiment where stars read naturally |
| Yes / No | A two-option toggle | Binary questions |
| Date | A single date picker | Deadlines, event dates |
| Date range | A start and end picker | Sprints, periods, availability |
The NPS preset
Section titled “The NPS preset”NPS is a one-click preset, not a separate type. Applying it gives you a Scale question configured 0–10 with endpoint labels — the standard Net Promoter Score shape. Once applied it behaves as an ordinary scale question, so you can adjust it like any other.
Switching types without losing work
Section titled “Switching types without losing work”Questions in the same family preserve their content when you switch:
- Choice ↔ Multi-select keeps the options you’ve written
- Scale ↔ Rating keeps the bounds you’ve set
Switching across families — say Long text to Choice — starts that question fresh, because there’s nothing meaningful to carry over.
Choosing well
Section titled “Choosing well”Prefer Long text over Short text for anything open-ended. A short box signals “give me a few words” and that’s what you’ll get.
Reach for Choice only when the options are genuinely complete. If you find yourself needing an “Other” escape hatch for a large share of people, the question probably wants to be Long text.
Scale questions need labelled endpoints to be comparable later. “1 to 5” means nothing without knowing which end is good.
If you’re mainly trying to learn why, consider an Agent Survey instead. Fixed questions are good at measuring something you can already name, and poor at discovering something you can’t.