Long surveys, low drop-off

Forty questions is fine — if each one appears alone, irrelevant sections skip themselves, and a closed tab doesn’t erase twenty answers.

The recipe · a user-research survey

Three moves, one result.

SCREEN

Two screener questions up front — people outside the target group exit early to a polite ending

ASK

Matrix grids, 0–10 scales, rankings and open text — randomized choices where order could bias

KEEP

Autosave holds every answer — participants finish in two sittings instead of abandoning at question 23

Tag cohorts with URL params (panel, newsletter, social) — responses arrive pre-segmented.

Length is not what kills a survey. Density is.

Forty questions one at a time beats fifteen crammed onto a page you have to scroll.

A question grid on one page

The scrollbar shows how much work is ahead, and half the panel decides not to start.

Everyone answers every section, relevant or not.

A closed tab erases twenty minutes of somebody else’s time.

You discover the screener let the wrong people through after the data is in.

A study that respects the participant

One question per screen — the length stops being visible and completion holds up.

Screeners route people out early and politely; sections that do not apply skip themselves.

Autosave means the long tail of the survey gets answered in two sittings, not abandoned.

Choice randomization keeps question order from bending your results.

What does the work

What does the work

Templates

Or start from a template

Multi-section studies with scales, matrices and open text — a working structure to start from.

Conduct research: all 17 templates →
FAQ

Survey & research, in practice

Is there a limit on how many questions a form can have?

No. Forms, questions per form, workspaces, logic rules and variables are all unlimited on every plan, including Free. What the plans meter is responses per month.

Can I screen people out mid-survey?

Yes — send anyone outside the target group to a dedicated ending screen at any point in the flow, so they never reach the main body of questions and never muddle your results.

Can answer options be randomized?

Yes, per question — so the order choices appear in does not bias what people pick. Checkboxes also take minimum and maximum selection limits.

How do I separate cohorts?

Tag each distribution link with a URL parameter — panel, newsletter, social. The value is stored on the response, so you can filter or branch on it without adding a "where did you hear about us?" question.

How do I get the raw data out?

Export any form’s responses to CSV, on every plan — or webhook each response into your own pipeline as it arrives. There is no lock-in: the data is yours to take out at any time.

Better data, question by question

Free forever for 100 responses a month — the whole builder, no card.