A form and a survey are built with the same tools and are not the same thing. Getting the distinction right up front saves you from the two most common failures in this space: a survey that behaves like a transaction, and a form that behaves like research.
The difference in one line
A form processes one person. A survey measures a population.
Everything else follows from that.
What that changes
Who the answer is about. A form’s answers are about the individual submitting it: their order, their application, their request. A survey’s answers only mean something in aggregate; a single response to a satisfaction survey is noise.
What happens next. A form triggers an action for that person: a ticket opens, a lead lands in the CRM, a registration is confirmed. A survey triggers a decision about a group, usually later, usually after someone has looked at a chart.
Whether identity matters. A form almost always needs to know who submitted it. A survey is frequently better without it, anonymity raises both response rate and honesty.
What length is acceptable. A form can be long if the transaction justifies it; people will fill in twenty fields to apply for a job. A survey is a favour someone is doing you, and the acceptable length is roughly ninety seconds.
Whether branching helps. In a form, branching is almost always right: skip what does not apply. In a survey, branching costs you comparability, two respondents who took different paths cannot be compared on the questions they did not share.
Question types diverge too
Forms want questions that capture a fact exactly: dates as date questions, files as uploads, quantities as dropdowns, addresses as structured fields. Precision matters because someone will act on the specific value.
Surveys want questions that aggregate: scales, grids, and choices. A free-text box in a survey of four hundred people is four hundred paragraphs nobody will read. A free-text box in a work request form is the entire point.
A useful rule: in a form, every question can be free text if it has to be. In a survey, every free-text question needs to justify itself against the reading time it will cost.
Where the two overlap
Plenty of real documents are both, and the honest answer is to be deliberate about which half is which.
A customer feedback form is a survey in structure, a rating you will average, with a form bolted on: “would you like someone to get back to you?” routes to a section that captures identity and creates a follow-up task. Survey for the measurement, form for the exception.
An event registration is the reverse: a form, primarily, that happens to produce aggregate data as a by-product. You are processing individuals, and the catering count is a side effect.
The pattern in both cases is the same: keep the measured questions identical for everyone, and put the transactional questions behind a branch.
Practical consequences
If you are building a form:
- Branch aggressively; nobody should see an irrelevant question.
- Ask for identity where you need it, and validate it.
- Notify a human, or a system, on submission.
- Let people fix their answers afterwards: transactions get corrected.
- Length is fine if the transaction earns it.
If you are building a survey:
- Fix the wording and never change it, or the trend line resets.
- Keep the core questions on the main path for everyone.
- Ninety seconds, five questions, one open box at most.
- Consider dropping identity entirely.
- Duplicate the form per cycle rather than editing a live one, so historical answers stay attached to the wording they were given.
If you are building both: measure first, transact second. Put the survey questions at the top where everyone answers them, and branch the transactional part off the end.
Why tools blur it
Most form builders market to both audiences with one product, because the editor genuinely is the same. The differences show up in the details: whether responses aggregate into charts, whether you can branch per answer, whether a submission can trigger a webhook, whether the same form can be scored.
Questionful does all four, which means the distinction is yours to make rather than the tool’s. That is a feature and a responsibility. It is entirely possible to build a nine-question survey with three branches and then wonder why the results will not compare.
The short version
- A form processes a person; a survey measures a population.
- Forms want precision and branching. Surveys want aggregation and consistency.
- Forms can be long. Surveys have ninety seconds.
- Identity is usually required in a form and often harmful in a survey.
- When a document is both, measure on the shared path and branch the transaction.
If you are still deciding which one you are building, the use-case pages are organised along exactly this line, lead generation and internal workflows are forms; customer feedback is a survey.