Should I use a poll or a survey?

Use a poll when there is one question and the result should be visible — a decision, a vote, a show of hands. Use a survey when there are several questions and the results are for you. They share an editor; the type sets the defaults for results and voting limits, and you can change either.

Can I show a question only to some respondents?

Yes. A question can be shown only when one earlier answer matches — for example a “what went wrong?” box only for people who said their issue was not resolved. A condition compares against a single answer; it cannot test a range or combine conditions.

How do I keep a survey anonymous?

Do not add a name or email question, leave “Collect email” off, and send everyone the same link. Each response still records the time, the browser type and the page it was sent from. In a small group, also drop questions such as department and tenure that together identify one person.

How do I run the same survey every month?

Duplicate it from the dashboard for each round and give the copy the period in its name. Each copy has its own responses and its own results, so the rounds stay separate and comparable.

Can a poll be used for a formal election?

It should not be. Vote limits stop casual repeat voting but do not verify who a voter is, and ranked votes are reported as average rank rather than counted by instant runoff. Use election software for a binding vote.

← Docs

Survey and poll use cases, and how to set each one up

Six jobs that cover most of what surveys and polls are used for. Each names the template to start from, the two or three settings that matter, and what to read when the answers are in.

Customer feedback you can track

Start from the NPS survey template (all customer templates).

Ask: A 0–10 recommend question, one open “why”, and a segment.
Set: Results: Only me. Voting limit: none, or one per browser if you link it from a single email.
Read: The NPS and its promoter / passive / detractor split on the Results tab; the reasons in the CSV, sorted by score.

An employee survey people answer honestly

Start from the Employee engagement survey template (all employee templates).

Ask: Agreement statements grouped into matrices by theme, plus an eNPS question. No name, no email.
Set: Results: Only me. A closing date, announced up front. One link for everyone — never a personal link per person.
Read: One stacked bar per statement. Look for the statement people disagree with, not the overall average.

A decision in a meeting

Start from the Yes / no poll template (or make one with no account).

Ask: One question.
Set: Results: Always, on a shared screen — or After voting if early votes should not sway later ones. One vote per browser.
Read: The live results page at the poll’s address plus /results.

Choosing between several options

Start from the Ranked vote template (or a single-choice poll).

Ask: Everyone puts the options in order.
Set: Results: After voting closes, with a closing date, so standings do not invite tactical ranking.
Read: Average rank and first-place votes per option. It is not an instant-runoff count.

Deciding what to build

Start from the Feature prioritization survey template (all research templates).

Ask: A hundred points to spend across the candidates, then a ranking.
Set: Results: Only me. Add a segment question so you can weight answers in the export.
Read: Total and average points per feature — a high total from few people is a niche need.

Feedback right after an event or a session

Start from the Post-event survey template (or training feedback).

Ask: An overall rating, a matrix for each part, and “would you come again”.
Set: A closing date a week out. One vote per browser. Share the link before people leave the room.
Read: The lowest-rated matrix row and the split on attending again.

What all of them share

Respondents need no account. Every submission passes a bot check before it is stored. Every response stays in your dashboard and exports as CSV, and can additionally be sent to a webhook or a chat channel as it arrives.

Where these stop

None of these recruits respondents for you, branches on score ranges, or analyses one question against another. If a job needs that, this page says which kind of tool to use instead.

Common questions

Should I use a poll or a survey?: Use a poll when there is one question and the result should be visible — a decision, a vote, a show of hands. Use a survey when there are several questions and the results are for you. They share an editor; the type sets the defaults for results and voting limits, and you can change either.
Can I show a question only to some respondents?: Yes. A question can be shown only when one earlier answer matches — for example a “what went wrong?” box only for people who said their issue was not resolved. A condition compares against a single answer; it cannot test a range or combine conditions.
How do I keep a survey anonymous?: Do not add a name or email question, leave “Collect email” off, and send everyone the same link. Each response still records the time, the browser type and the page it was sent from. In a small group, also drop questions such as department and tenure that together identify one person.
How do I run the same survey every month?: Duplicate it from the dashboard for each round and give the copy the period in its name. Each copy has its own responses and its own results, so the rounds stay separate and comparable.
Can a poll be used for a formal election?: It should not be. Vote limits stop casual repeat voting but do not verify who a voter is, and ranked votes are reported as average rank rather than counted by instant runoff. Use election software for a binding vote.

Keep reading

TemplatesEvery use case here, ready to open in the editor.Live resultsThe four visibility settings.One vote per personChoosing a voting limit.Is Votense the right tool?What it does not do, and what to use instead.