Open text questions
Open text answers are read word for word, shuffled and undated. See the limits, the recurring-word themes, and how to keep written answers anonymous.
- Admins writing surveys
- 6 min read
- Updated Oct 29, 2026
Free text is the highest-risk question type in any anonymous survey; not because of how it's stored, but because of what people write in it. The controls below exist for that reason.
An open text question collects a paragraph: "Anything else you'd like to share?" The answer is encrypted in the respondent's browser like every other answer. Once results unlock, you read each answer word for word, in a shuffled order with no dates attached. If someone sent only topics, you'll see those topics instead of their words, once enough people picked them. See Send topics instead of your words.
#Open text vs. word cloud
They look similar and behave very differently.
| Open text | Word cloud | |
|---|---|---|
| Length | 200-5,000 characters (default 1,000) | Up to 50 characters |
| What you read | Written answers, word for word | Every word, counted by how many people used it |
| Best for | "Tell us what happened" | "One word for how you feel" |
If you want sentiment at a glance, use a word cloud. If you need the story, use open text; and accept that you're asking people to write something that could identify them if they aren't careful.
#Writing the question
Add a question in the poll wizard and pick Open text. Two fields matter:
- The prompt (up to 200 characters). Ask for the thing you'll act on. "What is one thing that would make next quarter better?" produces more usable answers than "Any other comments?"
- Helper line under the question (optional, up to 200 characters). This is where you set the safety expectation.
Please avoid naming individuals.is short, and it changes what people write.
Set the character limit deliberately. The default of 1,000 characters is roughly 150 words. Raising it to 5,000 invites the long account of a specific incident; which is sometimes exactly what you want, and sometimes the thing that makes an answer identifiable in a team of nine.
#Reading the results
Nothing's readable until the poll's minimum group size has answered, and that's never below three. While the poll's open, the first results appear three answers after that, then new answers reach you in random groups of three to five, with a few always held in reserve until the poll closes. At close, everything shows. The response count on your dashboard still goes up with every answer. How batched release works.
The results panel shows three things:
- How many people wrote something.
- Recurring words, counted by how many distinct people used each. A word appears only when at least two different respondents used it; one person's vocabulary isn't a theme, and surfacing it would point straight at them.
- The answers themselves, word for word, in a shuffled order that has nothing to do with when they were submitted. Export gives you a Written answers (CSV) download for your own records.
Two more things apply on top:
- Cohort filters narrow the list, once the poll has closed. Breakdowns only appear after close. If the poll has demographic dimensions, filtering to a group then shows only that group's answers. A group below the floor (five, or the poll's minimum group size if higher), or one that could be worked out from the others, stays locked rather than showing a short list, and no reason is given. See Demographic cohorts and thresholds.
- An optional summary, computed on your device. On a desktop browser, once at least five people have answered, you can download a small model once and ask it for themes and a short summary. It runs in your browser tab, and nothing about the answers is sent to us or to an AI provider to produce it. Full account: On-device AI summaries.
If you turned on anonymous replies for the poll, each answer also carries the control to reply to its writer without learning who they are.
#When people send topics instead of their words
Some people won't write anything if they think their phrasing could be recognised. So under every open-text box, respondents can choose Only send topics, not my words. They pick up to five topics, from suggestions drawn from what they typed or from a fixed list like workload, pay or recognition, and can add a tone. Their words stay on their device and never reach you. It's on by default for new polls, and you can turn it off per poll in the builder's Distribution step, under Comments.
Under the written answers, a Topics only group shows:
- How many people answered that way. Until five have, that's all it shows.
- Each topic at least two people picked, with its count.
- A tone line, if people tapped one.
The topic counts are in the aggregate CSV and the PDF report too. They're never in cohort breakdowns, shared slices, published results or the on-device summary. You can't reply to a topics-only answer, and speak-up polls don't offer the choice. See Send topics instead of your words.
#Written answers stay in the signed-in results view
If you publish a poll's results to a public results link, the recurring words and the response count are included. The written answers aren't, and neither are topics-only counts.
A public results link is forwardable. Paragraphs written by employees who were promised anonymity don't belong behind a URL that can end up anywhere. This isn't a setting.
The same rule holds on every other surface that leaves the admin's browser: the aggregate results CSV and the PDF export carry recurring words and counts only, which is what keeps the PDF's "Aggregate data only" footer honest (the written-answers CSV is a separate download that exists only in the signed-in results view); a shared slice for a manager is built from a format that has no field for written answers; and a series' trend snapshot archives counts, never words. An on-device AI summary is excluded from all of those too, and lands in a PDF only if you tick a box for that one export.
#Anonymity in practice
Encryption protects the answer in transit and at rest. It can't protect an answer whose content identifies the writer; "as the only person on the team who was at the Berlin office in March..." is self-identifying no matter how well it's encrypted.
Reduce that risk by using the helper line to ask people not to name individuals, keeping open text out of surveys aimed at very small groups, and preferring a word cloud when you only need the temperature. Nothing's readable until the poll's Minimum group size has answered (never fewer than three), so for a small group, raise it or leave open text out. The Anonymity Coach reminds you of this risk in the wizard.
See also Aggregate results and anti-inference controls, Viewing results, Communicating survey anonymity to employees. Each page covers the adjacent settings and workflow steps.