When you publish an anonymous survey, no identifying data is attached to any response. The dashboard shows aggregate counts only. Per-respondent CSV export is blocked. This page explains the rules in detail.
What anonymous means in practice
For an anonymous survey, Surveys does not record which Slack user gave which answer. The response row in the database has no user reference. That means:
- Charts show counts and percentages, not “Sarah picked option A.”
- The CSV export contains one row per answer, with no name column.
- You cannot follow up on a single response inside Surveys.
The three-respondent floor
Below three respondents, Surveys hides results entirely. The dashboard shows a message explaining why. This is a privacy guardrail. With one or two respondents on an anonymous survey, results can sometimes be linked back to a specific person by elimination.
If you need results from a smaller group, run the survey without anonymity.
The toggle locks at publish
Once you publish an anonymous survey, the anonymity setting is locked. You cannot turn it off after responses come in. This is deliberate. People who answered under the promise of anonymity should keep that promise.
If you accidentally published with the wrong setting, the workaround is to end the current survey, create a new one with the right setting, and re-invite the same audience.
Channels and DMs both honour anonymity
The anonymity setting applies to every response, regardless of whether the survey was sent to a channel, to a list of DMs, or both. Channel membership is not used to identify respondents.