Survey Distribution
A survey is addressed to two independent audiences: internal recipients who log into the system, and external participants who do not have an account. They are configured separately and behave differently.
Internal distributionβ
The internal distribution list accepts identities, not just named people:
| Identity type | Supported |
|---|---|
| User | Yes |
| User group | Yes |
| Position | Yes |
| Role | No |
| Organization unit | No |
Standing tasksβ
This is the mechanism worth understanding. When the survey publishes, the system creates one task per distribution entry β not one task per resolved person. A task addressed to a user group stays addressed to that group.
The consequence:
Somebody added to a group after the survey is published still sees the pending task.
Group membership is resolved when the task list is read, not when the survey is published. This is the opposite of how a mail-merge style distribution behaves, and it is deliberate β a survey addressed to "Production Supervisors" should reach whoever holds that role at the time they look.
Answering completes only your own taskβ
A task addressed to a group is shared. When one member answers, their own participation is recorded and their view of the task is completed β the task stays pending for everybody else in the group.
External distributionβ
External participants have no user account. They are addressed by adding external participant records to the survey's external distribution list.
On publish, each new external participant receives:
- A participation record with a unique access token.
- For participants whose contact preference is e-mail, an automatic invitation e-mail containing their personal link.
Participants whose contact preference is phone still get a record and a working link, but no message is sent automatically β there is no SMS integration. You must share their link manually.
Because the link carries a token, external participants answer without logging in, and each response is still attributable to a specific participant.
Publishabilityβ
A survey needs at least one internal distribution entry to be published. An external-only survey cannot be published β this is a known limitation, not a configuration problem.
Republishing after adding recipientsβ
When you add recipients and publish again, the system diffs both lists:
- New internal identities get a new standing task.
- New external participants get a new record and invitation.
- Existing recipients are untouched β nobody is invited twice.
A republish that adds only external recipients succeeds. A republish that adds nobody new is refused.
Prefer groups and positions over named users in the distribution list. It is the difference between a survey that stays correctly addressed as people join and leave, and one that has to be rebuilt every time the organization changes.