Anonymous, and the server is what makes it true.
A pulse survey can be marked anonymous, in which case the server stores no respondent identity with the answers.
Surveys are on Command and HQ. Anonymity is not an add-on and not a setting anyone can quietly turn off afterwards. It is decided when the survey is created and enforced where the answer is written.
1How is the new shift pattern working?
- 566
- 462
- 325
- 27
- 13
2What would you change?
- Better, but the handover is still rushed.
- We need one more pair of hands on the late shift.
No identity is written with an answer. A separate participation record knows that you answered, and never what you said. It is two records kept apart, not a promise in a policy.
And you can still ask the ones who have not answered.
The usual trade is real: promise anonymity and you give up the ability to chase anybody. Here you do not, because participation is recorded apart from the answers.
- Responded163
- Asked214
- Response rate76%
Participation recordThat they answered. Never what.
- Amelia DawsonAnswered
- Simon NaylorAnswered
- Zoe MitchellNot yetAsk again
A second submission from the same person is refused, and this record never holds what anybody wrote.
Three people, and what each of them gets.
- Dr Tessa PalmerHead of clinical governance
Asks 214 people how the new shift pattern is working, and asks the same question again next month so the answer becomes a line rather than an anecdote.
Comms Admin - Simon NaylorRegistered nurse
Says what he actually thinks, because on an anonymous survey the server writes no identity next to his answer. That is enforced where the answer is stored, not promised in a policy.
- Zoe MitchellPorter
Answers on the phone she already owns. She joined with an access code and has no company mailbox, and answering carries no role requirement at all.
Short, targeted, and repeatable.
Command and HQ workspaces can run pulse surveys of up to 20 questions, using rating, free text, multiple choice, yes/no or an NPS scale alongside content blocks that ask nothing. Target them at a saved audience segment, repeat them monthly or quarterly as a named series, track response rate against audience size, and export responses as CSV.
- Up to twenty questions, six kinds
Rating, free text, multiple choice, yes/no, an NPS scale, and content blocks that ask nothing and simply explain why you are asking.
Command · HQ - Aimed at a saved audience
Build the audience from department, role, job title, Space membership, onboarding status and the date somebody joined, combined with AND or OR, and see the member count before you send.
Command · HQ - A response rate you can read
Track how many answered against the size of the group you actually asked, rather than against the whole workspace.
Command · HQ - Repeat it on a cadence
Run it monthly or quarterly as a named series, so the same question becomes a line on a chart instead of a one-off. Leaving the cadence empty is a real answer too: some questions should only ever be asked once.
Command · HQ - Take the answers with you
Export the responses as CSV. Nothing in the product is held hostage.
Command · HQ
Where anonymity stops.
Every line here traces to something in the code, including the parts it would be easier to leave out.
- Anonymity covers the answers, not the fact of answering
A separate participation record notes that a person responded, and never what they said. That record is the whole reason a second submission can be refused and the people who have not answered can still be asked again.
- A small enough group is a way back
Response rows still carry a workspace id and a submission timestamp. On a very small audience segment that combination remains a theoretical route to working out who said what, and pretending otherwise would be the easy thing to do here.
- It is a pulse, not a research instrument
There are no ranking, matrix or branching questions, and a survey cannot change what it asks you based on an answer you gave earlier. A recurring series repeats monthly or quarterly, and that is the entire list.
- Two roles can create one, everybody can answer
Building a survey and reading its results sit behind the Super Admin and Comms Admin gate. Answering carries no role requirement at all, which is the half that matters: everybody in the workspace can be asked.
- An NPS scale is a question type
You can ask a nought-to-ten question and export the answers. There is no separate scoring engine and no benchmark set behind it, and nothing here is sold as one.
What sits either side of a survey.
- The recordThe opposite instrument. A named confirmation against a person and a time, where a survey deliberately keeps no name at all.
- AnnouncementsWhere you tell people what the answers changed. A survey that visibly changes nothing gets answered once.
- AnalyticsResponse rates over time, beside the rest of the numbers about the workspace.
Ask the people who never reply to email.
Staff who join with an access code and a phone number answer surveys the same way everybody else does. Kayden Connect is invitation-only before launch.
Request early access- Command and HQSurveys, anonymous or attributed.
- Two recordsThat you answered, and what was said. Kept apart.
- Yours to keepEvery response exports as CSV.