Want to see how this works in practice? Request a demo.
How AAwiz shows organisations how their teams are doing, while making it impossible to see what any individual answered.
Employees will only answer honestly if they are certain their answers cannot be traced back to them. Good intentions are not enough to guarantee that; rules built into the software are. These are the rules AAwiz follows.
Joining is not sharing. An employee who joins their organisation in AAwiz shares nothing. Adding five anonymous work questions a week is a separate choice that is off until the person turns it on, and it can be stopped at any time.
At least five, every time. A team result is shown only when at least five people in that team answered that week, and the team has at least five members. A smaller group shows no result at all. The organisation cannot lower this number.
No subtraction tricks. Hiding small groups is not enough on its own. If a department of twelve shows a result and one of its two teams shows a result, the other team could be worked out by subtraction. AAwiz therefore hides additional results whenever subtraction would reveal a hidden group. The same applies when one person's answers would dominate a result.
Ranges, not exact counts. The organisation sees how many people answered as a range, such as "5 to 9", never an exact number. Exact counts would reveal who joined or left.
Weekly, and never rewritten. Results are published once a week. A published week never changes afterwards, so comparing versions of the same week reveals nothing.
A quiet start. In the first three weeks after an organisation starts, nothing is published. This lets the group settle before results appear, so early joiners cannot be singled out.
Moving people does not help. When one to four people join or leave a team, its results pause for four weeks. An answer counts only if the person was still in the same team when the weekly results were calculated.
What the organisation never sees:
- individual answers;
- who answered and who did not;
- anything from the personal app, such as conversations with Elin or personal check-ins.
No free text, no AI on employee data. The five questions use a scale from 1 to 5, with no open comments that could identify someone. No AI analyses employee data for the organisation; results come from fixed, transparent rules.
The honest limit. In a team of exactly five, a result that is shown means all five answered that week. AAwiz says this plainly in the consent text, because trust depends on stating limits, not hiding them.
These rules were tested by independent reviewers who tried to identify individuals from the organisation's view. Every route they found was closed before release.