Confidential vs anonymous — use precise words
Truly anonymous usually means the platform never stores a link from response to identity. Confidential often means identity exists for administration (reminders, eligibility) but is withheld from managers and reports. Say which model you run — overclaiming “anonymous” when you mean “confidential with thresholds” destroys trust when someone asks a smart question.

K-anonymity thresholds by product
Corpmetrix enforces k-anonymity at query time: no dashboard, pack, or export shows a group smaller than the threshold. Defaults differ by design:
- Engagement and pulse programmes: default threshold of 5 — team reports unlock only when enough people responded.
- 360° feedback: default threshold of 3 — rater groups (peers, directs) are naturally smaller; a higher bar would suppress almost every cut.
- Clients can raise thresholds; lowering them is a governance decision, not a convenience toggle mid-fieldwork.
Access control and comment safety
Role-based access keeps board packs, HR analytics, and manager scorecards in separate permission lanes. Comment intelligence links themes to sample quotes only when k-anonymity rules are met — so managers discuss patterns, not people.
Communicate the rules before fieldwork
Trust rises when the invite states: who can see what, the minimum group size, whether open text can appear in manager packs, and how long identifiable admin data is retained. Put the same text in Arabic and English when you run bilingual programmes.
Frequently asked questions
Can HR see who said what?
In a properly configured programme, HR sees aggregates and k-anonymous quote samples — not a named comment feed. Admin identity for reminders is separated from analytic views.
Why is the 360 threshold lower than engagement?
Because peer and direct-report groups are small by nature. A threshold of 5 would hide most 360 cuts; 3 balances honesty with usefulness for development reports.

