Should 360 feedback be anonymous? Yes, for developmental 360s at mid-market companies, anonymity is usually the right default so peers and direct reports give candid behaviour examples - but it is not absolute. You still need enough raters per relationship (typically three or more peers and three or more reports) so individuals cannot be reverse-identified, a clear purpose (development versus pay), and a manager-led debrief. Anonymous-for-all-time with no follow-up coaching wastes the data.
Key takeaways
- Default to anonymous for peers and reports. Most schemes promise rater anonymity because it reduces fear of retaliation and supports honesty (British Psychological Society best-practice guidelines).
- Protect the floor: three to five raters per group. BPS guidance sets a minimum of three to five people so individual sources cannot be identified; combine groups if a team is smaller.
- Purpose beats volume. CIPD’s 2022 evidence review finds multi-source improvements over time are small, and more sources do not outperform feedback from direct reports alone.
- Developmental intent shows larger gains. Smither, London and Reilly’s 2005 meta-analysis of 24 studies found small average effects (about corrected d = 0.15 from reports and supervisors) that were larger when the purpose was developmental rather than administrative.
- Close the loop with a debrief. Anonymity without coaching leaves themes unread. Pair the report with a manager-led conversation and a short 90-day plan.
What “anonymous 360” actually means
In a multi-source or 360 process, anonymity usually means peer and direct-report ratings are aggregated so the recipient sees themes, not named comments from each person. The manager rating is often visible, and self-ratings are always named. Bracken, Rose and Church (2016), writing in Industrial and Organizational Psychology, note that anonymity has long been treated as a foundation of 360 feedback because it promotes greater honesty - especially for peers and reports, who face the highest risk of retaliation.
That is why the practical UK answer to “should 360 feedback be anonymous?” is: yes as the default for developmental 360s, with hard rules that keep anonymity real. BPS best-practice guidelines are explicit: most schemes promise anonymity for raters, and the sample must be large enough that individual sources cannot be identified - a minimum of three to five people depending on the circumstances. If a team has fewer, combine groups (for example peers with reports) rather than showing a named cell of one.
Anonymity is a design choice about trust, not a substitute for clarity. CIPD’s 2022 performance-feedback evidence review lists the conditions that make multi-source feedback useful: a developmental purpose, recipient involvement in choosing raters, help interpreting the results, follow-up discussions and organisational support to act. Hide names without those conditions and you collect quiet data nobody uses. For the wider shift away from one annual meeting, see our pillar on alternatives to annual performance reviews and the companion note on why feedback never sticks.
When anonymity helps - and when it does not
Use this table to decide what stays anonymous and what can be named. The goal is candid behaviour examples for development, not a secret court.
| Rater group | Usual anonymity | Why | Minimum to keep anonymity real |
|---|---|---|---|
| Direct reports | Anonymous, aggregated | Highest retaliation risk; honesty depends on cover | 3–5 reports; combine groups if fewer (BPS) |
| Peers | Anonymous, aggregated | Same honesty rationale as reports (Bracken et al., 2016) | 3–5 peers; do not show a single-peer cell |
| Manager / skip-level | Often named | Accountability sits with the line; named feedback is expected | One clear manager view, written and discussable |
| Self | Named | Self-other gaps are the coaching material | Same framework as others; no hidden self score |
| Pay / bonus input | Use with extreme care | Administrative purpose shrinks gains (Smither et al., 2005) | Prefer separate evidence for pay; keep 360 developmental |
Smither, London and Reilly’s 2005 meta-analysis of 24 longitudinal studies found small average improvements after multi-source feedback - corrected effect sizes around d = 0.15 from direct reports and supervisors and about 0.05 from peers - with larger effects when the purpose was developmental (roughly 0.25 across sources) rather than administrative (about 0.08). CIPD (2022) adds that improvements over time are small and that piling on more sources does not beat feedback from direct reports alone. Anonymity supports honesty; purpose and follow-up decide whether honesty turns into better work.
How to run an anonymous 360 without reverse-identification
- Write the purpose in one sentence. “This 360 is for development and coaching; it is not an input to bonus or redundancy.” Share it before invites go out. Administrative use shrinks the gains Smither and colleagues measured.
- Set the rater floor before you launch. Require at least three peers and three reports wherever those groups exist. If a manager has two reports, combine peers and reports into one anonymous pool rather than showing a two-person cell (BPS: three to five minimum).
- Let the recipient help choose raters. CIPD’s success factors include recipient involvement. Raters must be people with recent working contact - not a popularity contest of distant stakeholders.
- Use behaviour items, not personality labels. Ask for observed examples against skills and values, the same bar you use in manager training for performance management.
- Aggregate before anyone sees a report. Strip names from peer and report comments. Keep free-text that would identify a single person out of the shared pack, or edit it with the rater’s consent.
- Run a manager-led debrief within two weeks. Themes without a conversation die in the inbox - the same failure mode we cover when improving the performance review process. Agree three actions and a check-in date.
- Keep a written record of the plan, not the raw named gossip. Acas expects fair process and a written record of performance conversations. Store the development plan and follow-ups; do not recirculate identifiable rater rows.
Risks if you get anonymity wrong
- Reverse-identification. Two reports and a distinctive writing style effectively names people. That breaks the promise BPS says most schemes make and destroys trust for the next cycle.
- Anonymous forever, no coaching. CIPD’s review is clear that multi-source feedback needs interpretation support and follow-up. A PDF in a drawer is not development.
- Using 360 scores for pay. Smither et al. found smaller effects under administrative purposes. Mixing anonymity with bonus maths also invites fairness challenges.
- Feedback that harms performance. Kluger and DeNisi’s classic meta-analysis, summarised by CIPD (2022), found that over a third of feedback interventions left performance worse. Poorly framed 360 comments without coaching sit in that risk set.
- Changing access rules mid-cycle. BPS warns that changing who can see ratings is a common reason trust collapses. Lock access before launch.
If your wider review process already feels noisy, run our free performance health check before you add another survey layer.
How to measure whether anonymity is working
- Completion without chase drama. Peer and report completion above your usual survey baseline suggests people trust the cover.
- Comment usefulness. Count behaviour examples versus vague labels. Anonymity that only produces “great colleague” is cover without candour.
- Debrief rate. Percentage of recipients with a documented coaching conversation within 14 days of the report.
- Action follow-through. Share of plans with at least one progress note at 30 and 90 days.
- No retaliation flags. Track informal complaints or engagement dips in small teams after a 360 - a signal the floor was too low.
For bias and fairness checks around ratings more broadly, pair this with performance review bias and your performance review checklist for HR managers.
Our perspective
When mid-market People teams ask me whether their 360 should stay anonymous, they usually already know the answer they want: hide the names and hope the comments get braver. The part that fails is the floor. Two reports and a distinctive turn of phrase is not anonymity, it is a labelled whisper. Set three to five raters per group, keep the purpose developmental, and put a manager in the room within a fortnight to turn themes into a plan. Quiet data with no coaching is just another PDF nobody opens.
Ash
How Revolut People supports anonymous multi-source feedback
I work at Revolut People, so read this section with that in mind. On anonymity and access, what matters for this question is permissioning and minimum-group rules rather than a slogan. Revolut People lets you separate who can see ratings and feedback, hide colleagues’ sensitive performance data by default, and set survey anonymity thresholds so results only appear when a minimum number of responses and a minimum group size are met - the same logic as a three-to-five rater floor. Teams that already run scorecards and calibration can keep 360 themes in the same operating rhythm as goals and 1:1s instead of a side spreadsheet. Customer stories such as Ebury and Sertis show how mid-market and scale-up People teams put performance routines on one system; for access detail see Revolut People security and access.
If you want a light fit-check on whether your current 360 rules would survive a three-rater floor and a coaching debrief, book a 15-minute 15-minute review. Soft ask only - bring your rater counts and purpose statement.
90-day plan
- Days 1–30 - Rules. Publish purpose (developmental), rater floor (3–5), access list and debrief owner. Kill any bonus link for this cycle.
- Days 31–60 - Pilot. Run one leadership cohort with aggregated peers and reports, named manager view and a facilitated debrief within 14 days.
- Days 61–90 - Harden. Measure completion, comment quality, debrief rate and 30-day action notes. Adjust group-combining rules before a wider rollout.
Bottom line
Should 360 feedback be anonymous? Yes as the default for peers and reports in a developmental 360 - provided you keep three to five raters per group so anonymity is real, you separate development from pay, and you close with a manager-led debrief. Anonymity without coaching is just quiet data. Start with the rater floor and the purpose sentence; the survey tool is the easy part.
FAQ
Does 360 feedback have to be anonymous?
No. Anonymity is the usual default for peers and direct reports so people can give candid behaviour examples without fear of retaliation, which is why most schemes promise it (British Psychological Society best-practice guidelines). Boss and self ratings are often named. Keep enough raters per relationship so individuals cannot be reverse-identified.
How often should 360 feedback be used?
Treat a full 360 as an occasional development input, not a weekly ritual. Many mid-market teams run it once a year or every 18 to 24 months for leaders, and lean on regular 1:1s the rest of the time. Acas expects performance conversations at least yearly with informal check-ins; a 360 should feed those conversations, not replace them.
Is 360 feedback effective?
It can be, modestly, when the purpose is developmental and someone helps interpret the results. Smither, London and Reilly (2005) found small average gains across 24 studies (about corrected d = 0.15 from reports and supervisors; 0.05 from peers), larger when developmental. CIPD (2022) notes multi-source gains stay small without coaching follow-up.
Who should give 360 feedback?
People with recent, direct working contact: the manager, peers, direct reports where they exist, and sometimes cross-functional partners. BPS guidance stresses that raters must be credible to the recipient, and that the recipient should help choose who is in a position to comment. Skip distant stakeholders who only see polished presentations.
How does 360 feedback work?
The recipient and stakeholders rate behaviours against a shared framework. Peer and report responses are aggregated and usually shown without names; the manager rating is often visible. A facilitator or line manager then runs a debrief into a short development plan. CIPD’s success factors include a developmental purpose, help interpreting results and support to act.
Sources
- British Psychological Society with CIPD South West London Branch and the UK Department of Trade and Industry: 360 Degree Feedback: Best Practice Guidelines (BPS; PDF hosted 2022) - minimum of three to five raters so individuals cannot be identified; most schemes promise anonymity
- CIPD: Irina Cioca and Jonny Gifford, Performance feedback: An evidence review (2022) - multi-source improvements over time are small; more sources do not beat direct-report feedback alone; Kluger and DeNisi (1996): over a third of feedback studies found performance worsened
- James W. Smither, Manuel London and Richard R. Reilly: “Does performance improve following multisource feedback? A theoretical model, meta-analysis, and review of empirical findings”, Personnel Psychology (2005) - 24 longitudinal studies; corrected d ≈ 0.15 (reports/supervisors), 0.05 (peers); larger effects for developmental than administrative purpose
- David W. Bracken, Dale S. Rose and Allan H. Church: “The current state of 360-degree feedback research and practice”, Industrial and Organizational Psychology (2016) - anonymity as a foundation for honesty among peers and reports; not required for every rater group
- Acas: Reviews and appraisals: performance management (updated 2026) - regular reviews at least yearly, informal check-ins, written record and fair process