How to encourage team feedback without undermining trust or performance reviews

Executive overview

Leaders often conflate two distinct types of feedback: formative (coaching for growth) and summative (evaluation for ranking). Mixing them in the same process destroys psychological safety and kills the very improvement they aim to create. The nine-box grid has similar problems — its "potential" dimension is both hard to measure and shown to be discriminatory.

Separating feedback type from feedback process is the prerequisite for any feedback system that actually works.

Why anonymous peer feedback tied to salary backfires

• Telling employees their feedback is confidential, then using it in evaluations, breaks trust irreparably
• Employees will withhold honest feedback if they believe it could harm a colleague's pay
• People in a change process need psychological safety to fail and try again — evaluation stakes remove that
• Appreciation, coaching, and evaluation are three distinct feedback types (Sheila Heen, Thanks for the Feedback, ep. 143) — conflating them produces none of them well
• The manager's role in most organisations is to evaluate; peer evaluation is a high-trust practice that requires significant groundwork first

Building toward a feedback culture: where to start

• Establish team norms first: what does good feedback look like here, and why do we do it?
• Use feed forward (Marshall Goldsmith) as a low-stakes entry point: each person names something they want to improve; others offer one suggestion for the future only — no evaluation of the past
• Feed forward works even with strangers or low-trust groups; it separates improvement from judgment
• Incrementally build comfort with giving and receiving feedback before moving to anything more formal

Why the nine-box grid is problematic

• Originated at McKinsey; rates employees on a 3×3 grid of performance × potential
• The "potential" dimension is subjective, hard to measure, and bakes in bias
• Research shows women apply for roles only when they feel 100% qualified; men at 60–70% — this gap shows up directly as lower potential ratings for women
• Self-evaluations compound the problem: people underrate themselves, reinforcing the bias systemically
• Result: lower ratings for women and people of colour get formalised and repeated

Alternative frameworks for talent mapping

• Kim Scott's rock star / superstar model (ep. 638): focuses on growth trajectory rather than fixed potential — recognises that people move fast or slow at different times and in different areas
• Skill-based team mapping: list team members on one axis, required skills on the other; rate each cell as ready to execute / has potential to develop / unlikely to develop for this role — surfaces succession gaps and investment priorities without labelling whole people
  • Avoid letter grades (A/B/C) — too much baggage from school; use neutral symbols
  • Combine with strengths data (e.g. StrengthsFinder) to shape roles around energy, not just gaps
• Ongoing informal feedback embedded in culture makes formal annual evaluations less necessary

Finding meaning in leadership: practices that work

• Daily gratitude prompt: write one thing you are grateful for, even if small or obvious
• Compassion prompt: where did you feel or show compassion today? — keeps focus on human connection
• Start team meetings with a client or member success story — grounds the team in purpose, especially during hard periods
• Send specific, unsolicited notes of recognition — one precise observation lands far better than generic praise
• Servant leadership (Robert Greenleaf, 1970): the leader seen as follower because they serve so well — reframe thanklessness by actively serving others
• When leadership feels thankless, send encouragement outward — joy compounds