Many leaders still treat “candid” negative feedback as a performance accelerant. Harvard Business Review’s “When Feedback Crosses the Line” (Bin Zhao, Rebecca L. Dunkailo, Judith Clair, and Ryan L. Boyd — 2 Mar 2026) challenges that habit. When criticism feels belittling or humiliating, it often backfires and impairs performance — the opposite of what the manager intended.

The article draws on research involving employees’ firsthand accounts of destructive criticism. The pattern is familiar to mid-market CPOs: annual dumps, public roasting, character attacks, and judgment with no next step. “Radical candour” without guardrails is not a system — it is a liability.

What HBR is actually saying

Restating the core thread in plain language:

  1. Candid is not the same as useful — leaders often believe blunt negative feedback improves performance; when it feels belittling or humiliating, it tends to impair it instead.
  2. Destructive criticism has a shape — it attacks character rather than behaviour, lacks guidance, and often happens publicly. Related failure modes: judgment without guidance, unfairness, dismissiveness, and public shaming.
  3. The harms compound — undermined confidence, suppressed learning and innovation, disengagement, and turnover. People stop taking risks when the cost of a miss is humiliation.
  4. Leaders need operating rules — set clear guardrails for feedback; focus criticism on specific behaviours plus clear next steps; pause before delivering; repair trust quickly when harm occurs; and guard against bias disguised as criticism.

None of that is a soft skills workshop alone. It needs a living feedback surface — continuous notes, structured 1:1s, and a record that coaches behaviour instead of roasting character once a year.

The hurdle: feedback that humiliates or never sticks

For mid-market People leaders, HBR’s line exposes two failure modes that often coexist. Feedback is either episodic and punitive, or so vague it vanishes after the meeting. Neither builds a learning organisation.

  • Annual dump — twelve months of unspoken issues compressed into one review, so “candour” arrives as an ambush.
  • Public roasting — criticism in stand-ups, Slack, or calibration rooms that shames instead of coaching.
  • Character attacks — “you’re not strategic” or “you’re difficult” with no behavioural evidence or next step.
  • No living record — feedback lives in memory or a forgotten form, so 1:1s restart from zero and bias fills the gaps.

HBR’s guardrails only work if managers have a place to give behavioural feedback continuously — and a 1:1 rhythm that turns it into coaching, not theatre.

Book a 15-min discovery call Get the free PDF

Includes a price for your headcount + 3 fixes for your review cycle.

How Revolut People Feedback / 1:1s operationalises this

The product surface for this hurdle is continuous Feedback plus structured 1:1s — the same continuous feedback / 1:1s capability on the blog roadmap.

  • Behavioural, timely feedback — managers and peers capture specific observations when they happen, so criticism targets actions and outcomes, not character labels invented in December.
  • Structured 1:1s — a recurring coaching slot with an agenda and history, so “pause before delivering” and “clear next steps” become habit, not hope.
  • A living record — feedback and 1:1 notes accumulate against the person and their goals, which cuts public shaming (private by default) and makes repair of trust visible.
  • Bias guardrails in practice — a written trail of behaviours and next steps makes it harder to dress preference or stereotype as “candour,” and gives People leaders something to audit.

That is the stack HBR’s recommendations assume. Method context lives in the playbook summary and Week 1 — Why annual reviews fail; Feedback / 1:1s is how teams coach between review cycles. Pair it with live Goals and evidence-based Reviews, and “candour” stops being a once-a-year dump.

A 30-day start

  1. Publish feedback guardrails: behaviour + impact + next step; no character labels; no public shaming — share them in the next all-hands.
  2. Turn on continuous Feedback for two critical teams and require at least one behavioural note per direct report every two weeks.
  3. Put a standing 1:1 template in place (wins, blockers, one coaching topic, next step) and ban agenda-less “how’s it going” meetings for that cohort.
  4. Review the last cycle’s written feedback for character attacks or bias-coded language; rewrite two examples as behavioural coaching for managers to model.
  5. When harm happens, repair in the next 1:1 within 48 hours — acknowledge, restate the behavioural ask, and log the repair so trust is not left to memory.

HBR’s 2026 read is not “never give hard feedback.” It is stop confusing humiliation with high standards. Mid-market stacks that still dump criticism once a year — or roast in public — will keep impairing the performance they claim to raise.