Why annual reviews fail (and what to run instead)

Ash Taherian · Product Sales Executive, Revolut People · Week 1

Most companies still run one big review a year. Managers scramble for ratings. Ratings don’t match across teams. Promotions and exits feel political. Then everyone promises to “do better next year.”

The failure isn’t effort. It’s design — and CPOs feel it when talent density slips while the process still looks “complete” on paper.

Get the full PDF Book a meeting

What breaks

• Vague ratings — “meets expectations” means different things to different managers.
• No published talent bar — junior and senior get compared with the same fuzzy language.
• Once-a-year evidence — memory replaces scorecards; recency bias wins.
• No calibration — Team A’s “exceptional” is Team B’s “solid.”
• No consequences — grades don’t cleanly drive promo, pay, PIP, or exit.

What high-growth companies run instead

The method behind Revolut’s performance culture (detailed playbooks from QuantumLight — Ash did not author those playbooks) is simple:

1. Score three lenses — deliverables, skills, and culture. Not vibes.
2. Publish a talent bar by seniority — minimum standard for each level.
3. Sort honestly — A-players, above bar, underperformers (with deliberate caps).
4. Quarterly cycle + calibration — same scorecards, then align across teams.
5. Tie grades to outcomes — promo roadmaps, development, PIP, or pay.

Andrei Oprisor (GM, Revolut People) made the same case in his HR Chief cover feature: performance should create clarity — and winning raises morale. The process only sticks when scorecards, calibration, and outcomes live in one system instead of slides and spreadsheets.

How Revolut People helps

The primary product surface for this hurdle is Reviews.

• Standardised scorecards — managers record evidence against deliverables, skills, and culture, not free-text vibes.
• Quarterly cycle — reviews run on a rhythm short enough that underperformance doesn’t compound for a year.
• Calibration — align grades across teams, check outliers, and keep “exceptional” in a real band instead of politics.
• Outcomes — grades connect to promo roadmaps, development, PIP, or pay so the cycle isn’t theatre.

Method is the playbook (public summary). Reviews is how teams operationalise the quarterly filter Nik describes in How Revolut filters talent.

A 30-day start

1. Write the talent bar for two critical job families.
2. Switch the next review to a three-lens scorecard with evidence fields.
3. Hold one calibration meeting with a real A-player cap.
4. Announce grades with a clear next action for every person.

Related: Week 2 — Publish a talent bar by seniority · Week 3 — Goals that don’t cascade · All weeks