Should we remove performance ratings?
Question
Should we remove performance ratings?
The "kill the annual rating" wave peaked in 2015–2017 when Adobe, Microsoft, Deloitte, Accenture, and GE publicly dropped ratings. Many of them quietly brought ratings back in a different form by 2020. So the 2026 question is sharper: what role should ratings play in a continuous-feedback world, and what breaks when you remove them entirely?
This page is for HR leaders deciding whether to keep ratings, redesign them, or drop them.
Evidence
Four findings from our library that frame the decision — each tagged with where its source has a commercial stake, because a source that profits from its own conclusion deserves a second read:
Even fixing the ratings doesn't fix the bias. Harvard Kennedy School's 2025 field study at a multinational financial services firm used a natural experiment — a software glitch that hid self-ratings from managers for one review cycle. Result: hiding self-ratings reduced anchoring, but managers substituted historical ratings as the new anchor. Demographic disparities (gender, race) held. The implication is uncomfortable: rating mechanics matter less than the system around them. Source check: independent academic study — no product to sell, which is exactly why we lead with it.
Annual ratings are losing adoption, fast. 15Five's 2026 trends piece reports annual-review adoption fell from 82% (2016) to 54% (now). Only 14% of employees say performance reviews inspire improvement. The "annual cycle" itself is the practice under question, not ratings per se. Source check: 15Five sells continuous-feedback software — it profits from the exact conclusion it advocates. Calibrate.
Speed is possible — but it's downstream of infrastructure. Remote ran a global review of ≈1,800 employees in 48 hours using monthly check-ins, AI-assisted drafting, and live alignment dashboards. The 48-hour format itself isn't the lesson — the continuous-feedback foundation that made it possible is. Without that foundation, fast reviews produce worse reviews, not better ones. Source check: the experiment is Remote's own and showcases the HR infrastructure Remote sells — neutral on ratings, not neutral on infrastructure.
MENA workplaces should treat perf-mgmt as strategic, not compliance. SHRM's 2026 piece on performance management in MENA argues that under Vision 2030 / Vision 2040 talent-localization mandates, performance management is a strategic tool — if it shifts from evaluative to developmental. 91% of workers who feel their organization addresses their needs report job satisfaction; ratings that don't connect to development aren't doing this work. Source check: SHRM is a paid-membership professional body — no direct stake in the rating verdict.
Disagreement
The four sources don't actually agree on the verdict.
15Five treats annual ratings as a relic — continuous feedback wins. Their commercial interest is in continuous-feedback software, so calibrate.
HR Magazine (Remote) is neutral on ratings; their argument is about process speed and infrastructure. You could run their 48-hour cycle with or without a numeric rating output.
SHRM (MENA) is pro-perf-management but doesn't say abandon ratings. The piece argues for reframing feedback as developmental — which is compatible with keeping ratings, dropping them, or redesigning them.
Harvard Kennedy is the most uncomfortable source: their field experiment suggests the underlying bias persists regardless of rating design. Removing ratings might solve a different problem than the one most HR teams are trying to solve.
So the live disagreement isn't "ratings yes or ratings no." It's three nested questions:
- Are ratings serving comp / promotion / layoff differentiation, or just bureaucratic ritual?
- If you remove them, what replaces the differentiation signal those decisions need?
- Will removing ratings reduce demographic bias, or will the bias survive in whatever replaces them?
Peoplense Verdict
Do: kill annual-only review cycles. They produce "feedback amnesia" (15Five), low employee buy-in (14% inspired), and disproportionate admin burden. Move to continuous feedback infrastructure — monthly check-ins, clear role expectations, live calibration. This works whether you keep ratings or not.
Don't: remove ratings entirely unless you have a replacement signal for the consequential decisions ratings currently support — compensation differentiation, promotion eligibility, PIP triggers, layoff prioritization. Removing the rating without removing the consequence creates ambiguity, not freedom.
Watch out: ratings without rigorous calibration are the worst of both worlds — you get the demoralization of being graded and the unreliability of being graded by an uncalibrated grader. If you don't have manager-of-managers calibration sessions before ratings lock, you don't have ratings; you have manager opinions formatted as ratings.
What to do Monday
Three concrete actions for HR leaders this week.
-
Audit your last review cycle for rating distribution. Pull your last cycle's ratings. Did they actually differentiate, or did 80%+ cluster at "Meets"? Rating compression is the leading indicator that calibration is broken — not a sign of a strong team. If distribution is flat, your ratings are not doing the work you think they're doing.
-
Map every consequential link that depends on a rating. Make a list: comp adjustments, bonus modifier, promotion eligibility, succession planning, layoff prioritization, PIP triggers, learning budget allocation. For each item, ask: does this decision actually need a rating, or could it use a different signal? Items where the rating is the only input are candidates to keep. Items with no consequential link are candidates to drop entirely.
-
Run one calibration session before your next ratings lock. One afternoon. Manager-of-managers plus HRBP. Review ratings by demographic. Compare against any objective outcome you have (revenue numbers, project deliveries, customer NPS). Surface the gaps. This is the cheapest, highest-leverage intervention. If you skip this, every other improvement is decorative.
GCC Relevance
The KSA / GCC workplace adds three specific pressures to this decision.
High power distance amplifies rating inflation. Hofstede framework: KSA scores high on power-distance and collectivism. Managers feel more social cost from giving lower ratings than US/UK peers. The result is "Meets" inflation that destroys the rating's signal value within two cycles. Removing ratings doesn't solve this — it relocates the problem to whatever replaces them. The intervention that helps most is calibration discipline, not rating removal.
Vision 2030 / Nitaqat increases — not decreases — the need for performance differentiation. Saudization-banded scorecards make corporate performance KPIs more consequential. Performance ratings tied to Saudization progress can support the national agenda — but only if they're rigorous. Inflated ratings on Saudi nationals undermine the very localization story the rating system is supposed to support.
Wasta dynamics make "ratings without rigor" actively dangerous. In contexts where relational politics already shape decisions, ratings without independent calibration encode those politics into the formal record. The honest move is either strong calibration (so ratings reflect performance, not relationships) or no ratings (so relational decisions stay visible as relational decisions). The dangerous middle is a ratings system that pretends to be objective while reproducing wasta in a structured format.
Practical implication for KSA HR leaders: the question for most Gulf workplaces isn't "should we abandon ratings" — it's "do we have the calibration discipline to make ratings honest, or are we using ratings as cover for decisions we'd rather not defend openly?" Answer that first.
Sources
All four corpus articles below open in our admin reader with the editorial-summary contract banner — our text summary on Peoplense, full text + author voice + figures at the original publisher.
- Performance Management Trends to Watch in 2026 — 15Five, 2026-03 (annual-review adoption decline, the 14% inspiration stat, six 2026 trends)
- The 48-hour performance review — and why it might be the future — HR Magazine, 2026-03 (Remote's experiment, the continuous-feedback infrastructure that made it work)
- From Process to Purpose: Making Performance Management Your Greatest Asset in Times of Change — SHRM, 2026-04 (MENA-specific, Vision 2030 talent-localization, developmental-not-evaluative reframe)
- Self-ratings and bias in performance reviews — Harvard Kennedy School, 2025-10 (natural-experiment field study at a multinational financial services firm)
Get the Monday Brief
Evidence-based people development research, summarized weekly. Free. No ads. Every article links to its source.
Email used only to deliver the brief. Unsubscribe anytime.
Want to join the editorial team?
We're building this slowly, with HR practitioners across the GCC. If you'd like to help shape what gets published, tell us about yourself.
Tell us about yourself