Why Performance Scores Should Target Projects, Not People: Rethinking KPI and OKR

Recently several readers shared their frustration with performance reviews that left them feeling devalued and even persecuted. The author discovered that the disappointment stemmed not only from unmet expectations but also from conversations with managers and HR that seemed to completely reject the individual.

Two core viewpoints:

Goal setting and performance should aim to improve organizational and personal efficiency, not merely to judge people.

People are complex ; they experience fluctuations and should not be easily dismissed. Performance should assess work, not the person.

The author believes scores belong to projects, products, departments, or code rather than to individuals, yet most companies still assign them to people, leading to a feeling of being “purged.”

Historical anecdotes—such as the early struggles of Su Buqing, Hua Luogeng, Zheng Yuanjie, and the aristocratic Lev Tolstoy—illustrate that early failures do not predict lifelong inability, mirroring how low performance scores should not define a career.

Modern tech giants (Adobe, Dell, Microsoft, Amazon, Deloitte, Accenture, PwC, GE, Tencent, Baidu, Xiaomi, and others) are abandoning strict KPI systems, recognizing that treating performance like school grades stifles growth and encourages short‑term, superficial tactics.

OKR (Objectives and Key Results) is presented as a better alternative, but many organizations still misuse it by treating Key Results as KPI targets, causing employees to know *what* to do without understanding *why*.

Key characteristics of effective OKR:

Proposed by employees.

Goal‑oriented.

Shared transparently across the organization.

Examples of OKR implementation:

Objective: Enhance user experience. Key Results: Reduce user steps by >20%, cut support tickets by >40%, achieve 99.9% of operations under 100 ms.

Product team objectives: streamline registration, improve red‑packet algorithm, raise product quality.

Backend team objectives: define SLA/monitoring, automate operations, boost performance to meet specific QPS latency.

These objectives cascade from top‑level leadership to front‑line staff, ensuring transparency and allowing anyone to question or improve the direction.

The author critiques value‑based assessments, warning they can become political tools that suppress diverse thought, likening them to historical “cultural revolutions.”

Personal advice: treat performance scores with a calm mindset—they do not define your life—while taking personal development seriously. Success stories like Brian Acton (WhatsApp) and Jack Ma (Alibaba) demonstrate that early setbacks can precede great achievements.

Ultimately, the author urges readers to focus on doing the right things, not on chasing KPI numbers, and to seek environments that match their strengths, rather than letting external evaluations dictate their worth.