Silicon Valley performance reviews split on four axes —
impact, expertise, direction, and soft skills.
career-strategymeta-lifebigtech-comp-is-a-system
실리콘밸리 성과 평가는
임팩트·전문성·방향성·소프트스킬 네 축으로 갈립니다.
실리콘밸리의 가장 큰 혁신은 제품이 아니라 **성과 평가 시스템** 자체입니다. 잘 설계된 평가 시스템이 기업의 문화를 만들고, 직원들이 회사에서 나아갈 방향을 결정합니다. 그 평가는 임팩트, 전문성, 방향성, 소프트스킬 네 가지 축으로 분리됩니다.
**임팩트**는 수치로 측정 가능한 성과의 크기입니다 — 광고팀이면 매출, 인프라팀이면 비용 절감, 마케터면 프로젝트 규모. 회사의 그 해 우선순위에 따라 임팩트의 정의가 바뀝니다. 2023년 메타는 효율성을 목표로 잡았고, 그 해의 임팩트는 코스트 리덕션이 되었습니다.
**전문성**은 직군별 핵심 능력입니다. 엔지니어에게는 코딩의 깊이, 마케터에게는 고객 동원력, PM에게는 어려운 상품의 상품화 능력. 직군과 연차에 따라 기준이 명확히 제시되고, 실리콘밸리 회사들은 그 기준을 의도적으로 공개합니다.
**방향성**은 한 프로젝트를 끝낸 후 다음에 무엇을 할 것인지, 어떻게 더 큰 문제로 확장할 것인지 제시하는 능력입니다. 평가받는 사람이 더 큰 영역을 그릴 수 있는가의 척도입니다.
**소프트스킬**은 커뮤니케이션, 협업, 멘토링 — 다른 사람을 통해 일을 만드는 능력입니다.
같은 IC라도 직군과 연차에 따라 네 축의 비중이 다르게 매겨집니다. 그래서 평가가 그렇게 정교한 것이고, 그래서 시스템이 사람을 정확하게 선별합니다.Also in EnglishSilicon Valley performance reviews split on four axes —
impact, expertise, direction, and soft skills.
Silicon Valley's biggest innovation isn't a product. It's the **performance review system itself.** A well-designed eval writes a company's culture and points its people in a direction. The eval splits into four axes:
**Impact** — quantifiable scale: ad revenue for ads teams, cost reduction for infra, project scale for marketers. The definition of impact rotates with the company's yearly priority. In 2023 Meta declared an "efficiency year" and impact became cost reduction.
**Expertise** — domain depth: code quality for engineers, customer engagement for marketers, productization for PMs. Criteria are spelled out per discipline and per level, and Silicon Valley companies *intentionally* publish them.
**Direction** — what's next after this project ships? How does the result expand into a bigger problem worth pursuing? Direction-setting is its own axis, scoring whether the IC can map a larger surface.
**Soft skills** — communication, collaboration, mentorship. The work that gets done *through other people*.
The weighting differs by discipline and by level — that's why the eval is as precise as it is, and why the system selects people accurately at scale.