Understanding OKR and Its Role in Managing Uncertainty Compared to KPI

Many organizations are trying OKR, but without understanding its background and its relation to uncertainty, using deterministic thinking will corrupt OKR.

A new book "This is OKR: The Work Method Behind Google and Amazon's Explosive Growth" appears on airport bestseller shelves, reflecting a trend as organizations face agile transformation and performance evaluation challenges.

The author, not a HR specialist, views performance evaluation as a transitional form and sees OKR as a small step forward that has already yielded good results in some tech organizations.

OKR vs KPI – KPI is a deterministic, predictable metric essential for business survival, while OKR’s Objective is inspirational and encourages teams to reach for ambitious, collective goals; Key Results are quantifiable outcomes that support the Objective but are not direct performance scores.

Understanding OKR’s background is crucial; at Google, KPI fails because the business is highly innovative and uncertain, whereas OKR thrives by allowing flexibility and tolerating uncertainty.

OKR does add management cost, but this cost is intended to manage uncertainty rather than enforce predictability.

OKR and Uncertainty Management – The separation of Objective (O) and Key Results (KR) enables handling uncertainty; each KR is a factor toward O, not a sum that guarantees O’s achievement.

It is impossible to enumerate all possible factors; individuals may see different paths to the goal.

The Objective itself may need adjustment during KR implementation.

Example Objective: “Make the post‑90s scream for product X.” Sample KRs: KR1 – NPS among post‑90s reaches 95%; KR2 – market share among post‑90s reaches 50%; KR3 – TikTok videos about product X receive over 100,000 shares.

All KRs share a quantitative nature, reflecting the principle “no measurement, no management.” However, achieving KR1 alone may not fully satisfy the Objective, prompting a revision of O.

OKR Layering Relevance – In large organizations, higher‑level KRs are often “inherited” as objectives for sub‑teams, which can re‑introduce determinism and undermine the original purpose of managing uncertainty.

When a market team inherits KR1 (NPS 95%), they might set their own Objective like “Become a top‑city post‑90s influencer product,” with a KR such as “Increase NPS by 10 points in first‑tier cities.” The relationship between KR and O is logical, not causal; failure of KR1 does not automatically mean O fails.

Blind inheritance can lead teams to accept suboptimal objectives, contradicting the intent of OKR.

OKR and KPI Combination – Both are numeric performance‑management systems, but they suit different contexts: KPI for predictable, certain environments; OKR for uncertain, innovative settings. They are expected to coexist for the foreseeable future.

Guidelines for adopting OKR while transitioning from KPI:

Clarify the purpose of OKR and KPI; many organizations lack transparency in KPI creation.

Avoid a one‑size‑fits‑all switch; mature organizations retain areas of certainty.

Do not try to “save” the discussion process; dialogue is core to OKR.

When mixing OKR and KPI on the same product line, be cautious of assumed certainty that may not exist.

Implement OKR with honest, periodic reviews and be willing to adjust objectives in uncertain environments.

Finally, the author encourages readers to think beyond OKR, aspiring to visionary, innovation‑driven outcomes.

Recruitment Notice : LeansoftX.com is hiring technical talent, emphasizing deep DevOps and agile knowledge; the interview process involves over 20 hours of technical discussion with senior engineers, and candidates are invited to submit resumes to [email protected].