← All ideas
Curious

Creative Abrasion

Research and book from HBS professor Linda Hill · Collective Genius (2014)

Confidence: Medium

Creative abrasion is the deliberate practice of generating breakthroughs through rigorous debate and the collision of diverse ideas, rather than smoothing over disagreements to preserve harmony. It requires a culture where conflict is productive and trusted relationships allow for respectful challenge.

Core Concepts

The Problem

Organizations often default to politeness and consensus, which stifles the conflict needed for novel ideas. Avoiding disagreement leads to incrementalism and groupthink.

The Claim

Innovation emerges when leaders foster environments where people can disagree passionately but constructively, integrating conflicting perspectives instead of compromising.

Key Evidence

  • Case studies from Pixar, Google, and other innovative firms that institutionalize structured debate in creative processes.
  • HBS research on collective genius showing that the best innovations arise from groups that embrace cognitive diversity and friction.

Practical Implication

Leaders should design meeting practices, feedback rituals, and incentive systems that reward dissent and the synthesis of clashing viewpoints. Psychological safety is a prerequisite so that dissent is heard rather than punished.

Nuance & Limits

Creative abrasion can easily slide into personal attacks or unproductive conflict if facilitators do not actively manage power dynamics and emotional safety. The skill lies in keeping debates focused on ideas rather than egos.

Source Material

Collective Genius Linda Hill, Greg Brandeau, Emily Truelove, Kent Lineback (2014)

Citation Density

Moderate — widely referenced in HBR articles and innovation literature, but still primarily attributed to Hill's work.

Gaps

  • More systematic evidence on the conditions under which creative abrasion succeeds versus fails across different industries and cultural contexts.
  • Measurement of the specific practices that convert friction into novel output, and how to train managers in them.

Discuss Further

Open this concept in an AI assistant for deeper discussion, critique, or exploration.

Was this useful?