Skip to content

70-20-10-value-split.md

Latest commit

 

History

History
29 lines (20 loc) · 1.75 KB

70-20-10-value-split.md

File metadata and controls

29 lines (20 loc) · 1.75 KB
Raw
title 70/20/10 AI Value Split
source-id bcg-bffxai-2026
wiki-page (02_References/enterprise-ai/wiki/skills/ — synthesis from BCG BFFxAI 2026)
last-synced 2026-06-07

70/20/10 AI Value Split

What this reference contains

BCG Build for the Future × AI 2026 finding: of the total value AI creates in an enterprise, roughly 70% comes from People, 20% from Process, and 10% from Technology. The distribution is counterintuitive — the layer everyone optimizes (models, tools, infrastructure) produces the smallest value share.

Key claims (with citations)

  • 70% of AI value comes from People — skills, behaviors, leadership, culture. Source: BCG BFFxAI 2026, core finding.
  • 20% comes from Process — workflow redesign, decision architecture, measurement discipline. Source: BCG BFFxAI 2026.
  • 10% comes from Technology — model selection, infrastructure, tooling. Source: BCG BFFxAI 2026.
  • The 5% that capture nearly all enterprise AI value allocate effort accordingly; the 95% focus disproportionately on technology. Source: paired finding, MIT NANDA 2025 (see 95-5-genai-divide.md).

When to consult

  • For skill people-readiness-conversation: anchor the conversation in the 70/20/10 finding when surfacing why people gaps matter disproportionately.
  • For skill tech-stack-diagnostic: use the 10% figure to calibrate expectations — a perfect tech stack won't rescue a 70% people gap.
  • For skill general-maturity-assessment: the 70/20/10 split is the interpretive frame for why Stage 2 → Stage 3 stalls in organizations with strong tech but weak people/process investment.

Source

BCG Build for the Future × AI 2026. Core findings on value distribution across people, process, and technology layers in enterprise AI transformation.