Organizational Drag: Measuring the Invisible Tax on Every Decision
Seven dimensions, one board-legible index. Why 70–85% of enterprise AI initiatives stall for organisational reasons, not technical ones.
The model works. The pilot demonstrated value. Eighteen months later it is still a pilot. This is not a technology failure — it is drag, and drag is measurable.
Organizational drag is the accumulated friction between a decision being obvious and that decision being executed. It shows up as approval chains, unclear ownership, tribal knowledge locked in individuals, and incentive systems that quietly reward the status quo.
The seven dimensions
- Decision velocity — how long from insight to authorised action
- AI citizenship — whether accountability for AI outcomes is actually assigned
- Edge autonomy — guardrails versus gates at the operating edge
- Knowledge access — is institutional knowledge retrievable or tribal
- Incentive alignment — what behaviour the compensation system funds
- Talent fluidity — how fast people move to where value is
- Narrative coherence — whether the organisation tells itself one story
Build at the edge
Incumbents cannot rebuild the core while running it — the existing organisation's antibodies reject deep change. The alternative that survives contact with reality is to build a protected, AI-native parallel unit adjacent to the core, benchmark it honestly against the mothership, and merge it back only once it wins on the numbers.