The short
- Observation: Many failures occur without dramatic triggers.
- Cause: Systems optimised for efficiency lose slack.
- Pattern: Stability hides accumulating fragility.
- Risk: Small disturbances cascade into breakdowns.
- Lesson: Resilience requires room to bend.
Failure without a villain
When systems fail, we look for causes: bad actors, poor decisions, unexpected shocks.
Increasingly, none exist.
Supply chains stall without disasters. Platforms go offline without attacks. Organisations freeze despite competent leadership.
The failure is structural, not event-driven.
How efficiency quietly erodes resilience
Efficiency removes excess. Redundancy looks wasteful. Buffers appear lazy. Slack feels irresponsible.
Over time, systems become precise but brittle — calibrated perfectly for yesterday’s conditions.
They function beautifully, until conditions shift by a few degrees.
The illusion of stability
Long periods without disruption create confidence. Metrics stay green. Costs fall. Performance looks strong.
But stability can be deceptive. It allows hidden dependencies to accumulate.
When stress finally appears, there is no margin left to absorb it.
Where the resilience gap shows up
- Lean supply chains with single-source dependencies.
- Teams staffed exactly to workload, with no recovery capacity.
- Digital systems running near maximum utilisation.
- Decision processes with no time for reflection.
Each looks efficient. None tolerate surprise.
Why small shocks cause large failures
Resilient systems fail gradually. Fragile systems fail suddenly.
Without buffers, even minor disturbances propagate — a delayed shipment becomes a missed quarter, a staffing gap becomes operational paralysis.
The system has no elasticity.
Designing for resilience again
Resilience is not about predicting every shock. It is about creating capacity to respond.
- Redundancy as insurance, not inefficiency.
- Slack as recovery time.
- Margins as protection.
The cost is visible. The benefit appears only when it is needed.
The takeaway
Efficiency optimises for the present. Resilience prepares for variation.
Systems do not fail because nothing went wrong — they fail because nothing was allowed to go wrong.