Design Authority

Definition

A design authority reviews and approves architecturally significant changes, with the aim of ensuring consistency, quality and alignment.

When it adds value

• Genuinely significant decisions — a new data store, a new integration pattern, a change to a security boundary, a cross-team boundary.
• Cross-cutting concerns no individual team can see: duplication across the estate, cumulative complexity, a decision that constrains others.
• Bringing expertise the team does not have — regulatory, security, operational.

When it becomes harmful

• Reviewing everything, so it is a queue and coverage is shallow because reviewers cannot meaningfully examine that volume.
• Meeting fortnightly, adding a fortnight to every decision.
• Approving rather than advising, so teams optimise for getting through review rather than for good design.
• Reviewing at the end, when the work is complete and changing anything is expensive. This is the most demoralising pattern and it is entirely avoidable.

The model that works

Advisory by default, gating by exception.

• Publish what "significant" means, with concrete examples, so teams self-identify and most changes never come near it.
• Engage early, at the shape of the design rather than at its completion. A thirty-minute conversation before implementation is worth more than a review of finished work.
• Provide standards and paved roads so most decisions are already made and need no review at all. This is the highest-leverage thing the authority can do — reducing the number of decisions requiring attention.
• Reserve gating for genuinely high-risk decisions — irreversible, security-relevant, or organisation-wide.
• Turn around quickly, in days, or teams will route around it.
• Record decisions so precedent accumulates and the same question is not re-asked.

The measure of a good design authority

Not the number of reviews conducted, which measures activity. Better: the proportion of decisions that did not need review because standards and paved roads had already settled them, and the turnaround time for those that did.

Interview question

"How would you structure a design review function that improves decisions without becoming a bottleneck?"