Design Reviews
Design reviews are not approval meetings. They are quality systems for judgment, consistency, and shared product thinking.
- Audience
- Design managers, design leads, product teams
- Status
- In development
- Random stakeholder feedback
- Taste-led decisions
- Inconsistent craft
- Weak rationale
- Designers working in isolation
- 01Critique should improve judgment, not punish taste.
- 02Reviews need decision criteria.
- 03Product context must enter craft discussion.
- 04Repeated rituals build culture.
- 05The best reviews reduce future ambiguity.
Inputs
- Work framed with its problem, constraints, and intended outcome
- Shared quality standards the review measures against
Rituals
- Fixed-cadence design reviews and peer critique
- Signoff review before development handoff
Decision rules
- Feedback attaches to criteria, not to taste
- Unresolved debates leave with an owner and a deadline
- Product context can veto craft preference
Outputs
- Consistent quality across designers and product areas
- Documented rationale for significant design decisions
- Faster shared judgment over time
Ownership
- A named facilitator owns cadence and standards
- Designers own bringing reviewable framing, not just screens
Review cadence
- Weekly rituals; signoff at sprint boundaries
What must be true for this system to work.
- Reviews run against written criteria, not personal preference.
- Every review ends with clear decisions or owners.
- Critique is safe enough that juniors bring unfinished work.
- Signoff happens before the most expensive moment to fix quality.
- The ritual survives deadline pressure.
What breaks this system.
- Reviews as approval theater for stakeholders.
- Feedback by hierarchy instead of criteria.
- Skipping reviews whenever delivery gets busy.
- Critique that punishes taste instead of building judgment.
- Decisions that evaporate because nobody wrote them down.
This manual entry is intentionally in shell form. It will expand into a full operating guide with examples, checklists, and decision rules.