The most expensive part of almost every design project isn't the design — it's the revision cycles. And the most common source of revision cycles isn't bad design. It's feedback that describes a reaction without explaining the root cause, which leads to changes that address the symptom but not the problem.
Good feedback is a skill, and it's one that most people in decision-making roles were never taught. These four rules have eliminated more revision cycles in our projects than any other process change.
Rule 1: Describe what isn't working, not what to change
The most common feedback mistake is suggesting a solution instead of naming the problem. 'Make the headline bigger' is a solution. 'The headline isn't grabbing attention quickly enough' is a problem. When feedback prescribes solutions, the designer either implements a solution they think is wrong or has to push back on the feedback itself rather than on the problem.
You hired the designer to solve problems. Your job in the review is to describe the problems clearly. Their job is to find the right solution.
Rule 2: Anchor to the audience, not yourself
Every piece of design feedback should answer the question: 'will this work for our specific target customer?' Not 'do I like this?' The most useful reformulation is replacing 'I don't like X' with 'our customer would struggle with X because they...' This forces the feedback-giver to think about the audience rather than their personal preference — and it makes the feedback immediately more actionable.
- Personal: 'This feels too dark'
- Audience-anchored: 'Our enterprise buyers tend to distrust very dark, aggressive aesthetics — it might trigger associations with legacy software'
- Personal: 'I'd make the CTA more prominent'
- Audience-anchored: 'A cold visitor to this page may not have enough context to click a CTA this early — can we test a softer entry point?'
Rule 3: Separate must-fix from nice-to-have
If everything in your feedback is equally important, nothing is. Before sending feedback, go through each note and label it: 'must fix before this can go live' or 'would be better but won't block launch.' Most feedback lists are 20% must-fix and 80% preference. When everything is combined, designers treat every item as equally important and the project loses momentum.
Rule 4: Give feedback once
Consolidate feedback before sending it. Multiple rounds of asynchronous notes — one from the CEO, one from marketing, one from the account lead — are more damaging than one comprehensive round. Each fragmented feedback session introduces contradictions, forces the designer to reconcile conflicting direction, and adds calendar overhead to every iteration.
The standard we hold ourselves to: one consolidated feedback document per round, reviewed by all stakeholders before it goes to the design team. The added 24 hours of consolidation time saves a week of revision cycles on every major deliverable.