Introduction
How to tell the truth without breaking trust — and why speed is not the solution
A month ago, I watched a design review session collapse in real time before me.
Everything was 'right' on paper:
The feedback was clear
The designer was experienced
The suggestions were actionable
But... nothing moved.
The designer defended every decision.
The team leader began to lose patience.
Two people stopped participating.
We ended up with fifteen comments on Figma... and zero decisions.
The problem was not the quality of the feedback.
It was in the vessel it was placed in.
Speed does not fix broken critique.
Today, AI can review your design in seconds:
Figma AI identifies usability issues
ChatGPT suggests improvements
V0 generates alternatives faster than you can explain the problem.
Theoretically, this should make critique better.
Practically?
It revealed a bigger gap.
The gap between comments that are made and comments that actually change the work.
AI can tell you that the contrast is weak.
But it doesn't know:
Why the designer defends a decision they know deep down is weak.
When a person needs support before honesty.
And when they need direct honesty without softening.
Speed improves quantity.
Good critique requires judgment in timing and delivery.
The real problem: teams fix the wrong thing.
Most teams try to give feedback faster.
What you really need is to strengthen the framework in which criticism is received.
Mark Heiman, who has led design teams at IDEO, Facebook, and startups for over 25 years, summarised what makes criticism work in 10 principles.
These are not tips.
They are principles that hold true even when tools change and humans remain the same.
1. One team, one goal
Criticism is not a court.
It is a joint effort to improve the work.
The enemy is not the designer.
The enemy is:
Unclear thinking
Weak craftsmanship
Lazy solutions
When criticism turns into conflict, the culture is broken.
Strong teams always reset the frame:
Before touching the design, they restate the goal... out loud.
2. Psychological safety is the container
No one dares to be honest when they feel exposed.
Without psychological safety:
People play roles
Defensive
Or withdraw
Google's research (Project Aristotle) proved that psychological safety is the most important factor for team effectiveness — more than talent or resources.
Take care of the room before you criticise the work.
3. Honesty without harm
Truth matters.
And the way it is said matters too.
You can be direct without being destructive.
Precision is more important than harshness.
"This doesn't work" ≠
"The visual hierarchy is unclear, and the user won't know where to look first"
The first is a judgement.
The second is a direction.
4. Critique the impact, not the person
Design is not identity.
And the draft is not the designer.
Talk about:
Flows
Decisions
Clarity
Not about intelligence, understanding, or taste.
Once criticism touches identity, growth stops.
5. Don't break the good while fixing the bad.
Good criticism doesn't just focus on what's broken.
Naming what works is not flattery, but precision.
It tells the designer what should be protected during improvement.
Without it, doubt will return to everything.
6. Taste without reason = noise.
"I don't like it" is not a design note.
Real criticism is related to:
The user
The goal
The constraints
The system
If you can't explain "why", it's a personal preference.
Preference is acceptable in exploration.
Destructive in the improvement phase.
Also read:If your design needs explaining... then the design itself is the problem.
7. Turn impressions into direction.
"This is confusing" is a feeling, not direction.
Good criticism does the extra work:
What needs to change?
Where?
And why?
Direction drives the work.
Impressions do not.
8. Context before solutions.
Without understanding the intent, you are solving the wrong problem.
Ask first:
What is the goal?
What is constant?
What is fragile?
What are the invisible constraints?
Criticism without context is just a review.
9. Not all notes are equal.
Mixing all the notes at one level creates emotional chaos.
Clearly distinguish between:
Actual error
Evidence from tests
Personal preference
Artificial intelligence does not weigh importance.
You do.
10. Many voices, one owner
Democracy in inputs does not mean democracy in decisions.
Without a clear owner:
Everyone speaks
And no one decides
Good criticism requires decision engineering, not consensus.
What can artificial intelligence not do?
No matter how much it improves, it cannot:
Build psychological safety
Read defensiveness or openness
Know when a person needs encouragement or firmness
Identify what needs protection
Weigh notes according to their source and context
Value is not in the observation.
But in its delivery.
When criticism becomes infrastructure
Weak teams treat criticism as a meeting.
Strong teams treat it as a system.
A system that includes:
Psychological safety built over time
Clear ownership of the decision
A shared understanding of what 'good work' means
Artificial intelligence will continue to uncover problems faster.
The real advantage?
Those who create the conditions that make problem detection actually improve work.
And this... is not a tool problem.
It's a trust issue.
And trust does not have an API.
With Echo Media
At Echo Media – the echo of media for digital marketing,
We help teams and designers build feedback systems that do not consume energy, but rather turn observations into decisions and real improvements.
If your team is full of observations and short on decisions —
perhaps the problem is not in the design, but in the way it is discussed.