Design by committee, death by a thousand opinions
No single note killed the work. That's the trick of it. Each round of feedback was reasonable, and the sum was beige.
The first presentation goes well. Better than well. The concept is sharp, the room actually laughs at the right moment, and someone says the words every designer files away for the bad days: “I’ve never seen anything like this.”
Then the feedback rounds begin.
Legal would prefer softer language in the hero. The VP of sales wants the product screenshot higher, ideally above everything. Someone forwards it to a spouse, who feels the black is “a little aggressive.” Marketing asks if the headline can also mention the other three value props. Each note arrives politely. Every one is defensible on its own terms. And eleven defensible notes later, the thing nobody had ever seen anything like looks precisely like everything everyone has seen before.
No single opinion killed it. That’s the trick of it.
Committees don’t murder good work; they sand it.
A thousand small passes of fine-grit consensus, and the edges (the parts that made anyone feel anything) are gone. What ships is smooth, inoffensive, and invisible.
We’ve watched this happen enough times to say with some confidence: the problem is almost never the people. Individually they’re smart, and they genuinely want the project to succeed. The problem is that nobody structured how decisions get made, so feedback rushed in to fill the vacuum. Unstructured feedback isn’t collaboration. It’s a group of people politely fighting for the steering wheel.
There’s a second force underneath, and it’s cultural. In plenty of organizations, giving a note is how you prove you were in the meeting. A note is a fingerprint. It says I touched this, I mattered here, remember me at review time. So feedback becomes a ritual of ownership rather than an act of judgment, and the mockup collects fingerprints until nobody can see through the glass. People aren’t improving the work at that point. They’re initialing it.
Here’s the structure we push for instead. Five rules, in the order they matter.
1. Name the decider before round one
One person owns the call. Not the loudest person, not necessarily the most senior one, and never “the group.” Everyone else is a voice, and voices are valuable, but when input conflicts, one human breaks the tie and the work moves. The decider’s job isn’t having the best taste in the building, either. It’s keeping the work whole while it travels through the org. If a project can’t name its decider in one sentence, that’s not a detail to sort out later. That’s the whole storm, still offshore.
2. Give feedback as problems, not prescriptions
“Make the logo bigger” is a solution wearing a trench coat. Underneath it is a real concern, usually something like “I’m worried people won’t remember who this is from.” Say that instead. The problem statement is almost always legitimate and worth solving; the prescription is almost always the least imaginative fix available. When stakeholders bring problems and designers bring solutions, everyone’s doing the job they’re actually good at.
3. Tell each round what it’s for
Round one asks: is the concept right? Round two: is the execution right? Round three: is anything broken? The rounds narrow, always. Feedback that questions the concept during round three isn’t insight arriving late, it’s round one escaping containment, and the polite-sounding phrase “just playing devil’s advocate” is usually how it escapes. Devils, in our experience, have plenty of advocates already. Assign the round a question, and park everything that doesn’t answer it.
4. Weight opinions by proximity to the user
The intern who watched five customers stumble through the signup flow knows more about the signup flow than the boardroom does. Org charts measure accountability, not perception. When a debate stalls, the tiebreaker shouldn’t be who outranks whom; it should be who’s standing closest to the person this thing is actually for. Sometimes that’s nobody in the room, which is its own answer: go find out instead of voting.
5. Let silence expire
Open-ended feedback windows are where schedules go to die. Set a date, and make it known that silence past the date is approval, not a pause button someone can press three weeks later with “sorry, just seeing this.” Late notes don’t reopen the round. They ride the next train. This sounds harsh exactly once, and then everyone privately loves it, because it converts feedback from an ambient threat into an appointment.
None of this limits how many people can care about the work. Care away. Big feelings about the work are a sign the work matters, and the strongest projects we’ve shipped had passionate rooms behind them. Structure doesn’t reduce the number of opinions. It gives each one a door to knock on, instead of a wall to seep through.
Because here’s what the sanding-down process never admits: smooth isn’t safe. The inoffensive version offends no one and reaches no one, and the budget spent making it invisible is real money. Good work has edges. Somebody in your process has to be allowed to protect them.
Decide who that is before the first presentation. Preferably before the room laughs.
