Most projects fail before anyone opens Figma

Most projects that go badly didn't go wrong in execution. They went wrong in the first conversation, and everybody found out four weeks later.

Work was competent. It just answered a question nobody had bothered to check was the right one.

Nobody actually needs a redesign

A client asks for a redesign. That's a request. The actual problem might be that traffic converts badly, or that sales burns an hour a week explaining what the company even does, or that a competitor shipped something on Sunday that made the founder anxious.

Four different projects there. A redesign solves precisely one of them. Possibly none.

Taking the request at face value feels respectful and efficient. It's neither. All it does is defer the real conversation until after the work exists, when changing direction is expensive and everyone's emotionally invested in something.

The questions worth an hour

What happens if we do nothing? If the honest answer is "not much," you've just found the actual priority, and it's somewhere else entirely.

Who decides this is finished? If that person isn't at kickoff, they'll surface in week six carrying opinions that were always going to arrive.

What does this need to do, in numbers? "Feel more premium" can't be evaluated, so it can be relitigated forever, by anyone, at any time. "Reduce drop-off on the pricing page" can be checked on a Tuesday.

What's already been tried? There's nearly always a previous attempt. Why it failed is the single most useful thing in the room and almost nobody volunteers it unprompted.

An hour of this beats a week of revisions, because revisions can't fix a misdiagnosis. They just make it better looking.

Beware the room where everyone nods

Most dangerous moment in any project? The one where everybody nods.

Clean, modern, premium, bold. Each of those means something different to every person in the room, and they all sound agreed upon. Two people nod at "premium" while picturing opposite outcomes, and neither finds out until something concrete exists to argue about.

Vague agreement is just a disagreement that hasn't happened yet. The fix isn't more meetings, it's being specific early enough that disagreeing is still cheap. References, numbers, explicit non-goals. All of it beats another round of discussion.

If it isn't written, it isn't thought

A brief isn't documentation of the thinking. It is the thinking, and it isn't done until it's written down where someone else can read it.

Writing drags vagueness into the open. Try writing "modern but timeless, bold but understated" without noticing it constrains absolutely nothing. A brief that could describe any project isn't a brief, it's a mood.

Useful version is short and specific. What problem, for whom, judged how, and what's explicitly out of scope. If it doesn't name at least one thing you're choosing not to do, it isn't finished.

Week one is the cheap week

Pressure always runs toward starting. Briefs feel like overhead, clients want to see something moving, and nobody in the history of this industry has been thanked for spending a week on a document.

But the cost of a wrong brief isn't the brief, it's everything built on top of it. That bill lands late and it's always bigger than the time you saved.