the self-review problem

there’s a specific kind of confidence that sets in after you’ve worked on something for a while. not arrogance — just familiarity. you know what you meant, so you see what you meant, even when it’s not there.

i notice this most when i share work i’ve been close to. i’ll review it myself three times and think it’s tight. then someone asks a question — “what’s the user supposed to do here?” or “what’s this section actually for?” — and i realise i haven’t answered it. i thought i had. from inside the work it felt answered. from outside, it’s a gap.

this isn’t a discipline problem. you can be a rigorous designer and still have blind spots in your own work. it’s structural. the more you understand what you were trying to do, the more your eye fills in what’s missing. you read your own intentions back into the design. someone coming in fresh reads only what’s actually there.

the standard fix is to step away. leave it overnight, come back with fresh eyes. this helps, but not as much as we pretend. you still have all the context. the gap is still invisible to you because you’re the one who knows where it is.

the only thing that actually works is a different person. not someone who’ll tell you it looks good or suggest a different shade of blue — someone who’ll tell you where the argument breaks down. where the design assumes something the user won’t know. where the flow makes sense to you and nowhere else.

i’ve gotten this kind of feedback rarely. most of the time i got preference notes. the structural gaps — the places where the design was wrong at the level of what it was claiming — mostly got caught by users in testing, or by a PM who asked an inconvenient question, or by noticing in production that nobody was clicking the thing i thought was obvious.

that’s a slow and expensive feedback loop. and it’s avoidable if you can get the right eyes on work before it ships.

if you’re at that point — work that feels done but you’re not sure — i offer structured design critique — analysis at the level of what the design is claiming and where it doesn’t hold up. $75.