Product Craft Has Always Mattered. Now You Have No Excuse.
Product Craft Has Always Mattered. Now You Have No Excuse.
Great products have always been obsessed over. Apple didn’t ship the iPod with visible screws and call it good enough. The obsession was always the point. What changed is the time between “this isn’t right” and “now it’s right” has collapsed. The iteration that used to cost a sprint costs an afternoon. The one that cost an afternoon costs twenty minutes.
I built Homeboard, a Kanban board for my family. For the first time in years of product work, I got to stop asking “is this worth the time?” and just ask “is this actually good?”
The decisions were real
No login. Not because auth is hard, but because a household doesn’t need a retention loop. The board is for people who already trust each other. The interface should say that without saying it.
One board, shared view, no swim lanes. Personal lanes feel organized. They also feel like a chore chart. The whole design premise is collective ownership. Forcing structure onto that would have undermined everything.
No due dates. No assignees. No priority flags. Every one of those features has a legitimate use case somewhere. None of them belong here. Home should feel different from work. The tool should make that argument quietly, through what it doesn’t have.
The craft was in the details nobody asked for
The product decisions that mattered most weren’t the feature list. They were the interaction quality calls, the things that don’t show up in a spec but break perceived quality the moment someone uses the thing.
Drag felt jarring. Not broken, just jarring. Fixed it. Card refreshes were abrupt. Menus were hiding behind cards on mobile. The clock was ticking visibly and it was distracting. The filter icon looked odd. None of these were feature requests. All of them were wrong.
I pushed for TV-at-distance readability because the board lives on a screen across the room. I pushed for long-press on mobile and hover affordances on desktop because platform-appropriate behavior is the product respecting the context it lives in. I pulled the sidebar. Pulled the seconds display. Pulled simple mode. Every removal made the thing feel more like itself.
The chips got a specific border radius. The font hierarchy got explicit roles. Spacing targets were concrete numbers. Vague art direction is not art direction.
What actually changed
The thing AI-assisted building changed is the penalty for iteration. When the card color scheme was wrong, I changed it at no real cost. When something was technically present but visually off, and “works” is genuinely different from “right”, I fixed it without a tradeoff conversation.
That’s the restoration of craft. Not speed itself, but what speed removes: the budget for settling. There’s no longer a credible case for leaving something mediocre because fixing it feels expensive. The cost collapsed. The excuse went with it.
The obsession was always available. The question was always whether you’d pay for it. Now the price is low enough that the only remaining explanation for a mediocre product is a mediocre set of decisions.