Designing with intentions, not trends

Good interfaces don’t chase. They intend. When you open a file browser you don’t see translucency because it is fashionable — you see it because materials communicate hierarchy, and hierarchy is what lets you find things without reading.

Intent versus trends

Trends are shortcuts. Intentions are systems. I rebuilt this index six times to get the sidebar spacing right. Not because twelve pixels versus sixteen is visible on its own, but because it changes how quickly your eye can scan a column of categories.

Design that pays rent is design that reduces cognitive load. Every single time.

The test I keep coming back to: if I removed this, would anyone be slower? If the answer is no, the element is decoration wearing a job title.

Calm software

Calm is not the same as minimal. Calm is predictable. The category list is always on the left. The preview is always on the right. The dock is always at the bottom. You stop looking for the interface and start looking at the work.

• Links rather than JavaScript filters, so every view has a URL
• Preview as an enhancement, never a requirement
• Keyboard navigation that matches what the metaphor promises
• Scroll containers that behave like the thing they imitate

Keep the shell consistent and let the content fill the space. That is the whole rule.

Hierarchy is a promise

Every surface in an interface makes a claim about importance, and the user learns those claims within seconds. A panel that sits above another panel is saying “I am the thing you are working on.” If that stops being true on some other screen, you have not made a visual mistake — you have broken a promise, and the user now has to read the interface rather than glance at it.

This is the argument for consistency that actually holds up. Consistency is not about looking tidy. It is about not lying to someone who has already learned your grammar.

Three surfaces is usually enough

Recessed, level, and raised. A recessed surface holds input. A level surface holds content. A raised surface holds something temporary that will go away. Almost every interface I admire uses roughly this vocabulary, and almost every interface I find exhausting uses eight surfaces with no rule about which is which.

Density is a decision, not a default

Dense interfaces are not worse. A trading terminal that wasted space would be a bad trading terminal. But density has to be chosen for a reason connected to what the person is doing, and most products inherit their density from whatever component library they installed.

The useful question is: how many of these will a person look at in one sitting? If the answer is three, give them room. If the answer is two hundred, every pixel of padding is a row they cannot see, and generosity becomes a cost you are passing to them.

Whitespace is not a style. It is a claim about how many things you expect someone to hold in their head at once.

Where taste actually lives

People talk about taste as though it were a preference for particular colours. In practice, the designers I learn most from are simply the ones who notice more. They see that the icon is a half-pixel off centre, that the transition overshoots, that two labels use different verb tenses. None of that is preference. It is attention, applied consistently, for long enough that it compounds.

Which is good news, because attention is trainable and preference is not. The way to get better is not to develop opinions about gradients. It is to look at your own work for longer than is comfortable, in the state a real user will see it, on the day you would rather ship it.

The thing I keep relearning

Every project, at some point, I add something clever. A transition that reveals structure. A layout that adapts in an interesting way. And every project, some months later, I take it out, because the cleverness was solving a problem I found interesting rather than a problem the user had.

The work that survives is nearly always the work that was slightly boring to build. That is not an argument against ambition. It is an argument for spending the ambition on the parts nobody will notice individually and everybody will feel collectively.