One surface, no hidden settings

The fastest way to make a tool feel complicated is to give it one more place to look. Not one more feature. One more place. A setting that could live in a single obvious spot, scattered instead across the form where you create the thing, a settings page, the thing itself, and a detail panel two clicks deep. Each copy seemed reasonable when I added it. Together they turned a simple control into a small scavenger hunt, and the screen that held them started to feel busy in a way I couldn’t quite point at. That is the moment worth naming, because the instinct that produced it is a good instinct pointed slightly wrong.

Here is the thing I had wrong. I thought a control that mattered should appear everywhere it might plausibly be wanted. The opposite is true. A control that matters should appear in exactly one place the person already expects to find it, and nowhere else.

The instinct that scatters

When you build something and believe in it, you want it to be visible. So a reminder option goes on the new-deadline form, because you set things up when you create them. Then it goes on a settings page, because settings live in settings. Then it goes on the deadline card itself, because surely you’d want to change it there. Then it shows up again in the detail view, for completeness. Four placements, and each was defensible in isolation. You can argue for any one of them with a straight face.

What you can’t argue for is all four at once. The reminder isn’t more available now. It’s more diffuse. A person who wants to change it has to remember which of the four surfaces holds the version that sticks, and whether the form’s copy and the card’s copy mean the same thing, and which one wins if they disagree. The feature got louder and the product got harder. That trade is almost always bad, and it’s easy to make because every individual step toward it feels like generosity.

There’s a quieter cost too. Every place a setting lives is a place it can drift out of sync, a place the wording can diverge, a place a future change has to be made four times instead of once. The scatter doesn’t just tax the person using the tool. It taxes the person maintaining it, which means it taxes the person using it again later, when the four copies have quietly grown four slightly different behaviors.

Two failures that look like opposites

Here’s the part that took me a while to see clearly. There are two ways to fail at simplicity, and they look like opposites, so fixing one tends to push you toward the other.

The first failure is the busy screen. Every control on display, nothing hidden, the interface a wall of switches. This reads as complicated, and the usual reaction is to start removing things from view.

So you overcorrect into the second failure: the empty screen that hides everything. Now the surface is clean, and the person can’t do anything without going to find where you put the controls. This reads as simple in a screenshot and feels terrible in the hand, because the work didn’t go away. It just moved somewhere the person has to go looking for it. An empty screen that hides the controls is not simpler than a busy one. It has only relocated the difficulty from your screenshot to the user’s afternoon.

A busy screen full of controls is not the opposite of simple. An empty screen that buries them is not simple either. Both make the person work. The actual opposite of complicated is not “fewer controls” and it is not “hidden controls.” It is something quieter. Everything you need is where you would reach for it.

That sentence is the whole discipline. It refuses both failures at once. It doesn’t tell you to strip the interface or to bury it. It tells you to put each control where the hand already goes.

Where the hand already goes

So where does the hand go? The honest answer is that it depends on what the person is doing, and there are only a few patterns.

You set things up where you create them. The moment you add a deadline is the moment you have the context to decide how it should behave, so the relevant choices belong on the form, with sensible defaults so you can ignore them and still get something reasonable. You change things where you act on them. If you’re looking at a deadline and want to adjust how it nudges you, the place to do that is on the deadline, right where your attention already is, not on a separate page you have to navigate to and match back up by memory. And the rest, the genuine account-level settings that aren’t tied to any single item, live in one obvious place that’s called what it is. Nothing important behind a secret menu. Nothing that only reveals itself if you happen to right-click or hover the correct pixel.

That was the resolution to the scattered reminder. Set the cadence on the form when you make the deadline. Change it on the deadline when you’re looking at one. Keep the account-wide default in settings, once. The same control, consolidated to where each version of the decision actually gets made, and removed from the two places that were only there because I wanted the feature to feel important. The feature lost nothing. The screen got quiet.

The test I use now is small and it works. For any control, ask where a person would instinctively reach for it, with no training and no tooltip. Put it there. If the honest answer is “in more than one place,” that usually means there are genuinely two different moments, setup and adjustment, and the control belongs in both because the hand goes to both. If the answer is “I’m adding it here so people notice it exists,” that’s not placement. That’s advertising, and it belongs in the empty space between the two failures, where things go to make a product feel busy.

A simple tool is not one that does little. It’s one that never makes you hunt.

Common questions

Isn’t simplicity just having fewer features or fewer buttons? No, and that’s the common trap. A tool can be simple and capable at the same time. The complexity people actually feel comes less from how many things a tool can do and more from how hard it is to find the thing they want right now. A screen with several controls that are each exactly where you’d expect can feel far simpler than a near-empty screen that makes you go looking. Simplicity is about findability, not subtraction.

Where should a setting live? Where the person is when they make that decision. Setup choices belong on the screen where you create the item, with good defaults so they’re skippable. Adjustments belong on the item itself, where your attention already is when you want to change it. Account-wide preferences belong in one clearly named settings area. The failure mode is putting the same setting in all of those places at once, which feels generous and reads as clutter.

Doesn’t putting controls on the main surface make it cluttered? It can, if you put everything there indiscriminately. The discipline isn’t “show everything” or “hide everything,” it’s “show each thing where the hand reaches for it.” A control the person needs while looking at an item belongs on that item. A control they almost never touch belongs in settings. Clutter comes from ignoring that distinction, not from having controls on the surface.

What’s the harm in offering a setting in more than one place, just in case? It costs more than it looks. Every duplicate placement is somewhere the wording can drift, the behavior can diverge, and a future change has to be repeated. For the person using the tool, it means remembering which copy is the one that sticks. “Just in case” tends to multiply the surfaces you have to keep honest, and a setting that lives in four places is a setting that can quietly start meaning four things.