Fit, then consistency

“Consistency” might be the most trusted word in design. It sounds like care, like craft, like someone has thought things through. Nobody wants a mess.

I’ve come to think the word is more often a trap than a gift. Not because consistency is bad, but because it lets us stop thinking. Once two things are supposed to match, we check whether they do and stop asking what the problem actually needs. We get the order wrong: consistency should follow fit, not come before it.

By fit I mean how well a design meets what people need from it. When a design fits, you barely notice it. When it doesn’t, you feel it: you hesitate, work around it, or give up on what you came to do. Design is about finding those moments and fixing them one by one, until it fits.

The trap

It’s easy to see a pattern work in one place and reuse it everywhere, without asking whether the new place has the same problem. At that point we’re designing for a tidy system, not for the people using it.

In 2025, Apple took the glass look it designed for its headset, called it Liquid Glass, and extended it across every platform. On a phone, where you mostly need to read things, many people found text and controls harder to see. A few months later, Apple added a setting to make it more opaque.

Consistency is easy to defend. It comes with arguments that sound smart in any meeting: one system, less work, faster development. Sometimes that trade-off is worth it. But none of it helps the person using the product if the result doesn’t fit them. The problem starts when “keep it consistent” becomes the answer to every question.

The gift

Consistency isn’t the opposite of fit. It’s what fit looks like when the same need keeps coming up.

Pull-to-refresh is a good example. In 2009, Loren Brichter built it into Tweetie 2, a Twitter app for the iPhone, so refreshing could be part of scrolling instead of yet another button. Nobody required it. It fit how people read a feed, so other apps picked it up, and in 2012 Apple made it a standard part of iOS. The fit came first. The consistency followed.

In practice

Start by understanding the problem in front of you and what it needs. Then look at the patterns you already have. Does one of them fit, not just on this screen, but for the person moving through the whole product? If it does, use it. If it doesn’t, design what fits.

When you’re unsure, ask: if I made this one different, what would actually go wrong? If the only answer is “it would be inconsistent,” you’re protecting the rule, not the person using the product.

Chase fit, and consistency shows up on its own wherever the same answer keeps working. Chase consistency first, and you can end up with a product where everything matches and nothing quite fits.