Building with friction | Adam Hutchinson

Jul 10, 2026<br>Building with friction

I spend an unreasonable amount of time removing friction from software.

Then I spend my weekends carrying heavy packs up mountains.

Good tools are easy to use. Good lives aren’t necessarily easy to build.

A builder absorbs difficult work once, so thousands of users can avoid it. People train for hours so the rest of life becomes easier. The payoff beats the cost.

Meaningful resistance leads to growth, while meaningless resistance gets in your way for no reason.

The Jobsian gold standard of a great product is stripping it back to its core without losing its power. If you have a choice between a convoluted flow and a single button: pick the button. It’s better in its simplicity. Don’t make the user think.

The button example may be contrived, but the same kind of thinking goes into every feature added to an app like Linear. Instead of the user paying in friction, we absorb it early in the build process. We wrestle with ambiguity, edge cases, and difficult decisions, so the finished interaction feels obvious. Hard work turns into clarity for the user and better judgement for the builder. The next thing we ship is a little better by default. Experience turns to instinct. Shared hardship raises the team’s bar.

Seek the friction that leaves you more capable. You’re not going to get stronger sitting around doing nothing. You need to put in sweaty work — legs aching, a little voice telling you to stop, yet you push through. Next time, the same weight feels lighter.

What friction is telling you

Friction isn’t automatically good. Left in the wrong place, it kills products.

It’s also a gift. The product is telling you plainly that something is wrong.

Friction can live in the words, the interaction, the workflow, or the idea itself. It can begin before someone even opens the product, when marketing promises one thing and the experience delivers another.

A confusing interaction may indicate poor copy.

Words mean things.

Copy should be treated with as much care as the rest of the product — it’s part of the product.

A cumbersome workflow may require a shortcut, or a UX pass — turning three screens into one, or one into three. Less friction doesn’t always mean fewer steps. Sometimes one overloaded screen adds more friction than five clear ones.

You don’t always need a hundred metrics to figure this out. Using the product and applying judgement goes a long way. Notice where things feel a bit wrong. A few small papercuts may seem harmless in isolation, but together they become the product.

Talking to customers fills gaps in your perspective and helps avoid tunnel vision.

Sometimes friction is telling you there’s a simpler way. Or that the feature shouldn’t exist at all. It was never for this world. Sometimes a solution flies straight over the problem’s head.

Maybe two features should’ve been one. Merge them. Or perhaps the feature adds nothing. It was a mistake. Let it go. The product will be better for it — and you’ll have less to maintain.

Attachment is the fundamental cause of product suffering.

Not all product friction is bad

A confirmation dialog before deleting something important is just enough to prevent data loss. Sometimes the user needs a beat to reconsider an action’s consequences.

Most apps, unfortunately, can’t be reduced to a single button. Showing the user every possible option at once might seem low friction, but overloading the user is high friction. The calmest interfaces are deliberate about what they show the user. Frequently used options front and center, the rest further away. The friction in finding them is lower than the friction of always showing them.

It should still be fast and easy to get to the goal, once the user knows what they want. Keyboard shortcuts keep things fast for power users without crowding the interface for everyone else.

The trick is to pay attention to friction. Place it correctly and remove it where it doesn’t serve.

Some products even manufacture it through fake loading states designed to make users feel that the system is working hard on their behalf. It’s pretty human to assume hard work takes time, even when machines can run trillions of operations per second.

Personally, I prefer honesty and speed.

More notes... Get future posts by email:<br>Email address Subscribe<br>Share<br>Copy link<br>Copied X LinkedIn Hacker News Reddit