Back
NoteJun ’242 min read

Redesigning a tool people know by heart

Figma showed UI3, its third interface redesign, at Config this week. In a tool people use all day, muscle memory is the user’s investment, and a redesign spends it.

Typeface

At Config this week Figma showed UI3, its third interface redesign, with more of the screen given to the canvas. It’s rolling out slowly in a limited beta, and for now people can switch it off. Figma’s plan is to move everyone eventually.

Dylan Field told Creative Bloq, “I can’t tell you how scary it is designing a new UI for Figma for designers.” He’s right to be scared, and pickiness has little to do with it.

In a tool people use all day, muscle memory is the user’s investment, and a redesign spends it on their behalf. A designer who has used Figma for five years doesn’t look for the layers panel. Their hand goes there. Move it and every one of those reaches misses, hundreds of times a day, until the new place is learned.

A consumer app can usually get away with a redesign, because most people use it for a few minutes and relearn it without noticing. A professional tool is closer to a musical instrument. Its users pay for the change on every working day until they’ve relearned it.

When I redesigned Uber’s experimentation platform, I threw out the two lists engineers had been using and made each flag its own page. I’d make that call again: the old screens couldn’t say whether a launch was hurting anyone, and the new ones could. It still asked every engineer who knew the old tool to unlearn it.

What I learned is that the case for a redesign has to be made in time saved per task for the people who already know the tool. “It’s cleaner” doesn’t count. If an expert is slower in week three, the redesign hasn’t paid for itself yet, however it looks.

Figma is doing the most important thing already: letting people opt out while it learns. A few things I’d add:

  • Keep every keyboard shortcut exactly where it was, so experts have one thing that didn’t move.
  • Measure how long experts take on common tasks, not only how new users rate the look.
  • Say when the opt-out ends, well before it does.

Keeping two interfaces costs Figma real engineering time, and it can’t last forever. It’s still cheaper than making everyone who uses Figma relearn it on the same morning.