Skip to content

Novelty Budget

Say a word twenty times in a row and it stops meaning anything. Rauno Freiberg points out that interfaces do the same thing: the flourish that delighted you on day one is invisible by day ten and in the way by day one hundred.

He borrows a rule from film, where around 10% of the frame is the accent color, and suggests making 90% of an experience familiar and 10% novel. Think of that 10% as a novelty budget. The question isn't whether to spend it, but where.

Finish a few tasks and clear a list, in both modes.

When everything animates, nothing stands out and the whole app feels slower. When only clearing the list animates, that moment lands.

Spend by frequency

The rarer the moment, the more expressive it can be. Benji Taylor calls this the Delight-Impact Curve: the potential for delight goes up as the frequency of use goes down.

  • Hovering, switching tabs, checking a box. These happen constantly. Spend nothing here.
  • Opening a dialog, sending a message. Occasional. A short, quiet transition is plenty.
  • Finishing onboarding, clearing your inbox, a yearly recap. Rare. This is where the budget goes.

Some moments should only happen once. On Rauno's Devouring Details, the welcome transition plays right after you log in and never on a reload.

Count the cost

Small delays compound. Drag the sliders.

6.1hoursa year, per person

A 300ms transition you see 200 times a day costs about six hours a year. That is a lot of waiting to spend on something you stopped noticing in week one.

Novelty taxes learning too

New interaction patterns cost more than new visuals, because people have to learn them before they can use them. Rauno calls this the novelty tax, and points to Arc: a browser its fans loved, that most people found too much to learn.

None of this means "no delight". It means a calm baseline, so the one loud moment has something to stand against.

Resources

Search concepts

Search for a command to run...