Designing for the cold start

The first 30 seconds of an app are the only 30 seconds most users will ever see. You can design the best tenth-screen in the world and never need to — because nobody gets past the third.

I've shipped two products that died at the cold start and a few that didn't. Here's what actually separated them.

What "cold start" means

Not literally boot time. The product cold start is the stretch between "I opened this" and "I understand what it does for me."

For a CLI that's the first --help output. For a web app it's the empty dashboard. For a social app it's the first scroll, with zero content.

Every one of these is a moment where the user is free to leave. They cost you nothing to build poorly, and everything to lose.

The anti-patterns

The obvious:

• An empty dashboard with "get started" and no content to start with
• A signup flow that asks seven questions before showing value
• A tutorial that explains the interface before the point

The less obvious:

• Defaults set to "show nothing" to avoid opinionated choices
• Settings panels that default to locked — every feature behind a toggle
• Copy that describes the app instead of what the user wants

What works

Pre-populate. Seed the account with something. A sample project, a starter document, a fake teammate. Empty state is where apps die.

One default action. Not three buttons of equal weight. One "this is the thing to click." The others can be there; they just shouldn't compete.

Show the shape of the thing. A data app should show data on day zero, even fake data. A writing app should already have something written. Never open on a void.

Defer the setup. The account you set up later is twice as likely to be set up, because by then you know why.

The honest version

All of this boils down to one idea: the cold start is a writing problem, not an engineering problem. You're convincing a stranger to spend another 30 seconds, then another minute, then an hour.

Engineers tend to reach for features. Writers reach for flow. For this specific 30 seconds, be a writer.