Every product makes a promise in its first two seconds. A banking app that opens on a list of account numbers promises bookkeeping. One that opens on what you can afford this week promises help. The features behind those screens can be identical. The products are not.

I learned this the slow way on the Halvard project. We had spent five weeks improving transfers, cards and savings. Each flow tested well on its own. Then we put them behind the old home screen and nothing changed. People still opened the app once a month, checked a number and left.

A home screen is an argument

The first screen argues for a way of using the product. If it lists everything, it argues that everything is equally important, which nobody believes. If it shows one clear next step, it argues that the product knows why you came.

If the first screen lists everything, it argues that everything matters equally. Nobody believes that.

So we stopped asking what should be on the home screen and asked what the customer was deciding when they opened the app. For Halvard the answers were short: can I pay this, and am I on track. The new screen answers both before a single tap.

How to test the promise

You do not need a finished product to test a first screen. You need one honest screen and five people. I use three checks.

• Show the screen for five seconds, hide it, and ask what the product is for.

• Ask what they would tap first, and why.

• Ask what they expected to see that was missing.

An early concept board for the Halvard home screen. The final version kept one idea from it.

If the answers to the first question vary, the promise is unclear. If the first tap is not the thing the business needs, the promise is wrong. Both are cheap to fix at this stage and very expensive later.

Halvard’s weekly active use rose 31% after launch. I would like to credit the forty screens behind the first one. The data says most of the change came from the first.