Dossier’sCase studies

Cargolyne moves freight across eleven states with four hundred drivers and a dispatch team of nine. Their software worked; using it did not.

The brief was to redesign the driver app. The work turned out to be deciding what the app was for.

The product had four years of feature requests behind it and no subtractions:

• A dashboard of thirty-four widgets, none of them optional
• Two taps of navigation before the next job appeared
• Delay reporting four screens deep, behind a free-text field
• A 4.1 second median load on the handsets drivers actually carry
• No offline state, in a job that is mostly out of coverage

Three things had to be true before we drew anything:

• A driver accepts or declines from a cold start in one tap
• A delay is reported in under ten seconds, one-handed, on a bad connection
• A dispatcher sees what needs attention without reading the whole board

Two days in a depot, watching. Then two weeks of prototypes tested against the same drivers, in the same cabs, at the same hour of the morning.

Nothing survived the first session intact, which is the cheapest possible time to find that out.

What shipped:

• The app opens on the next job; accept and decline are the two largest targets on screen
• Delay reporting is one tap and a reason, with the reasons taken from what drivers had been typing
• The dispatcher board keeps its density and gains a single needs-attention column
• Everything reads from a local cache first, so the app is useful with no signal

Support calls fell 38% in the first month. Median load dropped from 4.1 to 1.7 seconds, most of it by deleting things rather than optimizing them.

The dashboard was never the problem. The problem was that the product had been designed for the person who signs the contract, and every decision after that inherited it.