The brief
Ashgrove moves 4,000 pallets a week across seven depots. Dispatch was coordinated by phone and a shared spreadsheet that three people edited simultaneously. Nobody trusted it, so everybody kept a private copy.
They asked for a dashboard. We asked to watch a dispatch first.
What we found
The spreadsheet was not the problem. The problem was that depot managers had no way to see a load's status without ringing the depot that held it — so they rang, and the person who answered stopped what they were doing to look.
Sixty-one calls on the Tuesday we counted. Each one interrupted somebody mid-task.
Depot 3 on a Tuesday morning. The whiteboard was the real system of record.
What we made
A single load status, visible to every depot, updated by scanning at each handover. No new hardware — the scanners they already used now post to an API instead of a local file.
The dashboard they asked for exists, but it is the least-used screen we built. The scanner flow is where the value actually landed.
“I stopped getting phone calls in the second week. That was when I knew it had worked.”
What happened
Dispatch cycle time fell from just over four days to eleven hours within six weeks of go-live. The cost saving came mostly from reduced overnight holding, not from headcount — nobody lost a job.
Two depots still keep the whiteboard. We left it alone. It is now a summary of the system rather than a competitor to it.