Improving dispatch decisions

Emergency jobs were assigned based solely on technician availability, resulting in repeat visits and slow response times.

Every repeat visit consumed a second slot, which made the schedule tighter, which made dispatch reach for whoever was free. The problem fed itself.

Dispatch was optimising the only variable it could see. Skill coverage and location existed in people's heads, not in the dispatch view, so they could not be part of the decision.

Nearly a third of emergency callouts were second visits to a job already attended — capacity the business had already bought and was spending twice.

Change the decision inputs before changing any tooling. A better-informed human dispatcher beats an automated one working from the same thin data.

Deliberately parked: routing software. It was worth revisiting only once the SOP proved the skill and geography inputs actually changed outcomes.

Redesigned the dispatch SOP so every emergency assignment weighs technician skill match and geographic routing alongside availability.

Trained dispatch staff on the new process and instrumented it against a baseline, so first-time fix and response time are reviewed weekly rather than assumed.

A decision is only as good as the inputs in front of the person making it. Most dispatch problems are information problems wearing a scheduling costume.

Repeat work is the cheapest capacity in any service business, because you have already paid for it.