Monday 06:30, one email. Last week's jobs, one line each. Hours booked, materials cost, invoice value, margin. The three worst at the top, in red.
Most workshop owners know roughly which jobs went badly. Roughly is the problem. The one that felt like a disaster was fine because the materials were cheap. The one that felt easy took 31 hours instead of the 18 you quoted, and nobody noticed because it was spread across three people.
The system pulls three things you already have. Hours from wherever people book time, a sheet, a time app, a note on the job card. Materials from the matched supplier invoices. Invoice value from Xero or QuickBooks. It puts them against the job number and does the sum.
It does not need perfect data. If a job has no hours booked, it says no hours booked. A blank is better than a margin that looks fine. If an invoice has not been raised yet, the job shows as open.
Every week, same time, same format. After two months you have a list of job types that lose money, and that list changes what you quote.
This is the kind of thing we build. It runs in your own accounts. Which job last month took the most hours against what you quoted?