By Greg Allan, tradesman and creator of Data First Coaching
Why does a full order book not mean the business is making money?
Because busy and profitable measure two different things. Busy is a job by job reading, how many sets of shop drawings are moving, how much steel is going through cutting and folding, how many welders are on the tools. Profit is a reading across the whole workshop at once, including the time between jobs and the jobs still waiting on a finish.
A fabricator can have every bay full, the laser cutting running, aluminium and stainless jobs going out the door on schedule, and still be losing money on half of them without knowing which half. Each job gets priced, built and invoiced on its own, so it can look fine in isolation. The business only shows the truth once every job is counted together against what it actually cost to run the shop that week, not what the quote said it should cost.
Where does the money actually go missing on a fabrication job?
It goes missing in the parts of the job that never get a line on the quote. Non-billable workshop time, cleaning up offcuts, moving steel between stations, waiting on material to arrive, does not disappear, it gets absorbed quietly into the jobs that are billable instead.
So does rework, when a weld does not pass certification first time and has to be redone. So does the dead time waiting on powder coat or galvanising turnaround when a gate or balustrade is finished in the shop but sitting on a rack until the finisher can take it. None of that shows up as its own cost. It gets spread across every job as part of the overhead, so the quote still looks healthy and the fabricator still gets paid, while the actual hours spent on site or in the shop quietly blow past what was ever allowed for.
What number actually shows whether a fabrication job made money?
The number that tells the truth is billable hours measured against hours actually paid for, across the whole workshop, not job by job. This is a data first way of looking at the business, because it is built from what actually happened that week, not from how the job felt when it was quoted.
Gross profit margin is the figure to watch here, and it needs protecting and growing every month, not just checked once when a big structural steel job comes in. A fabrication business can have GST and PAYG sitting in the account, which makes the balance look healthy right up until that money is due to be paid out. Knowing the true cost of running the workshop against what is actually billable is what tells a fabricator whether they are profitable, instead of hoping the number in the account means they are.
What changes when an AI agent takes on this tracking?
An AI agent is software that does a job the way the owner would do it themselves, using the business's own prices, wording and rules. For a fabricator, that means an agent that pulls the real wages and overhead figures straight from the business's own data and works out the true cost of running the shop, with billable hours factored in properly instead of guessed at.
Picture a fabricator who could never get an accurate cost of operations figure. Every time the overhead was spread across jobs the way it always had been, it loaded around eighty dollars an hour onto every job, making almost everything look unprofitable on paper. Strip the overhead out completely and the number dropped to around eleven dollars and fifty-four cents an hour, which was clearly too low to be real either. An AI agent built on the business's own live data worked out where the true figure actually sat, put all the overhead and non-billable cost into its own bucket instead of spreading it across every job, and kept the costing accurate from there on. The fabricator stopped guessing whether a job had been profitable and started knowing.
Should a fabricator build their own costing system or rent one off the shelf?
The honest answer is it depends on the business, and the fabricator's own data is what shows which way to go. Some off-the-shelf job management tools suit a small shop well enough, particularly early on, and there is no need to throw out something that is already working.
What changes the equation is when a fabricator is paying for software built for every trade at once, stainless benchmakers, aluminium window fabricators, structural steel yards, all lumped into the one generic tool, when what actually runs the business is shop drawings, cutting, folding, welding, and the wait on finishing. A system shaped around how this fabricator's own jobs run and how builders and engineers actually buy from them is a different thing to a rented tool built for nobody in particular. AI agents can sit inside either, but they do their best work once they are built around the fabricator's own numbers.
How does Data First Coaching approach this?
Data First Coaching, built by Greg Allan, starts from the fabrication business's own numbers rather than asking the owner to explain the business from memory. The Trade Business Intelligence looks first at billable hours against hours paid for, because that is where busy and profitable come apart. Greg has described the trap plainly: “the job still gets done, you still charge the money, so you think you're doing okay,” and the fix starts with “billable hours for the day, divided by hours paid.” Greg has been open that before this kind of tracking existed, he would say, “I wasn't tracking profitability on jobs. I was just going, yeah, I made some money there.” How this gets built for a fabricator's own workshop is worked through inside the Academy, with members, because the Trade Business Intelligence is powered by every member for every member, and that only works for owners who are open minded, innovative and willing to share.
