By Greg Allan, tradesman and creator of Data First Coaching
Why do defects at handover end up costing more than they show on paper?
A defect found at the landlord or tenant walk-through rarely gets logged against the job. A shopfitter sends someone back to the site or the workshop to fix the joinery, touch up a finish or sort a snag item, and that labour goes unrecorded. The quote still shows the margin it promised. What the job actually cost to finish never gets written down anywhere.
Most shopfitters know the feeling of a job that looked solid on the quote and came out thin once it was done. The programme ran to deadline, the progress claims went out, the fit-out looked finished. Then the project manager walks the site before the landlord sign off and finds a bench edge that was not sanded, a door that binds, a shelf bracket that is not quite level. Someone has to go back and fix it, usually after hours, usually under pressure to get the keys handed over on time.
That extra time rarely gets booked back against the job. It is not a variation, because the client did not ask for it and would not pay for it. It is not covered by a progress claim. It sits outside the quoted cost of sales as unbillable labour, and because nobody tracks it against that specific job, it disappears into general overhead instead of showing up as the real reason the margin fell short.
Where does that lost time actually land, if not on the job?
It lands in the gap between what a shopfitting business's jobs are supposed to earn and what the bank balance says at the end of a busy month. Owners can point to full order books and still find the business short, because the hours spent fixing defects before handover were real paid hours that never showed up as cost against the job that caused them.
A shopfitter juggling several fit-outs at once, each with its own project manager, programme and landlord deadline, rarely stops to ask where the rework time on any one job actually went. The crew that fixed a joinery defect on a hospitality fit-out this week might be the same crew booked on a retail job next week, and the hours blur together in a timesheet nobody reconciles job by job.
This is the same pattern Greg points to with unconverted quoting: time spent driving to and pricing jobs that never get signed off is unbillable, and it quietly erodes profit the same way defect rework does. Both are hours worked that never attach to a paying job, and both hide inside a business that looks busy and profitable from the outside.
What is the one number that actually tells you whether a job made money?
Gross profit margin tells you, but only if you check it against what a job actually cost once every rework hour, every call back to site and every fix before handover is counted, not just the hours the quote allowed for. Checked once at quoting, it tells you what the job was meant to earn. Checked again at the end, it tells you the truth.
Most shopfitters set gross profit margin once, at the quote, and never look at it again until the accountant asks why the quarter came out lower than expected. By then you cannot recover the detail of which job ate the time, which defect took three site visits to fix, or which tenancy caused the most after hours trips. It is gone.
If you check gross profit margin as work moves through the business, not just at the quote, you can see a job starting to slip before it reaches handover. That is data first thinking: start from what the job actually cost, not from memory of how it felt once it was finished. It is the difference between finding out a job lost money after it is done and catching it while there is still time to manage the defect list, the client and the deadline differently.
How can an AI agent change what you actually know about a job's true cost?
An AI agent is software that does a job the way you would, following your own prices, wording and rules. AI agents like this can be set to pull every hour worked against a job, billable or not, straight from your own data, so the real cost of finishing a fit-out is visible to you instead of guessed at.
Picture a shopfitter who could never get a straight answer on what a job actually cost to run. Every job on paper looked barely profitable once office overhead was spread across it, and when that overhead was stripped back out, the number dropped to something clearly too low to be real. Neither figure meant anything, because nobody could see which hours were billable site and workshop time and which were the unbillable hours spent on rework, quoting that went nowhere, and fixing snags before a landlord walk-through.
An AI agent built on that business's own job and finance data worked out the true cost of running it, with billable hours properly factored in. All the overhead and the unbillable hours, the defect fixes, the after hours call backs, the dead quoting time, sit in their own bucket instead of being spread evenly across jobs that had nothing to do with them. AI agents built this way keep working in the background, job after job, so the costing stays accurate as the business changes, not just at the one point someone decided to check it.
What can you do differently at the walk-through itself?
You can treat the defect list the same way you treat a variation, as something with a cost that gets tracked back to the job, not absorbed as a cost of doing business. That means knowing, before the landlord or project manager walks the site, roughly what it will cost you to clear the snag list, and deciding with the client what gets fixed now and what becomes a follow up visit.
If you know your gross profit margin on a job as it stands, not just as quoted, you can have a different conversation with the project manager about the defect list. Some items are genuinely yours to fix at no charge. Others sit closer to a variation, where the scope changed after the shop drawings were signed off, or a supplier delay pushed the install into a rushed final week. Separating the two protects your margin on the job without damaging the relationship with the builder or landlord you need for the next tenancy fit-out.
You can also decide, with better information, whether a snag is worth fixing tonight under pressure or scheduling properly next week. If you can see the real cost of after hours rework against the job, you are making that call on data, not on the pressure of a looming deadline.
How does Data First Coaching approach this?
Data First Coaching starts from the shopfitter's own job and finance data, not a guess about where defects are costing money. Through the Trade Business Intelligence, it looks at what gross profit margin did on recent jobs against what the quote promised, and works back to find where the unbillable hours, rework, snag fixes and after hours call backs are sitting. As Greg puts it, “If you're driving out to jobs and quoting all the time and not converting them, then that could be costing it,” and the same thinking applies to time spent fixing defects nobody ever charged for. The how, the systems that track it and put the fix in place, is worked through inside the Academy, with members, because the Trade Business Intelligence is powered by every member and built for owners who are open minded, innovative and willing to share.

