By Greg Allan, tradesman and creator of Data First Coaching
Why does a dividing fence job between neighbours so often end up costing more than it earns?
A dividing fence job is quoted once but negotiated twice: once when the fencer prices the boundary fence, and again when the two neighbours argue over who pays what. Most of that second negotiation happens before a single post hole is dug, and it usually shortens the scope, swaps a panel, or drops a gate from the plan. The fencer's quote gets adjusted to keep both neighbours happy, but the labour and materials behind it rarely get re-checked.
That gap matters because a job can look priced correctly on paper and still lose money once the actual hours on site run over what was allowed. An auger hitting rock, an unplanned retaining issue along the line, or a neighbour wanting the fence moved after digging has already started all eat into hours the quote never accounted for. None of that shows up until the job is finished and the fencer looks at what it really took to get those steel posts in and that run of paling or chain mesh standing.
What happens when the crew does extra work and nobody raises it as a variation?
On most dividing fence jobs, the extra work gets absorbed rather than invoiced, and that silent absorption is where the margin disappears. An extra panel because the fence line moved, a gate change one neighbour asked for, or additional footings because the ground was softer than expected all cost real labour and real materials, but they rarely get written up as a variation.
The crew is usually trying to keep both neighbours happy and get off site, so an extra trailer run for materials or an hour of extra auger time just becomes part of the job rather than a line the fencer gets paid for. Multiply that across a season of dividing fence work and a business can be running a string of jobs that each looked fine on the quote but quietly handed back margin on site.
How do you protect the margin on a dividing fence job once the neighbours start negotiating?
You protect it by treating gross profit margin as something to check the moment the scope changes, not something you look at once a month in the books. A dividing fence quote that gets shortened or split between neighbours needs the same check a full boundary fence job gets: does the labour and materials still add up to a margin worth doing the job for.
That means knowing, job by job, what the fencer actually billed for labour against what the crew actually spent on site, not just whether the final invoice got paid. If a post hole and footing run that was quoted for a set number of hours ends up taking noticeably longer because of rock, a retaining issue, or a late change to the fence line, that is the moment the margin needs checking again, not after the next three jobs have gone the same way.
What number actually tells you whether a dividing fence run made money?
The number that tells the truth is the actual labour hours on that specific run of fence measured against what was quoted, not the balance sitting in the bank account. A fencer can see money land, GST sitting there waiting to be paid, and assume the job was fine, while the labour efficiency on that one dividing fence job has quietly dropped well below what the quote was built on.
Memory and a feeling about how the job went will usually say it went fine, because the invoice got paid and the neighbours stopped arguing. The actual hours and materials on that job tell a different story, and it is only that number, not the bank balance, that shows whether a decent quote turned into a loss once digging, disposal and the inevitable extra panel or gate bracket were absorbed without anyone flagging it.
How can AI agents help a fencer catch this before it eats the margin?
An AI agent is software that does a job the way you would do it yourself, using your own prices, your own wording and your own rules. For a fencer, that means an agent can sit on a dividing fence job and flag the moment extra hours, an extra panel, or a changed gate position start being logged against it, so the conversation about a variation happens before the job moves on, not after it is finished and the margin is already gone.
This is what data first means in practice: instead of starting from how the job felt on site, you start from the actual hours, materials and quote sitting in your own numbers, and the agent sits right at the point where the gap usually opens, before it becomes silent extra work absorbed into a dividing fence run. AI agents do not settle disputes between neighbours. What they do is make sure the fencer's own numbers tell the truth about the job, whichever way that argument between the two neighbours went.
How does Data First Coaching approach this?
Data First Coaching, built by Greg Allan from running his own trade business, starts from the fencer's own job and finance data through the Trade Business Intelligence, not from a description of the dispute or the neighbours involved. It looks at the gap between what a dividing fence job was quoted for and what it actually took on site, because that gap, not the argument between neighbours, is usually where the real problem starts. As Greg puts it, a fencer can find “you're losing money on every job, but you still get money in the bank”. The how of finding and closing that gap is worked through inside the Academy, with members, because the Trade Business Intelligence is powered by every member who is open minded, innovative and willing to share what they learn.
