All insights
Operator Handbook6 May 20265 min read

The geo-report gap: why sitework eats margin between bore holes

A geotechnical report tells you what is at three or four points on the site. Builders who price sitework as a locked sum are pricing those points as though they represent the entire allotment. The gap between bore holes is where margin disappears.

Nathan Holloway

Founder · BuildHawk and Hawktress

Sitework kills more first-year margins than any other trade on the schedule. Not because builders are careless. Because sitework is the one stage where the ground decides what it costs, and the ground does not care what the estimate says.

The mistake is consistent: a builder prices sitework as a locked sum at tender, reviews the geotechnical report, and assumes the report tells the full story. They win the job. They mobilise. And somewhere between topsoil strip and slab prep, they hit conditions that the report never captured. Rock in a zone between bore holes. Unexpected fill in a corner of the site that was never tested. A perched water table that only reveals itself when excavation opens up the full footprint.

Every one of these is a cost the estimate did not carry. Every one of them comes off margin.


Why the geo report is not the whole picture

A geotechnical report is a mandatory document. No building permit issues without one. The builder has it at tender, and they price sitework on the basis of what it says. That is the correct starting point. It is not a complete picture of what is in the ground.

A standard residential geotechnical investigation involves three to four bore holes across the site. Those bore holes produce accurate data at the exact point they were drilled. What they do not produce is certainty about every other point on the allotment. The ground between bore holes is interpolated, not tested. On a site with variable geology, fill of unknown origin, or a history of previous structures, the conditions between test points can be materially different from the conditions at them.

A builder who reads a geo report and locks sitework as a confirmed sum is pricing three to four data points as though they represent the entire site. In many cases that assumption holds. In enough cases it does not, and the cost of being wrong lands entirely on the builder's margin.

Why year one is the highest-risk period

An experienced builder develops a calibrated sense of sitework exposure over time. They know which postcodes carry rock risk between bore holes. They know what variable fill looks like in a geo report and how to price around it. They have been caught out enough times that they have built a structural allowance for the unknown into their sitework approach.

A builder in their first year does not have that history. They read the geo report, price what it says, and trust the data. They are not yet pricing the gap between what the bore holes found and what the excavator might find. That gap is where the margin exposure lives.

The spoil calculation builders get wrong

Spoil removal is one of the most consistently underpriced line items in residential sitework, and the error is almost always the same. The builder calculates the volume of material to be excavated, applies a truck rate, and puts a number in the estimate. What they do not account for is bulking.

Soil does not leave a site in the same volume it occupies in the ground. Clay bulks at 20 to 30%. Rock bulks at 30 to 40% or higher. On a standard residential allotment requiring 200 cubic metres of cut, a 25% bulk factor adds 50 cubic metres of material that was never in the disposal calculation. At a haulage and tipping rate of $120 per cubic metre, that is $6,000 unaccounted for before a single trade has raised a variation.

The compounding factor is contaminated or classified material. Standard fill goes to a standard landfill at a standard rate. Fill containing asbestos, hydrocarbons, or treated timber products is classified waste, running at three to five times the standard disposal rate. A geo report with three to four bore holes cannot confirm what is sitting in every untested zone. A builder who assumes standard disposal across the entire site is carrying both a financial and a regulatory risk inside the same unpriced line item.

On the wrong site, spoil is a $15,000 to $40,000 exposure sitting inside a number the builder never questioned.

The correct position: allowance, not locked sum

Having a geotechnical report does not mean sitework should be priced as a locked sum. It means the builder has confirmed data at three to four points and interpreted data everywhere else.

Sitework at estimation stage should be structured as a base confirmed cost, derived from what the geo report actually establishes, plus an explicitly qualified provisional sum for the conditions the bore holes did not test. Not a vague contingency. A separately identified line item in the contract with a defined scope, a defined trigger, and a defined pricing mechanism if adverse conditions are encountered.

A locked sum transfers all geological risk, including the risk of what sits between the bore holes, to the builder the moment the contract is executed. If conditions are worse, the builder wears it. If conditions are better, the client keeps the saving. The asymmetry is entirely in the client's favour.

The margin arithmetic on a missed sitework call

A builder prices sitework at $45,000 on an $850,000 contract. The geo report confirmed reactive clay at all four bore hole locations. Sitework is locked at tender on that basis.

Excavation commences. Rock is encountered at 600mm across 40% of the site, in a zone between two bore holes that both returned clay. Rock breaking and removal adds $28,000. Spoil volumes run 30% over estimate because bulking was not applied. Disposal adds a further $8,000. Total sitework cost lands at $81,000 against a $45,000 allowance. Variance: $36,000.

That $36,000 is unrecoverable. The geo report was in hand. The contract was signed on a locked sitework sum. There is no contractual mechanism to pass the cost to the client. It comes directly off the $212,500 margin target, reducing the effective margin from 25% to 20.3%.

On one project. Before the frame goes up.

The upstream fix

Read the geo report for what it confirms, not what it implies. Data at bore hole locations is confirmed. Everything between those points is interpreted. Price accordingly.

Carry a provisional sum for inter-bore-hole risk on every site where geology is variable, fill history is unknown, or previous structures have been demolished.

Apply a bulk factor to every earthworks volume calculation. Minimum 20% for clay. Minimum 30% for rock. Confirm the rate with the excavator before finalising the disposal figure.

Obtain a current feature and level survey before finalising earthworks quantities. Planning-stage contour data is not a reliable basis for a cost.

Price rock separately and always. If the geo report shows rock at any bore hole location, carry a provisional sum for zones between test points with a day-rate mechanism for actual removal.

Review the dial-before-you-dig report at every tender. Service locations shift. An old report is not a reliable basis for a cost.

The sitework stage is where the ground between the bore holes takes over. The builder's job is to price that uncertainty correctly before signing, not to discover it mid-excavation and absorb it in silence.

— BuildHawk

Written by

Nathan Holloway

Founder · BuildHawk and Hawktress

Start your brief

Related insights

See all
Video
Operator Handbook4 May 2026

Variation Control: The System That Removes Discretion

A builder can price a job accurately, manage trades well, and still finish in the red. In most cases, the cause is the same: variations that were completed but never properly controlled.

By Nathan Holloway · 5 min read

Video
Operator Handbook8 Oct 2024

Why every builder underprices sitework in a new region

Excavation, drainage, retaining and site clean. Together they are usually 4 to 7% of the contract. Builders almost universally get them wrong on the first job in a new region. Here is what to actually budget.

By Nathan Holloway · 6 min read