The cost that belongs to no project can still sink the company.

Every cost on a project sits at one of three distances from the work itself.

Three layers of ownership

Direct cost is the innermost layer, and you have lived in it for the whole track. It is traceable to a work item: this steel, fixed by this gang, into this wall. When you built the unit rate in Week 2, when you measured 114 tonnes in Week 8, when you split the variance in Week 6 — that was all direct cost. It is $710,000 of measured work, and it is the easiest cost in construction to control, because you can point at exactly what it bought.

Site overhead — the preliminaries — is the middle layer. It belongs to the project, but not to any single item. The site agent, the tower crane, the cabins, the security, the temporary power. You cannot bill the site manager's salary to a brick, but no brick gets laid without him. That is the $85,200 of prelims, and Week 8 already taught you its defining feature: it is earned by the calendar, not the quantity, because it is the cost of the project simply existing for a month.

Head office overhead is the outer layer, and it is the one this week is really about. It belongs to no project at all. The managing director, the accounts team, the HR department, the estimators who won the job, the rent on the office none of the sites ever visit. It is shared across every project the company is running at once, and this project's slice of it is $74,051.

The pattern that matters: the further a cost sits from the brick, the harder it is to see, the harder it is to control, and the easier it is to get catastrophically wrong. Direct cost you measure. Prelims you spread. Head office you have to recover — and recovery is where companies quietly die.

What was inside the markup

Where did that $74,051 come from? It came out of a number you have carried, unopened, since Week 6.

Back in the very first tender, the price built up to a million dollars, and sitting on top was $124,051 of “overhead and profit”. We said the words together, as everyone does, and moved on. Now we open the box.

Of that $124,051, roughly $74,051 is head office overhead — a genuine cost, money the company must spend to exist at all. The remaining $50,000 is net profit: the actual reward for taking the risk of the job, the number that is left after every real cost, direct and indirect and head-office, has been paid.

And notice that figure. Fifty thousand dollars. It is the management reserve from Week 5 that you could never touch. It is the retention from Week 17 the client holds for a year. And it is the true profit on this whole million-dollar job. The same modest number keeps appearing because it is what is actually at stake — the thin real margin that everything else in this track has been circling.

The mistake nearly everyone makes is in the phrase itself. “Overhead and profit” is spoken as one word, so it gets managed as one thing. It is two completely different things. One is a cost you cannot avoid — the company has to run whether this job exists or not. The other is the only reason the company bothers. Confuse them, and you will do something very dangerous with the next bid.

Recovery, and the bid that kills you

Here is why the distinction is not academic. The company's head office costs money every year — call it $74,051 per project across a dozen jobs, near enough $900,000 a year of directors, accountants, estimators and rent. Every project must carry a slice of that back, or the company as a whole does not cover its own costs. That slice, expressed as a percentage on top of direct cost, is the recovery rate: on this job, about ten percent.

Now watch what happens under pressure. A big job comes up. A competitor is hungry. To win it, the estimator shaves the bid — and the easiest place to shave, because it does not feel like a real cost, is the head office recovery. Cut it in half and the price drops without touching a single rate.

You win the job. And here is the trap: the managing director's salary did not fall because you bid low. Head office costs exactly the same whether this project recovers its full share or half of it. So this job now comes up $37,026 short on head office — and that shortfall does not vanish. It lands somewhere. It lands on the other projects, which now have to over-recover to cover this one, or it lands directly on the year's profit. You have won work that actively drains the company, and every progress report on that job will look perfectly healthy, because nobody made a mistake on site.

Win three jobs this way and it is $111,076 of head office that no project is paying for — a hole in the company's accounts with no corresponding hole on any site. This is the exact twin of Week 18. Overtrading ran out of cash; under-recovery runs out of margin. Both kill profitable-looking companies, both are invisible on any single project, and both are caused not by bad work but by winning the wrong work at the wrong price.

Practical insight

Find your company's head office recovery rate — the percentage added to direct cost to carry the office. Then look at what your most competitive recent bids actually recovered.

If the winning bids are quietly carrying less overhead than the company needs, you are buying turnover at the expense of margin, and the busiest year can be the one that loses the most money. A company can be full of work, every site running well, and still be shrinking — because the jobs were priced to win, not to feed the office above them.

Then ask the question that separates a cost engineer from a commercial manager: not “is this project profitable?” but “is this project recovering its fair share of the company?” They are not the same question, and the gap between them is where the company's real profit lives or dies.

Key takeaways

✔ Cost sits at three distances from the work: direct (the item), site overhead / prelims (the project), and head office (the company).
✔ The further from the brick, the harder to see and control. Direct you measure, prelims you spread, head office you must recover.
✔ The $124,051 “overhead and profit” opens into $74,051 of head office cost and just $50,000 of real profit.
✔ That $50,000 is the same figure as Week 5's management reserve and Week 17's retention — the thin real margin this whole track circles.
✔ “Overhead and profit” is two things, not one: an unavoidable cost, and the only reason the company exists.
✔ Every project must recover its share of head office — about a 10% rate here — or the company loses money quietly.
✔ Cutting overhead recovery to win a bid does not cut the overhead. The $37,026 shortfall lands on other jobs or on profit.
✔ Under-recovery is overtrading's twin: one runs out of cash, the other out of margin, and neither shows up on site.

What is coming next

If cutting the price is the dangerous way to win work, there is a disciplined way to cut the cost instead — to build the same thing for less, or the same money for more, without touching the margin that keeps the company alive.

It has a method, and a trap of its own: the cheapest way to build something is not always the cheapest way to own it, and the fastest way is rarely the cheapest.

Next week: value engineering and the time-cost trade-off — cutting cost on purpose, and the difference between price and worth.