The Jobs You Lost Money On Before You Won Them

There's a particular kind of margin problem that doesn't show up on site, in delivery, or in project management. It shows up months earlier — in the estimating system — and by the time it reaches the P&L, it's untraceable.

It happens when a business quotes on a cost base that's no longer true.

How the cost base goes stale

Every project or job-based business runs on a set of assumed costs: labour rates, material prices, subcontractor rates, plant and equipment costs, and an allowance for overheads. At some point, someone built those numbers carefully. They were accurate — then.

Since then, wages have stepped up. Key materials have repriced, sometimes more than once. Subcontractors have lifted their rates. Insurance, fuel and site costs have all moved.

The estimating library, meanwhile, has stayed where it was. Updating it is nobody's job, there's never a convenient time, and the business is busy. So quotes keep going out priced against a cost structure that existed twelve or eighteen months ago.

Why stale rates feel like winning

Here's the part that makes this leak so hard to spot: quoting on old costs makes you cheaper. Win rates improve. The pipeline fills. The workshop and the crews are flat out.

Every visible signal says the business is going well — right up until the financials arrive. Revenue is strong, activity is high, and profit is thin. The natural instinct is to look for a delivery problem: blowouts, rework, poor supervision. Often there isn't one. The jobs were delivered roughly as planned. They were simply priced below their real cost from day one.

That's what makes this failure mode expensive. The margin gap was locked in at the point of quoting, and every month of healthy-looking sales was compounding it.

The fix is cadence, not heroics

Most businesses that recognise this problem respond with a one-off repricing project — a painful few weeks rebuilding the rates, followed by another long period of drift. The rates are accurate again for a quarter or two, and then the same decay begins.

The durable fix is structural:

• A scheduled review rhythm. Cost inputs get reviewed on a fixed cadence — quarterly as a floor, monthly for volatile categories like key materials or fuel-exposed costs.

• Actuals feeding estimates. Completed jobs are compared against their estimates, and the variances feed back into the rate library. Estimating stops being a static document and becomes a loop.

• Clear ownership. Someone owns the cost base. Not the estimating software — a person, with a recurring calendar entry.

• A trigger threshold. When a supplier or subcontractor notifies a price change above a set threshold, the library gets updated then, not at the next scheduled review.

None of this is sophisticated. It's the same principle that applies across financial management in operational businesses: the numbers you make decisions with have to be current, or the decisions are made against a business that no longer exists.

A full order book priced on stale rates isn't a strong position. It's a queue of commitments to yesterday's economics.