Calculating job margin: how to know upfront whether a job will lose money

By Redazione VERC · 2 settembre 2026 · 9 min read

There is a way of working that is very common in small manufacturing businesses: you quote from experience, you accept the job, you produce it, you invoice it. At the end of the year the accountant tells you whether the company made money. If the result is disappointing, no one can say which jobs produced it.

Calculating job margin serves exactly this purpose: moving the answer from twelve months later to before you confirm. It doesn't require a complex management-control system. It requires three numbers, built well and used every time.

In brief

• Margin is calculated on the price, markup on the cost. Confusing them makes you price below your intentions, always in the same direction.
• The machine hourly cost must be divided by the real productive hours, not by the opening hours. It is the error that underestimates everything else.
• The contribution margin, not the net margin, is the right number for deciding whether to accept a job.
• Quotes lose money for five recurring causes, all of them predictable.
• The margin should be seen while you build the quote, not after the fact.

Markup and margin: the confusion that costs the most

They are two different percentages calculated on different bases, and they are used as synonyms every day.

Markup = (price − cost) / cost Margin = (price − cost) / price

Cost 100, price 150. The markup is 50%, the margin is 33.3%.

The business owner who says "I work at 30%" almost always means a 30% markup and believes they have a 30% margin. They are working at 23%. On a turnover of 500,000 euros, the difference between believing you have 30% and having 23% is 35,000 euros a year, which is often the entire profit.

To achieve a desired margin, the markup to apply is:

Markup = desired margin / (1 − desired margin)

A 35% margin? A 53.8% markup.

The machine hourly cost

It is the pillar of all industrial quoting, and it is almost always miscalculated in one single place: the denominator.

The numerator: the annual costs

• Depreciation: purchase value divided by the years of useful life. If the machine is fully depreciated but still in use, include a replacement allowance anyway, because one day it will have to be replaced.
• Maintenance: service contract, spare parts, callouts. Use the average of the last three years.
• Energy: average power draw × operating hours × cost per kWh.
• Space: square metres occupied × cost per square metre (rent or notional cost).
• Operator: the full company cost, not the net pay.
• Share of indirect costs: administration, general utilities, insurance, allocated across the machines with a simple, stated method.

The denominator: the real productive hours

Everything is decided here. The theoretical hours in a working year on one shift are about 2,000. The hours in which the machine actually produces are far fewer: you have to subtract downtime, maintenance, setups, waiting for materials, job changeovers, low-season periods.

In a small manufacturing business, an effective utilisation of between 55% and 70% of the opening hours is a realistic figure. Take 60%: 1,200 productive hours.

Example. Total annual costs of the machine: 48,000 euros.

• Divided by 2,000 theoretical hours → €24/hour
• Divided by 1,200 productive hours → €40/hour

Whoever quotes at €24/hour is giving away 16 euros for every hour of work, and will never see it in any quote. They will see it in the accounts, without understanding where it comes from.

The way to know the real utilisation is to measure it for a few weeks, even crudely: a sheet next to the machine with the start and end time of each job is enough to produce a number infinitely better than an eyeballed estimate.

Contribution margin: the number for deciding

The net margin of a job, which includes the allocation of fixed costs, is for after-the-fact analysis. To decide whether to accept a job you need a different number.

Contribution margin = revenue − direct variable costs

Variable costs are those that exist only because that job exists: raw materials, outsourced operations, specific packaging, dedicated transport, direct labour if it is genuinely variable.

The practical sense is this: if you have spare production capacity, a job with a positive contribution margin is worth accepting even if it doesn't cover its share of fixed costs, because you pay those fixed costs anyway. If instead capacity is full, the same job has to be compared with the one occupying the machine in its place.

It is the reason why "this job doesn't cover its costs" can be a wrong decision in January and a right one in September.

The five reasons quotes lose money

1. Underestimated setup. The setup time is remembered on the low side, because what sticks in memory is the time it went smoothly. On short runs the setup is the dominant item, and getting it wrong by 40% means getting the quote wrong by 25%.

2. Waste treated as a percentage. The paper or material consumed during setup is a fixed quantity, not a proportion. Applying a flat 5% on a short run means accounting for a fraction of the real consumption.

3. Unbilled changes. The customer asks for a change mid-job, you accept it as a commercial courtesy, no revision is issued. These are real hours that no one pays for. The remedy is procedural, not technical: any change after confirmation generates a revision of the document, even at zero cost.

4. Absorbed one-off costs. Cutting die, plate, sampling done "to win the job". If the repeat order never comes, they remain a dead cost. They should be tracked as separate items, so that at least you know how much you invested in that customer.

5. The outdated price list. The cost of raw materials changes and the prices stay as they were six months earlier, because updating all the files takes an afternoon that never comes free. It is the problem described in managing quotes with Excel, and it is solved only by taking the price list out of the scattered spreadsheets.

Building the margin into the quote

After-the-fact margin analysis is useful, but it arrives when the decision has already been made. The real step up is having the margin in front of you while you build the quote.

The mechanism is simple, and it requires only one thing: that for every price item there is a corresponding cost. If the price list is structured as a price matrix of operation × quantity, place a cost cell next to every price cell. The quote, as it is assembled, adds up both.

At that point the salesperson sees in real time:

• The overall margin of the job
• The margin per individual operation, which tells you where you're losing money
• The immediate effect of a discount requested by the customer

This last point changes negotiations. A 10% discount on a job with a 25% margin burns 40% of it. Seeing this while the customer asks for it on the phone is different from calculating it afterwards.

The discount threshold to know by heart

You need a single table, to keep on hand: how much extra volume is needed to offset a discount.

How to read it: with a 30% margin, a 10% discount requires 50% more volume just to stand still. When a customer asks for the discount promising more work, this is the table that tells you whether the promise is worth it.

Where to start

1. Calculate the real hourly cost of two or three machines, with the right denominator.
2. Take five jobs closed in the last six months and recalculate their margin with those costs. This is the moment the surprises turn up.
3. Check whether you are confusing markup and margin in your current prices.
4. Place the cost alongside the price in the list, at least on the main operations.
5. Make sure the margin is visible at the quoting stage, not only after the fact.

The first two steps take a day and usually return the most useful piece of information of the year.