Written for people who run the work.

Take an operator at $38 an hour. Payroll taxes, workers' compensation, general liability, health coverage and paid time off routinely add 30 to 40 percent on top. That $38 operator costs you somewhere north of $51 an hour before they've touched a machine.

Now run that through a job. On a crew logging 1,200 hours, a 35 percent burden you didn't cost is roughly $16,000 of real money that never appeared in your estimate. On a job bid at 12 percent margin, that alone can be the difference between a good week and a wasted one.

The failure isn't that contractors don't know burden exists. It's that burden gets applied once a year, as a rough percentage, at the accountant's request — rather than being carried into every estimate and every job cost as it happens.

There are two fixes and you want both. First, calculate burden by employee classification rather than as one company-wide number: a foreman with family coverage and a first-season laborer do not carry the same rate, and averaging them means you overprice one crew and underprice the other. Second, apply it automatically at the point the hours land on the job, so nobody has to remember.

The test is simple. Pull a job you closed last month and ask what one field hour cost you on it. If the answer is a wage, you have a burden problem. If the answer is a wage plus a rate you can defend by classification, you don't.

Most contractors have decent job costing. It just arrives too late to be useful.

Consider a 12-week mill and overlay running 8 percent over on labor. Discovered in week two, you have ten weeks to change how the remaining work is crewed, to price the change orders properly, and to have a conversation with the customer while you still have goodwill. Discovered at final reconciliation, you have a number and a feeling.

The same information is worth something entirely different depending on when it shows up. That's the whole argument for real-time job costing, and it's why 'we do job costing' and 'we know where our jobs stand' are not the same claim.

The practical bar is lower than most people assume. You don't need perfect data. You need field hours entered the same day, material invoices coded to the job when they arrive rather than at month-end, and someone whose job it is to look at estimated against actual once a week.

If that sounds like more admin, notice what it replaces: the reconciliation scramble at closeout, the argument about which change orders were approved, and the annual surprise about which job types actually make money.

A contractor with sound unit pricing can still lose money consistently on small work, and usually doesn't know why. The unit price is right. The problem is everything that happens before the first ton goes down.

Getting a crew and equipment to a site, set up, and productive costs roughly the same whether you're placing 400 tons or 4,000. Lowboy time, fuel, the hour the crew spends staging, the trip back — none of it scales with the size of the job.

On a large job that fixed cost disappears into the unit price. On a small one it can be a third of the total cost, and if your bid recovers it proportionally rather than explicitly, you've priced it as though it scaled.

The fix is to break mobilization out as its own line rather than burying it in the unit rate. It makes small jobs look more expensive, which is uncomfortable, and it's also accurate. Contractors who do this usually find two things: some small jobs they were happy to win they should have been declining, and some they thought were marginal are fine.

The related discipline is minimum job size. If you can state yours, and know the arithmetic behind it, you're ahead of most of the market.

Ask an owner what constrains growth and you'll hear cash, or people, or equipment. Ask what would happen if they disappeared for a month and the real answer usually surfaces.

Owner dependency is rational in the early years. You priced the work because you knew the costs. You handled the problems because you'd seen them. Every one of those decisions was correct at the time, and together they built a business that only functions with you in the middle of it.

The cost shows up in three places. Growth caps out at your personal reach. Good people leave because there's no decision they own. And the business is hard to value, because a buyer is really buying you.

The way out is narrower than 'delegate more'. Pick the three processes that break first when you're away — usually pricing, scheduling, and problem escalation. Write down how each one actually works, including the judgement calls. Give one person ownership of each, along with the numbers they need to make the call without you.

The numbers part is where most attempts fail. You cannot hand someone responsibility for job profitability and then keep the job costs to yourself. Visibility isn't a nice-to-have in delegation; it's the mechanism.

Expect it to be worse before it's better. The first month of someone else pricing work will produce decisions you'd have made differently. That's the cost of finding out whether the process you wrote down was the real one.

An independent contractor negotiates material pricing perhaps once a year. The supplier's representative negotiates it every day, with full knowledge of what every other contractor in the market is paying.

That asymmetry matters more than volume does. You can be a perfectly good negotiator and still be working from a single data point — what you paid last year — while the person across the table has the whole distribution.

This is the least discussed reason contractors join buying groups. The headline is bulk pricing. The underrated part is simply knowing where your number sits relative to everyone else's, which changes the conversation before it starts.

You can do a version of this yourself. Pick your top three material spends. For each, find two contractors outside your market — genuinely outside, so nobody is protecting a competitive position — and compare notes honestly. Most owners are surprised in at least one category.

The uncomfortable part is that suppliers know which of their customers benchmark and which don't, and price accordingly. Being visibly informed is itself worth something.