What Your Last Twenty Jobs Actually Cost

Name the last job that beat its bid.

Not the one that felt good. The one you can prove, from a report, without calling anyone.

Most owners cannot do it, and the reason is not carelessness. It is that the estimate and the actuals live in two systems that were never reconciled, so the only place they have ever been compared is in somebody's memory of how the job went.

Key Takeaways

• You cannot name the last job that beat its bid without calling someone to ask.
• Your profit and loss looks acceptable while the jobs underneath it are a mix of winners and losers you cannot separate.
• Costs are landing on jobs a month after the work happened, and nobody flags it because the report still prints.
• You suspect your estimates carry padding, but you have never measured how much or told your estimators it is there.

Why Can't I Tell Which Jobs Made Money?

Because the number you bid and the number you spent are recorded in different places, on different timelines, by different people, and nobody owns comparing them. Your estimate lives in an estimating tool or a spreadsheet. Your actuals accumulate in the general ledger as invoices clear and payroll posts. Unless someone deliberately ties one to the other job by job, the comparison never happens.

That is why the profit and loss can look fine while the individual jobs underneath it are a mix of winners and losers you cannot separate. The blend hides both.

The fix is not a better report. It is naming who owns the reconciliation and when it happens.

What Makes a Job Cost Report Wrong Even When the Math Is Right?

Three things, and all three are input problems rather than reporting problems. Costs land in the wrong period. Costs land on no job at all. And the estimate they are being measured against was never a clean number to begin with.

The period problem is the most common. A piece of equipment gets used on a job in April, the invoice arrives in May, and the job looks profitable in the April review because the cost has not shown up yet. Nothing is wrong with the report. The report is describing a cost base that is incomplete.

The unassigned-cost problem is quieter. Costs post to cost of goods sold with no job attached, which means they are correctly reducing company margin and invisibly missing from every job. The company number is right. Every job number is optimistic.

The third one is the hardest to see, because it requires someone to distrust the estimate.

How Do Estimating Cushions Distort Job Margin?

A cushion built into an estimating template and never documented makes every job look worse than it is, which changes bidding behavior in exactly the wrong direction.

A commercial contractor client of ours had this running for years. Their estimating tools carried safety nets that had been added at some point and never written down. Perceived margins came in lower than actual margins as a result, and estimators, reading the numbers in front of them, priced defensively. The business ended up using incentive compensation to push bidding back up, which is treating the symptom. The cushion was never the problem. The cushion being invisible was.

The same distortion runs the other way when the template is optimistic, and it is harder to catch, because the jobs come in under and everyone blames execution.

What Does It Look Like When the Estimate Is Never Reconciled?

It looks like three different companies discovering the same gap in three different places.

A specialty construction client of ours builds large, unique, highly variable jobs, and had no repeatable estimating model. Each significant bid was constructed from scratch. That is defensible once and dangerous as a system, and it came to a head when a customer wanted a complete quote inside forty-eight hours covering travel, insurance, and overhead. Nothing existed to build it from quickly: no checklist, no cost-category template, no captured institutional knowledge. Speed and accuracy were in direct conflict because the estimating process had never been made repeatable.

A trades client of ours in the glass and glazing business had the reverse problem. The job-level data existed. Labor and materials were being captured in their scheduling and production system, and the structure to support back-costing was there. What was missing was consistent entry of actuals and any reconciliation against the accounting system, so producing a monthly view required assumptions and pro-rating. The reported blended gross profit ran near fifty-two percent over six months while the trailing twelve sat at forty-one, and historical apportioning inconsistencies made the backward-looking numbers unreliable enough that the team treated them as directional. They judged it good enough for now and prioritized capacity, which was a reasonable call and a knowingly deferred one.

A past construction client of ours had solved the schedule and not the calendar. Their work-in-progress reviews ran monthly with real cost pooling behind them, and the reporting structure was mature. What kept the month open was cost timing and a close date that had never been locked, so the reconciliation happened against a moving base.

What Is Missing When You Have Job Costing and Still Cannot Trust It?

One person accountable for whether the reported job position reflects what actually happened on the job. Recording is happening, often well. Verification against operating reality is not, and no report can perform that check on itself.

Concretely, that means owning three disciplines. Costs matched to the period the work occurred rather than the period the invoice arrived. Every cost carrying a job, with a standing report that catches the ones that do not. And the estimate reconciled to actuals at closeout, with the variance explained rather than displayed.

That is CFO-level work in a contracting business, and it is the specific work a bookkeeper is not positioned to do, because it requires the standing to tell you your estimating template has been wrong for three years.

What Changes When You Can Answer the Question?

Bidding stops being defensive. When your estimators trust the margin they are looking at, they stop padding against a number they have learned to distrust, and you stop paying incentive compensation to correct for a reporting artifact.

Closeout becomes a decision point rather than an administrative step. You learn which job types actually carry margin, which customers cost more to serve than they pay, and which crews consistently beat or miss the plan, from data instead of impression.

And the next bid gets priced against what the last twenty jobs actually cost.

Your Next Step

You should be able to name the last twenty jobs that beat their bids, from a report.

You do not need a new system to begin. Pull your last twenty completed jobs, put the bid next to the actual cost, and see how many you can populate without asking anyone a question. The ones you cannot fill in are the answer.

The Financial Control Score Quiz scores whether your job-level numbers are solid enough to bid from. Seven questions, about a minute.

Take the Financial Control Score Quiz Book a Discovery Call

Common Questions

Frequently Asked Questions

What Is a Job Cost Report?

A job cost report compares what a specific job was estimated to cost against what it actually cost, broken out by category such as labor, materials, subcontractors, and equipment. Its value depends entirely on whether costs are complete, assigned to the right job, and recorded in the right period.

Why Don't My Job Costs Match My Financial Statements?

Usually because costs are landing in a different period than the work, or posting to cost of goods sold with no job attached. Both leave the company-level number correct while every individual job reads optimistic, so the two views legitimately disagree.

How Often Should Job Cost Reports Be Reviewed?

At minimum monthly alongside the close, and again at closeout for every completed job. The monthly review catches drift while the job is still running; the closeout review is the only point where estimate-to-actual variance can be explained while anyone still remembers why.

What Should a Job Cost Report Include?

Estimated versus actual by cost category, percent complete, costs committed but not yet invoiced, and a variance figure with an explanation attached. A report that shows the variance without the reason moves the analysis onto whoever reads it.

Who Should Own Job Cost Accuracy?

One named person, and not the person entering the transactions. Accuracy of entry and verification against operating reality are different jobs, and asking one person to check their own work removes the point of the check.