Why is cost to complete always wrong in construction?

Four failure modes cause it: PMs estimate percent complete based on schedule, not cost burn; costs are not coded to the job so actuals are incomplete; change orders add scope without updating the cost-to-complete baseline; and rework costs hit the job without being reflected in the revised estimate. Any one of these corrupts the number.

What happens when cost to complete is wrong?

Everything downstream breaks. WIP schedule shows the wrong over/underbilling position. Projected job margin is wrong. Cash flow forecast is wrong. The owner makes decisions based on a financial picture that does not match reality. By the time the real number surfaces at closeout, there is nothing left to do about it.

How do you get an accurate cost to complete in construction?

Accurate cost to complete requires three things: actual costs coded to the job by cost category so you know what has been spent, a PM who estimates remaining work in cost dollars not just schedule percentage, and a monthly review process that compares actuals to the estimate by cost code before finalizing the projection.

Who is responsible for cost to complete on a construction job?

The PM owns the cost-to-complete estimate. The controller owns the actual cost data it feeds on. The CFO owns the review and challenge process — catching when the PM estimate is optimistic and the actuals say otherwise. When all three are working, cost to complete is reliable.

JOB PROFITABILITY

COST TO COMPLETE IS
ALWAYS WRONG.

THE SHORT ANSWER

Cost to complete drives your WIP schedule. Your WIP drives your projected job margin. Your projected job margin drives your cash forecast. When cost to complete is wrong — and it almost always is — every decision downstream gets made on a number that doesn't exist.

BY JOSH LUEBKER UPDATED MAY 2026 THE CONSTRUCTION CFO

WHY IT BREAKS

FOUR WAYS COST TO COMPLETE
GETS CORRUPTED.

It is not that project managers are bad at estimating. Most of them are actually good at knowing what work is left. The problem is the inputs they are working from — and the process that is supposed to challenge them when the number looks too good.

PERCENT COMPLETE BY SCHEDULE, NOT COST

The PM says 60% complete because the schedule says 60%. But the job has burned 72% of its labor budget. Schedule and cost burn are two different things. Using schedule to estimate remaining cost produces a number that is almost always optimistic — and the error compounds every month until closeout.

ACTUALS ARE INCOMPLETE WHEN THE ESTIMATE IS MADE

The PM reviews the cost-to-complete before the close is done. Invoices from the prior month are still processing. Two weeks of labor haven't posted. The actual cost base is understated by $40K–$80K and the PM doesn't know it. They project from a number that isn't real yet.

CHANGE ORDERS ADD SCOPE WITHOUT UPDATING THE BASELINE

A $120K change order gets approved. The work starts. But nobody updates the cost-to-complete baseline to reflect the new scope. The job now shows worse margin than it should because $120K of new cost hits without a corresponding estimate update. The PM looks at the job and thinks it's losing money. It isn't. The baseline is just wrong.

REWORK GETS ABSORBED WITHOUT REVISION

Rework happens on almost every job. It rarely gets billed. The cost hits the job — labor, materials, lost productivity — but the revised cost-to-complete never gets updated to reflect it. The PM absorbs the rework into the remaining budget and the margin quietly erodes. Nobody flags it until the final billing event.

THE CASCADE

WHAT BREAKS DOWNSTREAM
WHEN THE NUMBER IS WRONG.

Cost to complete is not just a job-level number. It is the input that drives four other financial outputs. When it is wrong, all four of them are wrong simultaneously.

THE DOWNSTREAM FAILURE CHAIN

WRONG COST TO COMPLETE

↓

WRONG PERCENT COMPLETE → WRONG WIP POSITION

↓

WRONG OVERBILLING / UNDERBILLING BALANCE

↓

WRONG PROJECTED JOB MARGIN

↓

WRONG CASH FORECAST → WRONG OWNER DECISIONS

The PM estimates cost to complete on a Friday. The CFO builds the monthly WIP from it. The controller closes the books with that WIP. The owner reviews it in the monthly meeting and decides whether to bid the next project. Every step in that chain ran on a bad input.

Catching this early is the job. A cost-to-complete review run within 10 days of the prior month-end close catches labor trending over before it becomes a full-margin problem. T&M work that needs to be billed before leverage is gone gets flagged here. Rework that needs a change order gets surfaced here. At closeout, none of those levers exist anymore.

WHAT ACCURATE LOOKS LIKE

THE EXAMPLE THE BOOK
USES.

Real job costing means your PM can answer this question without calling the bookkeeper: How much have we actually spent on labor on Phase One as of last month, fully burdened? The answer should come from a screen, not from memory or a spreadsheet being emailed around.

TWO SCENARIOS — SAME QUESTION, DIFFERENT OUTCOMES

Budget

$32,000

Spent to date

$31,542

$28,480

% of budget spent

98.6%

89%

% of work complete

100% — phase done

70% — phase not done

Signal

Saved $458. Phase complete.

Trending $5K over. Fix it now.

Scenario B is not a crisis if you catch it at 70% complete. There is still 30% of the phase left to manage. You can adjust crew mix, accelerate material delivery, or tighten productivity targets. Catch it at closeout and the only thing left to do is explain the loss.

THE FIX

THREE THINGS THAT MAKE
COST TO COMPLETE RELIABLE.

CLOSE FIRST, ESTIMATE SECOND

Cost-to-complete reviews run after the books are closed — not before. The controller closes by the 10th. The CFO runs cost-to-completes after that. The PM sees actual costs as of the close date, not a partially-posted mess. Every cost code shows what has actually been spent before the PM estimates what is left.

ESTIMATE IN COST DOLLARS, NOT SCHEDULE PERCENT

The PM answers: how many labor hours are left at what burden rate, what materials need to be purchased, what equipment is still needed? That produces a cost number. The percent complete falls out of the math — it does not drive it. When PMs estimate in cost first, the number is grounded in what is actually left to build.

CFO CHALLENGES THE NUMBER EVERY MONTH

The CFO reviews the cost-to-complete against the actuals and asks one question on every job: does this estimate make sense given what we have already spent? If the job is 75% spent on labor but the PM says 55% complete, something is wrong. The CFO surfaces it before it becomes a loss at closeout. This is the review process most subcontractors are missing entirely.

The Job Profitability System inside CFOS builds this review into the monthly cadence. Cost-to-completes run within 10 days of the close. Every active job reviewed. Trending issues flagged while there is still something to do about them.

COST TO COMPLETE ISALWAYS WRONG.

FOUR WAYS COST TO COMPLETEGETS CORRUPTED.

WHAT BREAKS DOWNSTREAMWHEN THE NUMBER IS WRONG.

THE EXAMPLE THE BOOKUSES.

THREE THINGS THAT MAKECOST TO COMPLETE RELIABLE.

CLOSE FIRST, ESTIMATE SECOND

ESTIMATE IN COST DOLLARS, NOT SCHEDULE PERCENT

CFO CHALLENGES THE NUMBER EVERY MONTH

COMMON QUESTIONS.

THE GAP DOESN'T CLOSEWITHOUT THE SYSTEM.

COST TO COMPLETE IS
ALWAYS WRONG.

FOUR WAYS COST TO COMPLETE
GETS CORRUPTED.

WHAT BREAKS DOWNSTREAM
WHEN THE NUMBER IS WRONG.

THE EXAMPLE THE BOOK
USES.

THREE THINGS THAT MAKE
COST TO COMPLETE RELIABLE.

THE GAP DOESN'T CLOSE
WITHOUT THE SYSTEM.