Knowledge · Finance
Cost to complete forecasting,
the number that decides the job.
Every job has one number that says whether it will make money while that can still change, the forecast of what it will cost to finish. This reference covers how the forecast is built line by line, why uncommitted work drifts back to budget, the cadence that keeps the number honest, and how it feeds WIP reporting and cash flow.
01 / Overview
What cost to complete is
Cost to complete is the forecast of what it will still cost to finish a job from today. It sits at the centre of the wider cost control discipline because it is the piece that looks forward. Actual costs report what has already happened and committed costs report what has already been promised; the cost to complete is the judgement about everything that remains, and it is the only one of the three a builder can still influence.
The three combine into the forecast final cost, actual costs to date plus committed costs not yet invoiced plus the cost to complete on work not yet committed. Set against the contract sum with approved variations, the forecast final cost gives the current margin position on the job. Usage varies between builders (some fold committed-but-not-invoiced amounts into their cost to complete), and this library uses the terms in the strict sense above.
Why it matters
The estimate decided what the job should cost; the forecast final cost is the running answer to what it will cost, and it is the only version of that answer available while the outcome can still change. A job's profitability is not discovered at close-out, it is decided across the build, in orders placed, variations priced and problems caught early or late. A current cost to complete surfaces an overrun at 2 per cent, while it is a decision; a stale one delivers the same overrun at close-out, as history. The day-to-day tracking that feeds it is covered in the builder cost tracking guide.
02 / The lifecycle
Where the forecast sits in a residential job
The forecast begins where the estimate ends. On a win, the priced scope becomes the job budget (the seam covered by the estimate to budget handover), and from that day the budget stops being a plan and becomes the baseline the forecast measures against. Each cost review compares budget, committed, actual and cost to complete by cost code, which makes the forecast the forward-looking half of job cost reporting, the half that says what is left rather than what has happened.
Downstream, two disciplines consume the forecast directly. Work-in-progress reporting rests on it, because percentage of completion is a cost question before it is an accounting one. And cash flow forecasting is largely the cost to complete with dates attached; the remaining costs are the payments still to go out, and timing them against the claims still to come in is what turns a job forecast into a cash position. A forecast that is wrong quietly makes both of those wrong with it.
03 / Process workflow
The forecast review, step by step
Eight steps, from bringing the data current to acting on what the number shows. The walk through the uncommitted lines in the middle is where the forecast is either honest or decorative.
- 01
Bring the cost data current
Post the actuals, get the supplier invoices in, and make sure every order and subcontract raised since the last review is recorded as a commitment. A forecast built on stale data inherits the staleness, and the review spends its time arguing about inputs instead of judging the remaining work.
- 02
Walk the committed lines
For every line with a purchase order or subcontract in place, the forecast is the commitment value plus known variances, agreed extras, quantity adjustments and back-charges. These lines are largely settled; the work is confirming that what is known has actually been recorded against them.
- 03
Walk the uncommitted lines
For every line not yet ordered, ask what the work would actually cost to buy today, at current quotes and current rates, with what the drawings now show. The budget figure is the comparison, never the answer. This step is where the forecast is either honest or decorative.
- 04
Price the variations, both directions
Approved variations adjust the budget and the forecast on both the cost and the revenue side. Pending variations and unpriced site instructions get carried visibly as forecast risk, because work proceeding without a signed number is still work that will cost money.
- 05
Review the contingency balance
Confirm the remaining contingency reflects the draws actually made, and release allocations whose risks have expired as a deliberate decision on the record. A forecast showing the full allowance as available months after it was quietly consumed is wrong in the most dangerous way, confidently.
- 06
Roll up the forecast final cost
Actual costs to date, plus committed costs not yet invoiced, plus the cost to complete on uncommitted work, summed by cost code. Set it against the contract sum with approved variations and the margin position on the job is a current fact rather than a hope.
- 07
Explain the movement
Compare this forecast to the last one and give every material movement a reason, a variation, a rate movement, a scope gap found, a contingency draw. A forecast that moves without explanation teaches nothing and earns no trust.
- 08
Act while the number can still change
The point of the forecast is the decisions it enables, repricing a pending variation, resequencing a package, having the selections conversation early, or tightening supervision on the trade that is drifting. Then feed the result into the WIP position and the cash flow forecast, which both depend on it.
04 / Key mechanics
The six classes of line in a forecast
Every line in a job forecast belongs to one of these classes, and each class is forecast differently. Mixing them up, or treating them all like the budget, is how a forecast goes wrong without anyone lying.
Actual costs to date
Invoiced and paid costs against the job. This is the fixed part of the final number; nothing done at a review changes it, which is exactly why the review should spend its time elsewhere.
Committed lines
A purchase order or subcontract exists, so the forecast is the commitment value plus known variances. High certainty, small residual risk in back-charges, quantity adjustments and supply terms.
Uncommitted lines
Nothing is ordered yet, so the forecast is a fresh judgement at current rates and current quotes. This is where the real uncertainty lives, and where forecasts quietly default to the budget figure.
Approved variations
Signed changes to scope adjust both the budget and the forecast, cost and revenue together. Left out, the forecast compares the new job against the old plan and both sides of the margin are wrong.
Pending variations and instructions
Work proceeding on a site instruction or an unpriced change is forecast exposure, not future paperwork. Carry it visibly as a listed risk until it is priced and signed, or the forecast understates the job.
Contingency balance
The remaining risk allowance after logged draws. It is released into the forecast when risks expire, deliberately and on the record, never counted as margin while its risks are still live.
The optimism problem
The structural weakness of every cost to complete is that the uncommitted lines are forecast by a person who wants the job to be on budget. The budget figure is already on the page, restating a line above it means owning a problem months before anyone would otherwise notice, and so remaining work drifts back to budget by default. No single line looks dishonest. The total is simply the budget, restated monthly with increasing confidence, until the orders are raised and reality arrives all at once.
The countermeasure is procedural. Committed lines are forecast at commitment value plus known variances, which leaves little room for optimism. Uncommitted lines are forecast at what the work would cost to buy today, supported by a current quote, a current rate from the cost database, or a stated reason, and a line is not allowed to equal budget by silence. The remaining contingency is treated the same way it is set, as a ledger of named risks (the discipline covered in estimating contingency), not as a general reserve that makes any forecast add up.
Cadence, when the forecast must move
Common practice among builders who run this well is monthly at minimum, refreshed at every progress claim, and reworked on every major variation. The claim link is the non-negotiable one, because a claim asserts a percentage of the job is complete, and that assertion is only as good as the cost figures under it. The WIP reporting guide covers what happens downstream when those inputs are stale, over- and under-billed positions that look healthier than they are. How a WIP position translates into recognised revenue is an accounting treatment to settle with the builder's accountant; this page is about keeping the cost inputs worth accounting for.
05 / Best practice
How experienced builders run the forecast
The operator's observation is that the job that loses money rarely loses it in one event. There is no single day a job goes bad. It goes bad in a forecast that was never updated, a dozen small drifts, a package landing over allowance here, an unpriced instruction there, each one absorbed silently by a cost to complete still showing budget. The discipline that catches it is unglamorous, forcing every uncommitted line to justify, at every review, why it still equals budget. The lines that cannot answer are the forecast doing its job.
The same operators treat the review as a teaching instrument. A builder learns more from three honest cost to complete reviews on a live job than from any close-out post-mortem, because the review happens while the causes are still visible and the decisions are still open, and the post-mortem happens when both are gone. The pattern across reviews is the diagnostic, which trades drift, which lines were still open too late, how often a pending variation sat unpriced while the work proceeded. Jobs teach in the present tense.
Where software fits the workflow
Traditionally the forecast is a spreadsheet rebuilt each month from the ledger, a stack of invoices and memory, which is why it so often is not rebuilt at all. In VIABUILD, cost tracking holds budget, committed, actual and forecast against each cost code as orders are raised, and Oryn™ reads and codes supplier invoices as they arrive, so the fixed parts of the forecast are already current when the review starts. The review then spends its time where the judgement lives, on the uncommitted lines, and the builder makes every call with the trail recorded.
06 / Australian considerations
The forecast in the Australian environment
Cost to complete forecasting is an internal discipline, not a legislated one, but it operates inside a regulated contract environment and, at the moment, a hard market. The points below are labelled by evidence class; figures and requirements change, so confirm the current source before relying on any of them.
- Statistics. A record 3,596 Australian construction companies entered external administration for the first time in FY 2024-25, up 21 per cent on the prior year, and construction tops the industry count for insolvencies. Insolvency figures move each reporting period, so confirm the current numbers against ASIC's published statistics. The pattern behind the number is stable, though; building businesses rarely fail for lack of work, they fail on cash and on jobs whose losses were discovered late, which is the case for the forecast made in the building through a downturn guide.
- Legislation. The variations that move a forecast are regulated. Each state and territory's domestic building contract legislation sets rules for how variations to residential work are documented, priced and approved, and the rules differ by jurisdiction. A forecast carrying work that proceeded without an approved variation is carrying a compliance question as well as a cost one.
- Common practice. WIP schedules built from cost to complete are what accountants, financiers and home warranty insurance eligibility assessments commonly ask a builder to produce, and the quality of the forecast underneath decides whether those schedules survive scrutiny. How WIP flows into recognised revenue is an accounting treatment; take that question to the builder's accountant rather than to a web page.
- Professional recommendation. Industry commentary through the current contraction has been consistent that cashflow discipline matters more in a downturn, not less, and that builders who keep cost to complete current sit in a materially different position to those who do not. In a rising-cost market the forecast decays faster, so the cadence that was comfortable in a flat market is usually too slow.
07 / Common mistakes
Where forecasts actually go wrong
Each of these is recognisable, mechanical and avoidable, and none of them requires anyone to lie. Drift does the damage on its own.
Uncommitted lines forecast at budget
The budget is the number you want, so remaining work gets forecast at budget by default. Each line looks fine, the total looks fine, and the overrun stays invisible until the orders are raised and it is too late to do anything but absorb it.
The forecast never updated
A cost to complete produced once, early, then rolled forward untouched is not a forecast, it is a photograph. The job that loses money rarely loses it in one event; it loses it slowly, inside a number nobody re-examined.
Pending variations invisible
Work proceeding without a signed variation sits in nobody’s numbers. The cost lands in actuals eventually; the revenue side may never land at all. The forecast should carry every unpriced instruction as a named exposure.
Contingency shown as still available
Draws made by small site decisions and never logged leave the forecast showing an allowance the job has already spent. The cost to complete looks covered right up until the moment it is not.
The forecast owned by accounts alone
Accounts can total what has been spent; only the person running the job can say what remains and what it will really cost. A forecast assembled after the fact from the ledger reports history with a forward-looking label on it.
Movement without explanation
A forecast final cost that jumps between reviews with no stated reasons cannot be trusted or learned from. The explanation discipline is what turns a monthly number into an early-warning system.
08 / Practical example
A worked forecast at frame stage
Illustrative only, not a benchmark. A custom home is contracted at $850,000 against a budgeted cost of $760,000. At frame stage, actual costs sit at $320,000 and committed costs not yet invoiced at $240,000. The uncommitted lines total $180,000 at budget, so the lazy forecast writes $180,000 in the cost to complete column and reports a forecast final cost of $740,000, comfortably inside budget. Everyone relaxes.
The honest walk tells it differently. The painting quote has come in $6,000 over allowance, the cabinetry package is not yet ordered and current quotes are running $8,000 over, and the landscaping market has moved about $4,000 since pricing. An unpriced site instruction for extra drainage is carried as a further $3,000 exposure. The cost to complete is $201,000, the forecast final cost is $761,000, and the margin has slipped by roughly $21,000 against plan. At frame stage that is a set of decisions, price the drainage as a variation, requote the cabinetry, bring the selections conversation forward. At close-out the same $21,000 would simply be a smaller number in the bank, with nothing left to decide.
09 / FAQ
Common questions.
Cost to complete is the forward-looking piece, the forecast of what it will still cost to finish the job from today. Forecast final cost is the whole picture, actual costs to date plus committed costs not yet invoiced plus the cost to complete on uncommitted work. Usage varies between builders, and some use cost to complete for everything not yet invoiced, committed or not. What matters is that the definition is consistent inside the business, so a number said in a meeting means the same thing to everyone in the room.
Monthly at minimum, at every progress claim, and whenever a major variation lands. The claim link matters because claiming revenue against a stale cost forecast produces a margin position that is fiction in both directions. In practice the update is cheap when the cost data is current, because actuals and commitments are already recorded and the review only has to judge the uncommitted lines. Builders who find the monthly forecast a heavy exercise usually have a data currency problem, not a forecasting problem.
The person who runs the job, with accounts supplying the recorded actuals and commitments. The committed side of the forecast is arithmetic anyone can do; the uncommitted side is judgement about work not yet bought, and that judgement lives with whoever knows the site, the outstanding selections and the state of the drawings. Many builders find a short monthly review works best, the numbers prepared beforehand, the meeting spent only on the lines that moved and the lines still open.
Because the budget is the number the business wants to be true, and restating a line above budget means admitting a problem early and owning it. Anchoring does the rest; the budget figure is on the page, and confirming it takes a second while challenging it takes a phone call. The working countermeasure is procedural rather than motivational. Every uncommitted line must justify why it still equals budget, with a current quote, a current rate or a stated reason, and silence is not a justification.
Work-in-progress reporting compares what a builder has billed with the value of work actually done, and the percentage of completion behind it rests on cost figures, which makes the cost to complete one of the load-bearing inputs. A stale or optimistic forecast flows straight through to an over- or under-billed position that looks healthier than it is. How WIP translates into recognised revenue is an accounting treatment for the builder’s accountant; the builder’s job is keeping the cost inputs honest enough to be worth accounting for.
It converges. Early on, most of the job is uncommitted and the forecast carries real uncertainty; as packages are ordered, lines migrate from judgement to commitment and the band of possible outcomes narrows. The forecast final cost should move for stated reasons and settle as the job closes out, so the final account is a confirmation rather than a surprise. A forecast that stays glued to budget for eight months and then jumps at close-out was never a forecast; it was the budget wearing a different column heading.
10 / Terms
Glossary for this topic
Cost to complete (the forecast of remaining spend to finish the job), forecast final cost (actuals plus committed plus cost to complete), committed cost (money promised by an order or subcontract but not yet invoiced), actuals (costs invoiced or paid), uncommitted line (a budget line with nothing yet ordered against it), forecast movement (the change between one review and the next, always with a reason), WIP (the work-in-progress position that rests on these numbers). Definitions for the wider vocabulary live in the construction glossary.
The single biggest force that moves a forecast between reviews is changed scope, and how changes get priced, approved and kept out of the drift is the subject of variations in residential building work.
11 / Keep reading
Related knowledge, guides and features
12 / Further reading
Primary sources
- Australian Securities and Investments Commission , publisher of the insolvency statistics behind the external administration figures, the reference for current numbers.
- Australian Construction Industry Forum , publisher of national construction forecasts, the reference for the cost and volume environment a forecast operates in.
- Your state or territory's building regulator and fair trading body, for the domestic building contract rules that govern how the variations feeding a forecast must be documented and approved.
Know what the job will cost while you can still change it.
VIABUILD holds budget, committed, actual and forecast against every cost code as the job runs, so the cost to complete is a live number the business can steer by, not a spreadsheet rebuilt at close-out.