At 20 percent complete, a projected overrun is a problem with options. You can value engineer a scope, renegotiate a trade, adjust the finish schedule, or go to your capital partner early with a plan and your credibility intact.
At 90 percent complete, the same number is a disclosure. The work is bought, the trades are on site, and the only remaining variable is who absorbs the difference.
The number is identical. The timing is everything. This is what cost-to-complete forecasting is for, and why the frequency of the calculation matters more than its precision.
What is cost-to-complete forecasting?
Cost-to-complete forecasting projects the final cost of a project by combining what has been spent, what has been committed, and what remains to be committed, against the revised budget. The output is a projected cost at completion per cost code and for the project as a whole, alongside the variance from budget.
Four inputs drive it.
- Revised budget. The original budget adjusted by approved change orders. Not the original budget, which stopped being relevant at the first change.
- Committed cost. Signed contracts and purchase orders, including the portion not yet invoiced. This is money spent in every sense except the ledger.
- Incurred cost. Invoices received and posted, whether paid or not.
- Project manager forecast. The estimate to complete for work not yet committed, which is where human judgement enters and where most forecast error originates.
The calculation itself is arithmetic. The difficulty is keeping all four inputs current at the same time.
Why do manual forecasts arrive too late?
A manual forecast is a data-gathering exercise before it is a calculation. Someone pulls commitment balances from the accounting system, chases project managers for updated estimates, reconciles change orders that were approved in the field but not yet recorded, and assembles the result.
That process takes days, which means the forecast describes a project position that is already several days stale when it is delivered. On a monthly cycle, the practical effect is that decisions are made on data between one and five weeks old.
Three structural problems make it worse than the delay alone suggests.
Change orders lag the field
A change order approved on site changes the revised budget immediately in reality and weeks later in the system. Every forecast produced in that window is calculated against a budget that no longer exists.
Commitment data is incomplete
If contracts and purchase orders are not tracked systematically, committed cost has to be estimated, which converts a known quantity into a guess. Firms without commitment tracking consistently understate their exposure.
The estimate to complete is asked for, not calculated
Project managers are asked for a number, often at the end of a busy month, and they supply one. Without a structured basis, that estimate tends to be optimistic early and revised sharply late, which is exactly the pattern that produces late-stage surprises.
What changes when forecasting is automated?
Automation does not make the forecast more accurate in principle. It makes it current, which in practice makes it far more accurate at the moment you need it.
The calculation runs continuously
Budgets, commitments, change orders and forecasts live in one dataset, so the projected cost at completion recalculates whenever any input changes. The figure on the dashboard is current when you look at it rather than current as of the last time someone rebuilt it.
Change orders flow at approval
When the change order workflow is part of the same system, approval updates the revised budget immediately, and every downstream figure moves with it. The gap between field reality and reported position closes.
Variance surfaces without being asked for
Automated alerts flag cost codes trending over budget as the trend emerges rather than at month end. This is the difference between finding a problem and being told about one.
The forecast is auditable
Each revision retains its prior value, the person who changed it and the date. When a capital partner asks why the projection moved, the answer is in the record rather than in someone’s recollection.
Elevate’s real estate development software calculates cost-at-completion from budgets, commitments, change orders and PM forecasts across the portfolio, with drill-down from portfolio level into individual cost codes.
How to read a cost-to-complete report
The headline number is the least useful figure on the report. Three others carry more signal.
Variance by cost code, not in total. A project on budget overall can be badly wrong in both directions, with an underspend on one trade masking an overrun on another. The trades that are over are the ones that will keep going.
Percentage committed against percentage complete. If a project is 40 percent complete but 75 percent committed, the remaining scope has to be delivered within a quarter of the budget. That relationship predicts trouble earlier than the variance does.
Movement between periods. A projection that moves steadily in one direction across three reporting periods is describing a trend, not noise. The direction matters more than the current value.
A forecast that only ever reports the total is not being read properly, whatever system produced it.
Where forecasting connects to the rest of the project
Forecast quality depends on data originating outside the accounting function. Change orders come from the field. Commitments come from procurement. Progress comes from the site. When construction and development accounting run on the same platform, that data arrives without re-entry, and the delay between an event happening and the forecast reflecting it collapses. The same structure applies to raw land, where cost-to-complete is complicated further by phasing and lot takedown schedules.
Elevate is built on Acumatica, whose construction accounting platform supplies the underlying job cost and commitment structure.
What to ask a vendor about forecasting
- Approve a change order and show me how long until the cost-at-completion figure reflects it.
- Show me committed cost against incurred cost for a live project, per cost code.
- How does a project manager enter an estimate to complete, and what basis is it recorded against?
- Show me the same project’s forecast as it stood three months ago and explain the movement.
- What triggers an alert on a cost code trending over budget?
- Show me percentage committed against percentage complete across the portfolio.
The first question is the most diagnostic. If updating the forecast requires anyone to do anything, the forecast will lag the project permanently, and the lag is where the risk lives.
Getting forecasting right
Two practices separate firms whose forecasts are trusted from firms whose forecasts are argued about.
The first is a disciplined change order process. If changes are approved verbally on site and documented later, no system will keep the forecast current, because the input itself is late.
The second is a structured basis for the estimate to complete. Asking project managers for a number produces a number. Asking them to build it from remaining scope, current unit costs and known risks produces a forecast that holds up when the project gets difficult.
Neither is a software problem, but the right software makes both considerably easier to sustain.
Elevate is an Acumatica Gold Certified Partner founded by CPAs, with nearly 40 years supporting real estate developers. Book a 20-minute discovery call to see cost-at-completion forecasting on live project data.
