Why Most Custom Build Budgets Fail
at 80% Complete

Custom construction projects have a recognizable failure curve. The budget tracks well through the early phases. The monthly forecast holds steady. Then, somewhere between seventy and eighty percent complete, the forecast cracks. By the time the project is in its final phases, the original budget bears no resemblance to the actual final cost.

The owner is presented with explanations: scope creep, change orders, late finishing decisions, material price increases. Each explanation is partially true. None of them, individually or collectively, accounts for the full variance.

The pattern is not random. It is structural. And it is almost always produced by the same forecasting flaw: a methodology that distorts in predictable ways during the middle phases of a project, then unwinds the distortion during the final phases. The variance was always there. The methodology was hiding it.

The 80% Surprise

Owners who have been through this experience describe it the same way. It tracked fine, then it didn't. The shift feels sudden, but it is not. The shift was always going to happen at roughly the same point in the project. The forecast was producing readings that were artificially smooth during the middle of the build, and the smoothness was unsustainable.

This pattern is so consistent that we now treat it as the default expectation when reviewing a custom build's forecast methodology. If a contractor presents a project that is 60% complete with a forecast that has barely moved from the original budget, our first question is not how did you do this? Our first question is show me the methodology, because the math suggests this is structurally improbable.

"The 80% surprise is not a surprise. It is the methodology unwinding the smoothness it produced earlier."

Estimate at Completion, Defined

Estimate at Completion (EAC) is the running forecast of what the project will cost when finished. The basic formula is simple:

EAC = Costs Incurred to Date + Costs Remaining to Complete

The first term is observable. The contractor knows what has been spent. The receipts and invoices are in the books. The challenge is the second term: costs remaining to complete. This is a forecast. It is a prediction about future expenditures based on the work still to be done.

How that prediction is generated is the entire question. There are two dominant methodologies, and they produce very different results.

Why Percent-Complete Forecasts Lie

The most common methodology is percent-complete. For each scope on the project, the contractor estimates what percentage of that scope is finished, multiplies the contracted value by the remaining percentage, and sums the result to produce the costs-remaining number.

It is intuitive. It is widely used. And on fixed-price contracts, it is structurally distorting.

The distortion arises because percent-complete reporting on fixed-price contracts is not a measure of cost. It is a measure of progress, and progress does not predict cost. Consider a fixed-price subcontract for $100,000. The trade reports they are 50% complete. The percent-complete methodology says costs remaining are $50,000. But the trade's actual costs to date may be $30,000 (if they front-loaded their efficient work) or $70,000 (if they front-loaded their difficult work). The contracted value does not change. The trade will receive $100,000 regardless.

But the methodology cannot detect early-phase efficiency or inefficiency, because the methodology is reading completion, not cost.

The Stair-Step Problem

The distortion compounds at scope-level transitions. When a major fixed-price scope (cabinetry, for example) transitions from 0% complete to 25% complete, the EAC drops by 25% of the contract value. When it transitions from 25% to 50%, EAC drops again. The forecast moves in stair-steps that correspond to the contractor's percent-complete estimates, not to actual cost behavior.

Now consider what happens late in the project. Most of the fixed-price scopes are at 75-90% complete. The remaining work is concentrated in finishing scopes (paint, trim, detail work), which do not have the same fixed-price structure and which tend to overrun. Plus open allowances that have been accumulating variance the methodology did not surface. Plus change orders that were verbally agreed but never priced.

All of this hits the EAC at once, in the final 20% of the project. The methodology that smoothed the variance during the middle phases now has nowhere to hide it. The forecast jumps. The owner sees a budget failure. The contractor offers explanations. None of the explanations, individually, accounts for the magnitude of the shift, because the shift is the methodology unwinding, not a single failure event.

The Milestone-Binary Alternative

The fix is to abandon percent-complete reporting for fixed-price scopes. In its place: milestone-binary triggers.

Milestone-binary methodology divides each fixed-price scope into a set of defined milestones (mobilization, rough-in, fixtures set, substantial completion, deficiency cure). Each milestone is binary: either it has been achieved or it has not. There is no partial credit. The cost incurred against the contract is calculated by summing the value of milestones achieved, not by estimating completion percentage.

The milestone-binary method produces a forecast that looks worse during the middle phases. Early on, scopes that are partially executed but have not hit their next milestone show as zero progress, even when work is clearly happening on site. This appears overly conservative. Owners and contractors sometimes resist it on these grounds.

The conservatism is the point. The methodology refuses to credit work until it can be definitively measured. It produces a forecast that is harder to manipulate, harder to smooth, and harder to use to defer recognition of variance. It also produces, at the end of the project, a forecast that did not surprise anyone.

What Owners Should Demand

Owners do not need to specify the methodology themselves. They need to ensure the methodology is appropriate to the scope structure of the project, and they need to verify that it is being applied consistently. The minimum standard:

Monthly EAC reporting with explicit methodology documentation. The owner should know how the forecast is being calculated, not just what number it produces.

Milestone-binary triggers for all fixed-price scopes, with the milestones defined in writing at contract execution. Trades agree to the milestones. Reporting tracks against them.

Allowance tracking in committed dollars, separately from fixed-price scopes. An allowance is not at 60% complete; it has $30,000 committed of a $50,000 budget. That is a different sentence and a different number.

Variance attribution. When EAC moves, the owner should receive a written explanation of which scopes drove the change and why. Variance without attribution is a methodology that is hiding something.

An honest middle-phase forecast. The EAC should be permitted to show variance early, even when that variance is uncomfortable. A forecast that holds steady through the middle of a custom build is not a virtue - it is a warning sign.

The Architect's Stake

For architects, the EAC distortion problem has a specific consequence. When the budget cracks at 80%, the explanations offered by the contractor often point upstream to design changes, late decisions, or specification revisions. Some of these are real. Many are not. The variance the methodology was hiding, when it surfaces, gets attributed to whatever recent events are available to absorb the blame.

Architects who certify progress on projects using percent-complete methodology are, in effect, validating a forecast that will likely produce a late-project surprise. When that surprise arrives, the architect is sometimes positioned as having missed warning signs that, in fact, the methodology suppressed. The defense is to insist on milestone-binary reporting at the start, not to defend a percent-complete methodology after the fact.

"Forecast methodology decisions made at contract execution determine which ending the project gets - clean close-out, or a finger-pointing exercise about who was responsible for the late-stage variance."

What This Means for Project Selection

For owners and architects evaluating a contractor at the proposal stage, the EAC methodology is one of the most useful early diagnostics. The questions to ask:

How will you produce the monthly EAC forecast?

For fixed-price scopes, will you report on percent-complete or on milestone-binary triggers?

Can you walk me through how you forecast a scope that is 40% physically complete but where the trade has not yet hit its next milestone?

The contractor's answer to these questions reveals the maturity of their financial governance. A contractor who has not thought carefully about EAC methodology will give imprecise answers. A contractor who has built their practice on the lessons of past projects will answer specifically, with reference to the actual mechanics they use. The difference is observable in five minutes of conversation.

This is not the only diagnostic. It is one of the most reliable ones, because the answer cannot be faked from a marketing brochure. The methodology either exists in the contractor's working practice or it does not.