Ask most owners how they arrived at their change order contingency line and the honest answer is: it's what they used last time, or what their PM rounded to. That's not change order risk mitigation — it's a placeholder standing in for math nobody actually did. The real number is calculable, and it's usually more specific, and more actionable, than a flat percentage carried forward from a different project with a different set of drawings.

The distinction matters because a contingency line and a risk exposure number answer two different questions. A contingency line asks "how much extra budget should we hold." A risk exposure calculation asks "how much of that budget is actually at stake, and why" — and only the second question tells you whether spending money now to catch a conflict is worth it against what that conflict costs later.

What a contingency percentage actually represents

Contingency isn't supposed to be a flat number picked once and left alone — it's supposed to track how much is still unknown about the project, and that unknown shrinks as design matures. AACE International's cost estimate classification system, the reference framework capital projects use industry-wide, ties contingency directly to how developed the design is: a project early in design, still at a conceptual estimate class, can reasonably carry a double-digit contingency percentage against everything that hasn't been resolved yet. A project doing a final check estimate against a nearly complete, issued-for-construction set should be down to a low-single-digit contingency, because by that point most of what a contingency exists to cover should already be resolved on the drawings.

That's the gap most owners don't act on. Contingency percentages get set at the concept stage and then coast, unadjusted, all the way to bid — even though the set in front of them at bid is far more resolved than the one the original percentage was set against. Change order risk mitigation means updating that number as the documents mature, not defending a percentage picked eight months earlier.

The real formula: exposure, not contingency

The standard risk exposure formula is simple: probability of the risk occurring, multiplied by the cost if it does. Applied to design-related change orders specifically — the category driven by cross-discipline coordination conflicts, not owner-requested scope changes — that means two inputs an owner can actually estimate before bid:

ILLUSTRATIVE EXPOSURE CALCULATION$40M hard cost, IFC-stage set — not a specific project
Design-related change order range (industry)3%–5% of hard cost
Unweighted exposure range$1.2M–$2.0M
Document-level review cost$9,000 one-time, or $7,000/mo unlimited / 48 hours
Review guarantee thresholdFree unless $50,000+ found

That $1.2M–$2.0M range isn't a prediction that this specific project will see that much in change orders — it's the size of the exposure sitting in an unchecked set, based on where design-related change orders have historically landed as a share of hard cost. The second input, whether that risk is still live, is the part a contingency line can't tell you and a document-level review can.

Where the 3%–5% range comes from

That figure isn't specific to Preempt Global's work — it's a consistent range across published construction cost studies, and it's worth being precise about what it does and doesn't include. Total change order costs on major projects are commonly reported in the 10%–15% range of contract value, with some projects running past 25%, but that figure blends every category: owner-requested scope changes, unforeseen site conditions, and design-related errors and omissions together. The design-error slice specifically — conflicts between disciplines, contradictions between specs and drawings, the category a coordination review is scoped to catch — is consistently estimated at roughly 3%–5% of total project budget on its own, separate from scope changes an owner chose to make.

Independent Project Analysis, which has tracked project performance across more than 12,000 capital projects, has found that over a third of construction projects experience at least one major change over their lifecycle. Not every one of those is a design-coordination conflict, and not every design-coordination conflict becomes a change order — some get caught and redlined before they cost anything. But the base rate is high enough that treating a design-related change order as an edge case, rather than a near-certainty on a set this size, is the wrong assumption to build a contingency number on.

Worth knowing

The FMI/PlanGrid industry study, based on nearly 600 construction professionals, put total U.S. construction rework cost at roughly $177.5 billion a year, with poor communication between project participants responsible for about 26% of it. A duct routed through a beam because the mechanical and structural sheets were never checked against each other is exactly the kind of communication gap that number describes — and it's the gap a document-level review before bid is scoped to close.

Why the number should shrink before bid, not after

The exposure number isn't fixed — it moves depending on when you check. As covered in how a single missed MEP clash turns into a $400,000 change order, the same conflict costs almost nothing to fix as a redline before the set is issued, and can clear $300,000–$400,000+ once it's caught in the field instead. That means the 3%–5% exposure range isn't a fixed cost of doing business — it's the cost of not checking, and it collapses toward the low end of that range the moment someone actually reads the full set for cross-discipline conflicts before bid.

That's also why timing changes what the same review is worth. As discussed in pre-bid vs. IFC review timing, a conflict caught before the documents go out to bid is still a redline nobody has priced around yet. The same conflict caught later, after subcontractors have already bid the set as drawn, means change orders that reconcile the difference between what was priced and what actually has to get built — a more expensive and more contentious version of the identical fix.

Turning the math into a decision

Once exposure is a real number instead of a habit, the decision an owner is actually making becomes clear: is it worth a small, fixed cost to find out how much of that $1.2M–$2.0M range is actually sitting in this specific set, versus carrying the full range as unpriced risk into bid? A document-level review scoped specifically to cross-discipline conflicts — reading the mechanical, structural, and spec sections against each other, the way covered in anatomy of a findings report — doesn't eliminate the exposure range, but it replaces "somewhere between $1.2M and $2.0M, unknown" with a specific dollar figure tied to specific findings, each one still cheap to fix because the set hasn't gone to bid yet.

Key takeaways

  • A contingency percentage and a risk exposure number answer different questions — contingency is a budget buffer, exposure is what's actually at stake and why.
  • AACE's cost estimate classification system ties contingency to design maturity: it should shrink as the set gets more resolved, not stay flat from concept through bid.
  • Design-related change orders — driven by cross-discipline coordination conflicts, not owner-requested scope — are consistently estimated at roughly 3%–5% of total project budget, separate from total change order costs of 10%–15%+.
  • Independent Project Analysis has found over a third of construction projects experience a major change across their lifecycle, based on more than 12,000 projects tracked.
  • The exposure range collapses toward its low end once a document-level review actually checks the set for cross-discipline conflicts before bid — timing is what turns an unpriced range into a fixed, cheap-to-fix findings list.

A contingency line carried forward from the last project isn't change order risk mitigation — it's an unexamined guess dressed up as a budget line. The actual math is available before bid, not after the first RFI: a historical exposure range, and a specific check of how much of that range is still live in the set in front of you.

Frequently Asked Questions

What's the difference between a contingency line and a change order risk exposure calculation?

A contingency line is a budget buffer, usually a flat percentage carried forward from past projects. A risk exposure calculation multiplies the probability a specific category of change order occurs by its likely cost, based on the actual maturity and condition of this project's documents — it's a calculated number, not a habit.

How much of a project's budget is typically at risk from design-related change orders specifically?

Published industry estimates put design-related change orders — driven by cross-discipline coordination conflicts and errors, not owner-requested scope changes — at roughly 3%–5% of total project budget. That's separate from total change order costs across all categories, which commonly run 10%–15% of contract value on major projects and can exceed 25% on some.

Does contingency percentage stay the same throughout a project?

It shouldn't. AACE International's cost estimate classification system ties contingency to how developed the design is — a conceptual estimate early in design can justify a double-digit contingency, while a set nearly ready for bid, with most decisions resolved, should carry a much lower percentage. A contingency that hasn't been revised since the concept stage is probably overstated or understated for the set actually in front of you.

Why does catching a conflict before bid change the math, not just the timing?

Because the exposure range is a range of what's unchecked, not a fixed cost. A conflict caught before bid is a redline that costs close to nothing. The identical conflict caught after subcontractors have priced the set as drawn becomes a change order that reconciles the gap between what was bid and what has to get built — the same conflict, at a different point on the cost curve.

How do you turn a historical exposure range into a specific number for one project?

By checking the actual set for the conflicts that drive that exposure category — reading each discipline's drawings and specs against every other discipline's, not just against themselves — before the documents go to bid. That converts an unpriced range into a findings list with real dollar figures attached, while the fixes are still cheap.