The Real Cost of Poor Change Order Management

Change orders are not the problem. Unmanaged change orders are the problem. A project with zero change orders is a project with hidden problems; a project with disciplined change orders is a project that's behaving.

Three controls do most of the work: a contracted change-order process that names who can authorize what at what threshold; a contingency waterfall the lender approves up front; and a weekly OAC change-order review with hard accountability.

The threshold matrix is the most underused tool. A typical structure: PM up to $25K, owner's rep up to $100K, owner up to $500K, board above. With the matrix in place, the team stops debating who can sign and starts debating whether the change should happen.

The contingency waterfall should be written, lender-approved, and treated as a contract document. We see projects with 5% contingency that get drained by month four because every change was treated as an emergency. With a waterfall in place, 70% of the contingency stays for actual surprises.

Weekly review is the third leg. The change order log is reviewed in OAC, with the owner present, and every open item has a name and a date. Change orders left without an owner accumulate like silt; weekly review keeps the channel clear.

What we see on poorly managed $50M projects: 12–18% change-order leakage by closeout. What we see on disciplined projects: 2–5%. The delta — call it 10% of $50M — is $5M of avoided spend that pays for owner-side PM 8× over.

Practical first steps for an owner who's already mobilized: name the authority matrix, draft the contingency waterfall, schedule weekly change-order reviews, and stop signing anything that doesn't fit the process. The discipline pays back inside one billing cycle.