You Built to the Permit. Change Orders Came Anyway.
Change orders are driven by scope that wasn't drawn, existing conditions nobody could see, and code interpretation you don't control. Each source has a different mitigation — design discipline, physical investigation, inspector engagement — and you have a role in managing all three. With construction pricing where it is, a 10% change-order run on a $5M build-out is $500K.
With construction costs as high as they are, a change order is no longer a rounding error. It's real money that arrives mid-project, when your leverage is lowest and your schedule is hostage to the answer. Founders ask me all the time how to push that risk away from them. The honest answer: you can move some of it, price some of it, and absorb the rest knowingly. You can't eliminate it. Managing it well starts with knowing where change orders actually come from.
People are repeatedly surprised that your team can do everything right and change orders still happen. The drawings get reviewed, the contract gets negotiated, the permit gets stamped...and the corrections come anyway. It's unfortunately not a failure of diligence. It's a structural feature of the nature of construction: three different kinds of information simply cannot fully exist on the day you kick off the project.
The three sources: scope that wasn't drawn, conditions that couldn't be seen, and interpretation you don't control. Budgets fail when all three get lumped together as "overruns." They behave differently, they price differently, and they're managed differently. A single contingency line is not a risk mitigation strategy.
What scope gaps cause change orders?
The first source is the most controllable: work that wasn't in the drawings when the price was set. Drawings freeze at a moment in time. Your operations don't. The clinical director you hire in month three walks the space and wants the med room chagnged. The phlebotomy chair needs a sink the program didn't call for. A payer contract shifts and suddenly two exam rooms need to flex for telehealth.
None of those are bad decisions. The problem is when they get made: after bid day, every one of them gets priced by a single contractor with no competition, at 15–20% margins instead of the 2–3% carried to win the job. The decision costs the same either way. The timing decides who has pricing power.
The mitigation is decision discipline, not design perfection. Set freeze dates tied to design milestones and put your clinical leadership in front of the plans physically before they go to bid, in a walkthrough, not on a PDF. When a late change is genuinely worth it, make it with the premium visible. Sometimes the med room really does need to change. Pay for it as a decision, not as a surprise.
What existing conditions cause change orders?
The second source lives behind the drywall. In second-generation space the as-builts are aspirational, the panel schedule hasn't been accurate since 2009, and the slab has opinions nobody recorded. Demo opens the walls and the building tells you what it actually is: undersized electrical service, plumbing two feet from where the drawings promised, an HVAC unit with three winters left in it.
You can't eliminate this risk, but you can shrink it dramatically before you're exposed to it. Walk the building with your GC and engineers before you sign, not after. Pay for the investigation that's cheap relative to what it de-risks: a panel load study, a slab scan where new plumbing lands, exploratory demo in the suspect corners. Size a real construction contingency (5–10%, toward the high end for older second-gen space) because whatever the investigation misses will lead to a change order.
Who prices that invoice is a function of your contract. I broke down how lump sum, GMP, and cost-plus each allocate this risk — the short version is that unforeseen conditions burn visible contingency under a well-governed GMP and generate marked-up change orders under a lump sum. Same wall, same surprise, very different math.
Why do inspectors issue corrections on work built to the permit?
The third source is the one no amount of upfront rigor removes: you can build exactly to the permitted design and still get corrections. Plan check approved your drawings. The field inspector inspects your building. Those are (usually) different people, on different days, applying a code that leaves generous room for interpretation, and the field always wins.
I've stood in spaces where the accessibility clearance matched the approved plans and the inspector measured the built condition differently. Where fire-caulking details passed plan review and got written up at rough inspection. Where a corridor rating question that three professionals considered settled got reopened at the walk. The correction is legal reality even when the drawing was approved. Arguing that plan check signed off is not a timeline and cost management strategy, and it won't get your next inspection scheduled faster.
This is why "we'll catch everything up front" is not a plan. Interpretation risk can't be caught up front, because it doesn't exist yet. It materializes when a specific inspector meets your specific building. What you control is the relationship, the response time, and the budget line that absorbs it.
The management here is orchestration, not prevention. Meet the building department before you mobilize. A pre-construction meeting surfaces how this jurisdiction reads the gray areas. Get inspectors on site early and often rather than saving everything for finals; a correction caught at rough-in costs a fraction of one caught behind finished drywall. Treat the inspector as a professional doing their job, because they are. The relationship you build across visits is pays dividends when you're in a time crunch. I wrote about orchestrating the human side of the city in the regulatory bottleneck piece; the same logic runs through inspections. And carry a small allowance — budget and schedule — specifically for corrections, so the first one doesn't blow the plan.
How do you manage the change orders you can't prevent?
Once construction starts, the goal shifts from preventing change orders to controlling how they get priced, decided, and absorbed. This is process, and it's worth being rigid about:
- Every change order gets priced before the work proceeds — itemized, from the labor rates and unit costs in the contract, with schedule impact stated in days. "We'll true it up later" is how a $12K change becomes a $40K dispute.
- One owner-side decision-maker, on a clock. Route every CO through a single person empowered to approve, reject, or negotiate within 48–72 hours. A change order that sits without engagement results in delay claims.
- Keep a running log against contingency. Know your total exposure weekly. The jobs that end in shock aren't the ones with change orders; they're the ones where nobody was adding them up.
- Separate the ledger by source. Scope decisions, field conditions, and corrections are different negotiations. A tagged log also tells you on the next site which source actually hurt you, and where the next dollar of prevention should go.
One more piece of honesty: a project with zero change orders usually isn't a clean project. It's a padded one. The risk was priced invisibly into the number before you ever saw it. You're not buying zero risk on a clinic build, because that product doesn't exist. You're buying visibility and control over the risk that is always there.
So the reframe I'd offer: stop aiming for no change orders and start aiming for no surprises about who pays and why. Scope changes should arrive as decisions you made with the premium visible. Field conditions should land against a contingency you sized for this building on purpose. Corrections should hit an allowance you carried because interpretation risk is real. When each source has a home in the budget, a change order stops being a strain on your relationship with architects and contractors and becomes what it actually is: the cost of building in the physical world, managed.
Key takeaways
- Change orders come from three structurally different sources — scope not drawn, conditions not visible, interpretation not controlled — and a single "overruns" bucket hides which one is hurting you.
- Late scope decisions get priced without competition at 15–20% margins; freeze dates and clinical walkthroughs before bid move those decisions back into competitive pricing.
- In second-generation space, cheap investigation — panel load study, slab scan, exploratory demo — is the highest-leverage dollar in the budget, backed by a 5–10% construction contingency.
- You can build exactly to the permitted design and still get corrections; plan check and field inspection are different people applying interpretable code, so manage with engagement and an allowance, not with the assumption you caught everything.
- Once on site: every CO priced from contract rates before work proceeds, one decision-maker on a 48–72-hour clock, a running log against contingency, and a ledger tagged by source.
Frequently asked questions
How much do change orders typically add to a clinic build-out?
Plan for change orders and field conditions to consume a 5–10% construction contingency, with second-generation space toward the high end, plus an owner's contingency for scope you add. The bigger variable is pricing: a change ordered mid-construction carries 15–20% margins versus the 2–3% in a competitive bid, so when a decision gets made matters as much as what it costs.
Can I get corrections from an inspector if I built exactly to the approved plans?
Yes. Plan check approves drawings; the field inspector inspects the built condition, and code leaves genuine room for interpretation on things like accessibility clearances, fire-stopping details, and rating continuity. The field ruling governs. Manage this with a pre-construction meeting, early and frequent inspections so issues surface at rough-in rather than behind finished drywall, and a budget and schedule allowance for corrections.
How do I reduce change orders on a healthcare build-out?
You reduce, not eliminate. Freeze scope decisions at design milestones and walk clinical leadership through the plans before bid, so changes get priced competitively. Investigate the building before you're exposed — panel load study, slab scan, exploratory demo. And choose a contract structure that puts unforeseen conditions against governed contingency rather than marked-up change orders.
Should I pay a change order before the work is done?
The sequence that protects you is pricing before work, payment per the contract's normal application cycle. Require every change order itemized from the contract's labor rates and unit costs, with schedule impact stated in days, and approved by your designated decision-maker before the work proceeds. Letting work run ahead of pricing converts a negotiation into a dispute.
What's the difference between a change order and a contingency draw?
A change order changes the contract sum — new scope, new price, contractor margin included. A contingency draw spends money already inside the budget, and under a GMP it's visible in the open book and governed by whatever approval rights you negotiated. The same surprise behind a wall can be either one, depending on your contract structure — which is why the structure decision matters before the first wall opens.
Mid-Build, or About to Be?
I help founders set up the design freezes, building investigation, and change-order discipline before the exposure starts — and get control of the log when a build is already running hot.
Schedule a Conversation