The bill that shows up after the switch
In February 2024, a commercial roofing contractor called me at 7:40 on a Tuesday morning. His crew was eleven days out from final inspection on a 42,000 sq ft re-roof. Two weeks earlier, someone in procurement had swapped the specified Johns Manville roofing products for a lower-bid alternative, and swapped the R-13 batts in the wall assembly for a cheaper board from another mill. Estimated material savings: just under $9,000 across the project.
The job got done. It got done fourteen days late, after a re-submittal, a third-party thermal verification, and 31 hours on my end coordinating overnight freight once the swap got reversed. The $9,000 didn't survive contact with reality. It came back as roughly $21,000 in overtime labor, expedited freight, and a partial holdback that took four months to release.
I've coordinated 300+ rush material orders over 9 years for commercial jobs, and this pattern is boringly consistent. Not sometimes. Almost every time a spec changes mid-project to chase unit price, the money comes back — just wearing different clothes.
The assumption nobody pulls the thread on
The assumption goes like this: unit price is the cost, and the brand name is markup. Both halves are wrong, but the second one is more expensive.
Here's the part that gets skipped in the procurement meeting. When you swap from Johns Manville R-13 insulation to a comparable-sounding product from another manufacturer, you don't just change a SKU. You change a spec chain that already passed.
The spec section referenced published thermal performance data — which JM provides in their technical data sheets, along with product-specific installation guidance. The third-party verifier who signed off on the assembly matched the RFI to that data. The submittal was approved on that basis. Your field superintendent ran a mock-up with that product. Now every one of those artifacts is stale.
In most wall and re-roof projects I've touched, the re-submittal and re-verification alone runs 6–11 business days. That's before anyone touches a pallet.
What's actually costing you money
Three things, and none of them show up on the purchase order.
The documentation restart. Every approved submittal that referenced the original product has to be reissued. On a project with multiple assemblies, that's not one document. It's six or eight, multiplied by however many reviewers sit in the chain.
The field-learning curve. Crews who've installed JM board and membranes for years have muscle memory for it — how it cuts, how it fastens, where it wants to pucker. A new product resets that to zero. Not because the new product is bad. Because it's unknown, and unknown costs hours.
The warranty ambiguity. Mixed-material assemblies are a grey area at claim time. If you've got one manufacturer's insulation behind another manufacturer's membrane, and something fails, you've handed both manufacturers a reason to point at the other one. I've watched a $40,000 moisture claim sit in limbo for nine months over this exact issue.
That last one is the part I wish more people understood going in. The product isn't the risk. The gap between products is.
What the delay actually costs, in hours
Let me put real numbers on it, because "it costs more" isn't an argument.
On that February job: fourteen days of schedule slip on a 60-day build. Twelve of those days were re-submittal and verification. Two were freight. The general contractor charged liquidated damages of $900/day for days 6 through 14. That's $8,100 the owner never sees on a scope sheet.
Add the overtime. Add the returns processing on the wrong product (restocking fees on insulation are not small — call it 15% on the returned pallets). Add my team's hours, which the contractor ate at cost.
The $9,000 in nominal savings became a $21,000 problem. Net loss of $12,000, before you count the relationship damage with a GC who now reviews this contractor's submittals line by line.
Looking back, I should have killed that swap the moment I saw the change order land. At the time, procurement had already signed it and I wasn't the one who owned the decision. That's not a good reason. That's just what I told myself.
Where the money actually is
The answer isn't "stick with the original spec no matter what." That's too simple, and sometimes the original spec is genuinely overpriced for the application.
The answer is: separate the two decisions.
Decision one — is this product line right for the assembly? If yes, don't move it. Instead, go after source, lead time, and freight, which are where the real savings usually live anyway.
Decision two — if the product line genuinely needs to change, price the restart before you sign. Re-submittal days. Re-verification cost. The learning curve in crew hours. If the change can't clear all three and still save money, it isn't a saving.
One more thing worth knowing: JM publishes installed-performance data alongside lab data in their technical documentation. If you're comparing against an alternative, compare on installed numbers. Lab numbers and installed numbers diverge, sometimes significantly, depending on the substrate and the assembly. Most side-by-side comparisons I see in the field are comparing the wrong columns.
What was a reasonable procurement rule in 2020 isn't anymore. Not because the arithmetic changed — because the documentation burden on the back end got heavier while unit prices flattened out. The fundamentals of why you specify a product haven't moved. The cost of changing your mind mid-project has.
A few unrelated things people ask me about
I get pulled into questions that have nothing to do with building materials, usually because someone lands here while they're in the middle of a project and just keeps scrolling. So — briefly, and with the usual caveat that this is well outside my lane:
If you're working through a shower valve replacement, the starting point is your existing rough-in and your local code, not the brand on the box. If you're comparing screen protector specs, the number that actually matters is the hardness rating measured against what's scratching your display — most of the marketing numbers are rounded up generously. And if you're trying to figure out how to change wallpaper on Mac, it's System Settings → Wallpaper on Ventura and later, or System Preferences → Desktop & Screen Saver on older builds.
None of those are my specialty. I'm not a plumber, I'm not a display engineer, and I'd rather say that than guess at your pipe diameter.
What I can tell you is that all three share the same failure mode as a mid-project material swap. People pick the option that looks cheapest or easiest up front, then pay for the redo. The redo always costs more than the careful choice would have.
One thing I should add: I've been the guy who skipped the careful choice plenty of times. That's why I write about it.