You bid the job in March. You’re framing in August. Somewhere in between, the number you wrote down stopped describing the job you’re actually building.
Nonresidential construction input prices were running about 3.3% year over year in December. By May they were at 8.4%. Steel, aluminum, diesel. Tariffs and supply disruption pushed all of it, and the pressure is only now working its way into bids as older contracts close out and new work gets priced.
If you’re mid-job on something you priced in the spring, you’re eating some of that. The question is how much, and what you do about the next one.
One thing up front: contract language is attorney territory, and this isn’t legal advice. What follows is how these clauses work in practice so you can have a better conversation with yours.
What an escalation clause is
A cost escalation clause says that if the price of specified materials rises more than an agreed amount between the contract date and the purchase date, the contract price adjusts.
Every workable version of one answers four questions:
What’s covered. Usually named materials: lumber, steel, copper, roofing, drywall. Blanket “all materials” clauses tend to get negotiated out, because an owner can’t underwrite an unbounded risk. Naming three or four volatile items is far more likely to survive the redline.
How the increase gets proven. Either an objective index (the BLS producer price index for that commodity is the common one) or documented supplier quotes. Index-based is cleaner and less arguable. Quote-based is more accurate to what you actually paid. Some contracts use both, index as the trigger and invoices as the measure.
Where the threshold sits. Almost nobody writes a clause that fires on the first dollar. A typical structure is a threshold, say 5%, where you absorb everything under it, and above it the increase gets shared or passed through. This matters more than the existence of the clause. A 10% threshold in a year when your worst commodity moved 8% is a clause that does nothing.
How long the price holds. Bid validity periods are the quiet half of this. A proposal that says “pricing valid 90 days” is doing a lot of work. Thirty is more honest in a market moving this fast, and residential clients are more used to short validity windows than most GCs assume.
What it doesn’t cover
Two gaps catch people, and both are bigger than the material delta itself.
Labor. Escalation clauses are almost always written around material commodities, because those have public indices. Labor doesn’t work that way. There’s no clean index your client will accept for what your framer now charges. And labor is where a lot of the 2026 pressure actually lives: the industry is running roughly 349,000 workers short this year, and that shows up as higher sub pricing and thinner crews. Your material clause won’t touch it.
Time. This is the one nobody frames as a pricing issue at all, and it’s the biggest.
Duration is exposure
Here’s the mechanic that gets missed.
You don’t buy a job’s materials on day one. You buy them across the whole job, as each scope comes up. Lumber goes on the truck early. But the tile isn’t ordered until the substrate is ready. Fixtures come late. Trim packages, appliances, countertops. All of it gets purchased weeks or months after you signed.
Which means every material you haven’t bought yet is priced in whatever month you finally buy it.
Now put a delay into that. Framing slips nine days in June. That doesn’t just cost you nine days of general conditions, though it costs you that too. It also pushes every downstream purchase nine days later into an escalating market. On a twelve-week job in a flat-price year, that’s noise. In a year where input costs are climbing 8% annually, three weeks of slip is roughly half a percent on everything you hadn’t bought yet, and that comes straight off your margin because your contract price didn’t move.
Run that across a season. Four jobs, each running two to three weeks long, each with 60% of the material spend occurring after the slip. That’s not a rounding error. That’s the difference between a good year and a flat one, and it never appears on any line item because it isn’t a cost overrun. It’s the same materials, bought later.
In a stable market, schedule slip costs you general conditions and crew idle time. In an escalating one, it costs you general conditions, crew idle time, and the delta on everything still unpurchased.
The schedule quietly became a pricing document somewhere around the 6% mark.
Four things that actually protect margin
Shorten bid validity. The cheapest change on this list, and the one you can make this afternoon. Thirty-day pricing instead of ninety. Clients push back less than you expect, especially if you explain why in one sentence.
Name your three worst commodities in the contract. Not everything, just the ones that move. A specific, narrow clause tied to a public index gets signed. A broad one gets deleted.
Buy the volatile scopes early where storage allows. Locking price on the two or three items with the most movement, when you have somewhere to put them, converts a market risk into a storage problem. Storage problems are much easier problems.
Protect the sequence. This is the one that isn’t in anybody’s contract template, and it’s the one you control completely. Every week the job runs long is a week of price exposure on unbought material. Which means the ordinary work of keeping the job moving, knowing which delays actually push the finish date and seeing the downstream effect the day it happens instead of on day nine, is margin protection, not just operations.
That last point is worth being concrete about. When a sub slips two days, the useful question isn’t “how annoying is this.” It’s whether that task is on the critical path, because only path tasks push your finish date and therefore push your purchase dates. A slip with float costs you nothing and buys you no exposure. A slip on the path moves every remaining purchase further into an escalating market. We walked through how to find that path on a real remodel here. It’s the difference between reacting to every delay and reacting to the ones that cost you.
The conversation with the client
The reason most GCs don’t ask for escalation language isn’t ignorance. It’s that raising it feels like leading with an excuse before you’ve done any work.
It lands better when it’s framed as the alternative to what you’d otherwise have to do, which is pad. Every GC pricing in a volatile market is carrying contingency in the number, and the client is paying for it whether the increase materializes or not. An escalation clause is the honest version: a lower base price, with a defined mechanism if a named material moves more than a stated amount, backed by an index either party can look up.
Most residential clients understand that trade. They’ve watched grocery prices for three years. What they don’t tolerate is a surprise, a number that changes with no mechanism behind it and no warning.
Which brings it back to the same place as everything else on this list. A price adjustment your client saw coming, tied to a rule you agreed to in advance, is a business conversation. The same adjustment delivered in month three with no prior mention is a fight, and often a lost referral.
What Relay does about this
Relay doesn’t price your materials and it doesn’t write your contracts. What it does is close the gap between a delay happening and you knowing what it cost. Move a date and the cascade runs immediately, so you can see whether the slip touched the critical path and pushed every remaining purchase further into the market, or whether it landed on float and changed nothing. That’s the difference between finding out on day one and finding out on day nine, and in a year like this one those two days are priced differently.
The short version
Get the clause. Keep it narrow and index-based. Shorten your bid validity. Buy the volatile stuff early when you can store it.
Then protect the schedule like it’s a financial control, because in this market it is. The full cost of a schedule slip was already ugly when it was just general conditions and idle crew. At 8.4% input inflation, every week you run long also re-prices whatever you haven’t bought yet.
You can’t control tariffs. You can’t control what steel does. You can control how long the job stays open, and right now that’s worth more than it’s been in years.
Relay shows you the downstream effect of a delay the day it happens, not on day nine. Try it free for 30 days on the Complete plan at relayconstruct.com. Card up front, $199.99/month when the trial converts.