by Provision
Push back first on the three risk-shifting clauses named above — consequential damages, indemnification, and limitation of liability — and triage everything else by how much exposure it actually moves. The reason to start there is time: you rarely get to negotiate a contract line by line before a tender closes, so the clauses that can detach your liability from the value of your work deserve your scarce hours. A flagged insurance schedule is verification you can clear in minutes; a one-way consequential damages waiver is a decision that can outlive the project.
Treat the flagged-clause list as a triage queue, not a checklist. The more a term exposes you to unbounded downside, the higher it goes; mechanical, verifiable terms drop to the bottom and usually just get confirmed.
How often contractors reject (refuse to accept as written) each category of legal clause, based on aggregated feedback on Provision's platform:
The split is clear. The top three terms cluster tightly in a 23–25% rejection band because they share one root cause: each one converts a bounded, fixed-price scope into potentially unbounded liability. The bottom three sit below 8% because they are routine and verifiable. Push back accordingly.
Push back on any clause that makes you liable for the owner's consequential or indirect damages, because it is the single most-rejected term in construction contracts — refused roughly one in four times it is flagged. A one-way consequential damages waiver exposes you to the owner's lost profits, lost revenue, lost rent, financing costs, and downstream business losses. None of those track the value of your actual work, which is why they are so dangerous: a routine delay or defect can generate a liability many times larger than the contract itself.
The standard ask is a mutual waiver of consequential damages — both parties waive indirect and consequential losses against each other. If the owner insists on carving out specific damages (liquidated delay damages, for instance), keep that carve-out narrow and defined. Never sign an open-ended, one-directional waiver that leaves your exposure to the owner's business losses undefined.
Push back on indemnification clauses that are broad, one-way, or include a duty to defend, because indemnity is rejected nearly 23% of the time and is one of the most expensive terms to get wrong. The language to watch for: an obligation to indemnify the owner for the owner's own negligence, a "duty to defend" that triggers your costs before fault is ever established, and sweeping third-party obligations that reach beyond the work you actually control.
Negotiate toward mutual, fault-based indemnity — each party covers losses to the extent of its own negligence — and cap the obligation to bodily injury and property damage rather than every conceivable claim. In jurisdictions with anti-indemnity statutes, indemnifying another party for its sole negligence may be unenforceable anyway. Treat any "duty to defend" language as a separate, deliberate decision, not a detail to skim past.
Push back whenever a contract caps your liability at an unreasonable number — or fails to cap it at all — because uncapped exposure is rejected almost a quarter of the time. A missing cap is not a neutral omission; it means your total downside is undefined, the opposite of how a fixed-price contract should allocate risk.
Ask for a liability cap tied to the contract value or your fee — a multiple of the contract price, or total fees paid, is common. Watch for caps riddled with carve-outs (for indemnity, IP, confidentiality, "gross negligence"), because enough exclusions can quietly swallow the cap and leave you effectively uncapped. Hold firm that your maximum exposure must be a known, bounded number before you sign.
Push back on warranties that auto-renew, extend indefinitely, or promise "lifetime" performance, because evergreen warranties are rejected at more than double the rate of standard ones — about 17% versus 7%. A standard one-year workmanship warranty is convention and rarely worth contesting. The problem starts when the period quietly becomes perpetual, rolls over automatically, or attaches "lifetime" obligations that outlive the contract, the project, and sometimes the company.
Anchor the warranty to a defined, finite period that matches the work and the applicable standard. If the owner wants extended coverage on specific systems, scope it precisely — this component, this duration — rather than accepting an open-ended commitment you cannot price, insure, or eventually walk away from.
Push back on liquidated damages that are not a genuine pre-estimate of the owner's loss, because they are rejected about one in six times. Liquidated damages for delay are normal, but the figure has to bear a real relationship to the harm a delay would actually cause. Contest punitive daily rates, damages with no cap on the total that can accrue, and overlapping "dual" penalties that charge you twice for the same delay.
Negotiate a per-day rate that reflects actual anticipated loss and an overall cap on the total assessed. Pair that with fair extension-of-time provisions so you are not penalized for delays outside your control — owner-caused changes, differing site conditions, or force majeure. A clause you can live with is bounded, proportionate, and balanced by your right to legitimate schedule relief.
Push back on intellectual property clauses that assign more than the project deliverables, because IP terms are rejected about 15% of the time. On construction work this surfaces around design contributions, proprietary details, models, or software. The language to flag is a blanket assignment of "all" IP created in connection with the project, which can sweep in your background IP, internal tools, and reusable know-how. Limit any assignment to the project-specific deliverables the owner is actually paying for, and reserve your pre-existing IP with a license rather than a transfer — this matters most for design-build and engineered-systems work, where your reusable methods are a core business asset, not a project byproduct.
You can usually accept insurance, bonding, and standard warranty clauses, because contractors approve them more than 90% of the time. These terms are mechanical and verifiable rather than risk-shifting, so contesting them mostly burns leverage you will want for the clauses above.
The lesson cuts both ways: clauses with an established market standard get accepted; clauses that shift uncapped risk get pushed back on. Spend your negotiation capital where it changes your exposure.
These clauses get pushed back on because they share one root cause: they convert a fixed-price scope into open-ended financial risk. Consequential damages, indemnity, and liability caps are not rejected because contractors are difficult — they are rejected because each one can detach your liability from the value of the work you agreed to do. A $500,000 job should not carry the risk of a multi-million-dollar loss buried in a one-way waiver.
That also explains why a one-size-fits-all "risk score" misleads you. A clause rejected 25% of the time and one rejected 4% of the time are not equally urgent, and treating them the same wastes the scarce hours you have before a tender closes. Prioritize the consequential-damages waiver over the boilerplate insurance schedule — that is where real negotiation, and real exposure, lives.
Push back in writing, early, and with a specific alternative rather than a flat refusal, because owners respond to redlines that solve a shared problem. Three habits make pushback land:
Figures are drawn from anonymized, aggregated clause-level feedback submitted by construction contractors reviewing real contracts on Provision's platform — a dataset spanning more than 100,000 feedback items across many projects and companies. Each clause was counted once using its final review status (accepted versus not accepted), and clauses were grouped into categories by keyword. Insurance coverage requirements were analyzed separately from contractual liability caps so routine insurance approvals did not understate how often liability caps are actually contested. Rejection rate = clauses not accepted divided by clauses with a final accept-or-reject decision.
What is the most rejected clause in construction contracts?
Consequential and indirect damages waivers, rejected about 25% of the time — the highest of any legal clause category measured across the feedback data.
What should I push back on first in a construction contract?
The three risk-shifting terms: consequential damages waivers, indemnification language, and limitation-of-liability caps. They convert a fixed-price job into open-ended exposure and are rejected far more than any other clause type.
What contract clauses do contractors accept most easily?
Insurance coverage requirements (about 96% accepted), standard warranties (about 93%), and bonding or surety provisions (about 93%). They are verifiable and standard, so there is little to negotiate.
Why do contractors reject indemnification and liability clauses so often?
Because they shift open-ended, hard-to-cap financial risk onto the contractor — turning a fixed-price job into potentially unlimited exposure, often for losses unrelated to the value of the work.
Should I push back on insurance and bonding requirements?
Usually not. They are accepted more than 90% of the time because they are mechanical and verifiable. Confirm the limits and bond amounts are achievable, then save your negotiating leverage for the risk-shifting clauses.
This article is general information, not legal advice. Contract terms and their enforceability vary by jurisdiction, project, and contract form. Consult qualified legal counsel before negotiating, redlining, or signing any construction contract.
See Risk Review flag onerous terms, hidden obligations, and inferred-scope traps in minutes.
See Risk ReviewMore Articles