Every repair estimate has two parts: the part you can measure and the part you’re guessing at. You can measure the cracked drywall, the failed section of fence, the leaking fixture. You can’t measure what’s behind it until you open it up, and by the time you open it up, you’ve already quoted a price. This is the gap that turns profitable repair work into break-even work, and it’s different from new-build estimating in a way a lot of contractors never fully separate out.
Repair work and build work are priced differently, even when they look similar
On a new build or a full replacement, you’re pricing known quantities: a defined amount of material, installed a defined way. On a repair, you’re pricing a diagnosis, and diagnoses are wrong more often than anyone likes to admit. A water stain on a ceiling might be a failed gasket or it might be three months of slow leak into the joists above it. A sagging gate might need a hinge or it might need a post reset. The visible damage tells you where to start looking, not what you’ll find.
Contractors who treat repair estimates like build estimates, one firm number for the whole job, are the ones eating cost overruns or fighting with customers over invoices that grew after the fact. Neither outcome is necessary. The fix isn’t better guessing. It’s pricing the uncertainty on purpose instead of pretending it isn’t there.
Three ways to price the unknown
There are three honest methods, and which one fits depends on how much of the job is visible versus hidden.
Fixed price with a discovery clause. You price exactly what you can see and inspect, and you attach a clause stating that concealed damage found once work begins will be documented, priced, and approved before you proceed. This is the right call when most of the job is visible and only a small piece is a genuine unknown, a single suspect area behind a wall, one section of subfloor you can’t access without demo.
Allowance pricing. You carry a stated dollar amount in the quote for a specific unknown, “$400 allowance for subfloor repair if needed,” and true up at the end based on what’s actually used. This works well when the customer wants one number to plan around but you both know part of that number is a placeholder.
Range pricing with a not-to-exceed ceiling. You quote a range, say $2,200 to $3,400, based on best case and worst case, and commit to communicating before you’d exceed the top of it. This suits jobs where the unknown is large relative to the job, older homes, water damage, anything where “how bad is it” is genuinely a coin flip until you’re inside it.
Pick one method per unknown, not one method per job. A single estimate can carry a fixed price for the visible scope and an allowance for the one area you can’t inspect. Mixing them is normal and it’s more honest than forcing everything into a single number.
Writing a discovery clause that actually holds up
A discovery clause is only as strong as its specificity. “Additional charges may apply” is not a discovery clause, it’s a sentence customers ignore until they’re angry about it. A discovery clause that protects you says what’s included, what would trigger additional cost, and what happens next. Something like: “This price covers repair of the visible damage described above. If removal of the [drywall, decking, flooring] reveals additional damage to structural or hidden components, we will stop, document it with photos, provide a price, and get your approval before continuing.” That sentence does three jobs at once: it sets the boundary of the original price, it commits you to proof instead of just an invoice line, and it puts the pause-and-approve step in writing before anyone’s standing in a customer’s living room with bad news.
The sequence that makes discovery clauses work in practice
The clause only protects you if you follow it every time, including the times it’s inconvenient. When you find the hidden problem: stop before you fix it, photograph it while it’s still exposed, price the addition using your normal rates rather than a number pulled from the moment, and get approval before the work continues. Skip the stop, and you’ve quietly converted a written protection into a verbal agreement, which is no protection at all. The photo matters more than people expect. A picture of rotted framing or a corroded pipe ends the price conversation before it starts, because nobody argues with what they can see.
Set the expectation before you’re standing in the wreckage
The best time to have the “what if we find something” conversation is at the estimate, not mid-demo. One line in the quote does most of the work: “This price is based on visible conditions. Hidden damage, if found, will be priced separately and approved before we proceed.” Customers who hear this upfront treat a later discovery as expected professionalism. Customers who don’t hear it until the invoice treat it as a bait and switch, even when the work itself was necessary and fairly priced. Same discovery, same price, completely different reaction, and the only variable is when you said it.
Repair estimating rewards contractors who separate what they know from what they’re guessing, price each part honestly, and keep a paper trail when the guess turns out wrong. None of that requires better X-ray vision into other people’s walls. It requires a process you run the same way every time, whether the job is $600 or $16,000.
Keeping that process consistent is mostly a paperwork problem, and it’s one of the things we’re building into Punchlist for independent contractors and small crews: a cost catalog to price additions fast, discovery-clause language built into quotes, and change orders you can photograph, price, and send for signature from the job site. Join the waitlist if you want it when it ships.