Settle or Fight

Why this matters

The question arrives phrased wrong. Owners ask whether they are right, and rightness is not the variable that decides this. What decides it is what you can prove with the records you already have, what the loss actually costs including the other side's fees, and how much you have to spend before anyone learns anything new.

Settling a claim you believe is meritless feels like paying a bully, and sometimes it is the correct commercial decision. A shop that makes it deliberately, with the arithmetic written down, recovers in a week. A shop ashamed of the same decision makes it late, after spending the money that would have bought a better deal. This card is orientation, not legal advice about your matter, and the merits assessment at the centre of it is the one judgment you should not make alone.

The comparison that actually decides it

Not the claim against your sense of justice. The hours to reach the next checkpoint, against what that checkpoint can realistically change.

Convert the claim into lawyer-hours first: amount in dispute divided by your counsel's hourly rate. That unit and the phase-by-phase hour stack behind everything below belong to a sibling card and are not re-derived here. See related: What Litigation Actually Costs and How to Control It, which puts written discovery at roughly 30 to 70 lawyer-hours and depositions at 15 to 45 in a straightforward two-party dispute.

The word doing the work is "realistically". A phase that will surface the other side's weak documentation can change the number a great deal. A phase that will surface YOUR thin file changes nothing except who else has read it. That is why the merits axis is not a separate consideration to weigh alongside cost - it is the multiplier on the cost comparison.

Run the decision at checkpoints rather than once. After the pleadings, after written discovery, after depositions, after a dispositive motion is briefed. Each is a place where the case is worth re-pricing because something is now known that was not known before.

Axis one: what you can prove, not what is true

Separate two sentences that feel identical and are not. "We did that work correctly" is a belief about the past. "We can show a judge we did that work correctly with what is in the file today" is a claim about your records, and only the second one has any value in a dispute.

Walk the file before you talk about merits at all. A signed work order with the scope on it, dated photos from before and after, a signed change order for every deviation, job notes written at the time rather than reconstructed, a completion sign-off, dispatch records that put your people where you say they were, and the technician who did the work still employed and still able to remember the visit.

Each missing item moves you down a grade, and two of them move you a long way. A departed tech and no photos is not a strong case with a paperwork problem; it is a weak case, because the other side's account of the visit will be the only account anyone can produce. The most expensive error in this whole subject is an owner who is genuinely right, assumes that will be visible, and funds a discovery phase whose entire output is proof that it is not.

Axis two: the exposure is not the claim

What you lose if you lose is the claim plus three things that are invisible on the demand letter.

Fee-shifting. The default in United States courts is the American Rule, each side paying its own fees. A prevailing-party clause in your own contract, or a statute that shifts fees one way to a claimant, overrides it. Read your own contract form today, because it takes two minutes and it changes the whole calculation. The mechanics belong to the cost card.

Statutory multipliers. Many state consumer-protection acts allow doubled or trebled damages on a successful claim, and several also carry the one-way fee shift, so both arrive together. Whether your facts fall inside your state's act is a question for a lawyer in that state, but the existence of the multiplier is something you should know in week one, not month six, because it is the single largest term in the arithmetic.

The counterclaim. If you are the one suing on an unpaid invoice, the defendant's answer can convert a bounded receivable into an unbounded exposure. See related: The Counterclaim That Turns Your Collection Into a Defence.

One more that owners skip: a judgment against the business is a public record that shows up in credit checks, bonding applications and prequalification packages for commercial work. Where you bid public or institutional jobs, that is a real cost with no invoice attached.

Axis three: what the next phase costs, and who it costs more

Price the spend to the NEXT checkpoint, never to trial. Trial is the phase everyone imagines and almost nobody reaches, and estimating the whole case produces a number so large it stops the conversation rather than guiding it.

Then ask a question most owners never think to ask: which side does this phase hurt more. Discovery is symmetrical in law and wildly asymmetrical in practice. A shop whose job records live in one system, indexed by job number, with photos attached to the ticket, answers a document request in a few hours of office time. A shop whose records are in three systems and a shoebox pays for the same request twice, once in searching and once in what the gaps look like. If your file is the clean one, the phase you are dreading is the phase that pressures them.

The other three axes, and one of them is the largest

Can they pay. Winning against someone with nothing costs you the whole cost stack and returns a piece of paper. Run this check before you spend, not after you win. See related: The Other Side Has No Money and What That Changes.

Precedent with a customer base that talks. Residential shops overstate this and commercial shops understate it. One homeowner's refund does not travel unless your market is a single subdivision or an HOA. A concession to a property manager holding forty doors travels the same afternoon, to the other managers in the same portfolio. Where the concession would travel, buy confidentiality as a term of the deal rather than declining the deal.

Your own hours. The largest line and the only one with no invoice. Assembling records, preparing for and sitting a deposition, and attending hearings takes an owner out of revenue work in blocks that cannot be moved around a busy season. Count them explicitly. This is the same calculation you already make when you decline a job that pays and eats the crew.

Two cases, one frame, opposite answers

Case A, an irrigation shop. A homeowner sues over a backflow assembly installed the previous spring, claiming the work was defective and the price misrepresented, under the state's consumer-protection act. Claim divided by counsel's hourly rate: 40 lawyer-hours. That act carries treble damages and one-way fees to the consumer.

Merits: the shop is confident the install was to code and the inspection passed. The file holds the invoice and a permit number. No photos. The installing tech left in the autumn and his phone was reissued. The customer has an eight-message text thread with that tech, which she still has and the shop does not.

Exposure: 40 hours trebled is 120 lawyer-hours, plus her fees, which on the same phase table run 44 to 107 hours cumulatively from the pre-suit letter through written discovery, rather than the 30 to 70 that one phase costs on its own. At the bottom of that range the downside is 164 hours, which is at least four times the claim.

Next checkpoint: the close of written discovery, call it 45 lawyer-hours against the sibling card's 30-to-70 range for that phase. That alone exceeds the 40-hour claim, and it buys the shop nothing, because the discovery output is the shop's own thin file plus a text thread produced by the other side. The phase cannot improve the position; it can only publish it.

Answer: settle now, and settle in a shape that costs the shop labour rather than cash. Offered: pull and reinstall the assembly with a licensed tech and a permit, photograph everything, and extend the workmanship term on that assembly, against a mutual release and a confidentiality term. The cash component is small because most of the value transferred is work the shop performs at its own cost, which is the cheapest currency it has.

Case B, a commercial mechanical shop. A general contractor withholds payment and backcharges for delay on a tenant fit-out. Claim divided by the hourly rate: 300 lawyer-hours. The subcontract carries a mutual prevailing-party fees clause.

Merits: every change is on a signed change order. Daily reports name the days the shop's crew stood down waiting on ceiling access, and the GC's own superintendent signed two of them. Foreman still employed. The file is already indexed by date.

Next checkpoint: the same 45 lawyer-hours of written discovery. Against a 300-hour claim that is 15 percent, and this is the asymmetry case - the shop produces its file in a day, and the request going the other way asks the GC to produce a schedule record it has reason not to want read.

Fee-shifting runs the other way here. The clause is mutual, so the GC is betting its own fees on a documentary record the shop holds and it does not.

Answer: fund the next phase, and put a mediation offer on the table in the same week, because mediation is cheap relative to any phase in the table and costs nothing but the day if it fails.

Same frame, opposite answers, and the merits axis is what flipped it. Case A's 45-hour spend is 113 percent of its claim and buys nothing; Case B's identical 45-hour spend is 15 percent of its claim and buys leverage. Neither answer came from who was in the right.

What a settlement is made of besides money

Owners negotiate the number and accept whatever terms arrive around it, which is backwards. The terms are where a trade dispute finds the cheap resolution.

Work in kind. A redo, a replacement component, a credit against future service, an extended workmanship term. You are paying in labour and parts at your cost rather than in cash at retail, which is the single largest discount available in any settlement you will ever sign, and it is available almost only in the trades.

A payment schedule with security. Where you owe money you cannot pay at once, instalments plus a stipulated judgment that only enters if you default is a structure many defendants will accept. Whether your state permits that shape, and in what form, is a question for counsel.

Scope of the release, and mutuality. A one-way release leaves your own claims alive and theirs extinguished, or the reverse, and which direction the document runs is not always what the covering email says. The document that ends the dispute has its own card and every term in it does work. See related: The Settlement Agreement and the Release You Sign.

No admission, and confidentiality. Standard, worth having, and both are narrower than owners think.

Here is where you stop and hand it over. You can decide whether to settle. You should not draft or sign the release, agree the tax treatment of the payment, or accept a confidentiality clause without counsel reading it, because a release that is broader or narrower than you believe is the one mistake in this subject with no second chance. Walk in with the claim, your file index, your contract form with the fees clause marked, and the number you have decided you will pay.

References

• The American Rule on attorney fees and its contractual and statutory exceptions; state consumer-protection acts that add multiplied damages and one-way fee shifting
• See related: What Litigation Actually Costs and How to Control It, which owns the cost curve and the lawyer-hours unit used throughout this card
• See related: The Other Side Has No Money and What That Changes, The Counterclaim That Turns Your Collection Into a Defence
• See related: The Settlement Agreement and the Release You Sign, Mediation Arbitration and Court Compared