Lost profits claims fail more often on method than on merit. Courts do not reward the biggest number; they reward the number a party can trace, line by line, back to a defensible model of what would have happened without the wrongful conduct. That model is usually called the but-for scenario, and everything in a lost-profits calculation flows from how carefully it is built.

This guide walks through the standard structure: establishing the but-for baseline, isolating the effect of the wrongful conduct from every other market force, netting out costs the business avoided by not earning that revenue, and applying mitigation. It is an educational explainer of how these calculations are typically built, not legal or accounting advice, and not a substitute for a qualified damages expert.

Before you can say what was lost, you need a credible picture of what would have happened absent the breach, tort, or unfair conduct at issue. The most persuasive baselines use the business's own historical performance — a clean trend line from periods unaffected by the dispute — projected forward using the same growth assumptions the business was actually using before the disruption.

Where a clean historical trend does not exist (a new business, a volatile market, a seasonal product), analysts fall back to comparable businesses, industry benchmarks, or a yardstick approach using a similar company that was not affected by the same conduct. Whichever method you choose, the baseline must be built the same way for the full damages period, before and after the harm, so the comparison is apples to apples.

A drop in revenue is not proof of causation. Between the alleged conduct and the observed shortfall sit competitors, pricing changes, macroeconomic conditions, seasonality, management decisions, and simple bad luck. A defensible lost-profits calculation accounts for these confounding factors explicitly rather than attributing the entire gap to the defendant.

The claim is for lost profits, not lost revenue, and the difference is where many first-draft damages models go wrong. Every dollar of revenue the business did not earn also came with costs it did not incur — materials, variable labor, shipping, sales commissions, and any other cost that scales with volume. Those avoided costs must be subtracted from the lost revenue figure.

Fixed costs the business would have paid regardless (rent, salaried overhead, insurance) are generally not subtracted, because the business bore them either way. Getting the fixed-versus-variable split right, and being able to show your work with the company's own cost accounting, is usually the single biggest driver of whether a lost-profits number survives a Daubert-style challenge to the expert's methodology.

Most jurisdictions require a plaintiff to mitigate damages — to take reasonable steps to reduce the harm rather than let it compound. A defendant will often argue the plaintiff could have replaced lost sales through a new customer, a different market, or a pivot in product mix, and any such offsetting gain typically reduces the claim.

Courts also generally require lost profits to be proven with reasonable certainty, not mathematical precision. That standard tolerates some estimation, but it does not tolerate speculation. A model built on a single optimistic assumption, with no sensitivity to alternative inputs, invites exactly the challenge that reasonable-certainty doctrine exists to screen out.

The final figure should be traceable backward: from the bottom-line number, to the years and periods it was built from, to the underlying financial records. Analysts who can walk a jury or judge through that chain in plain language tend to fare better than those who present only a conclusion.

A structured simulation is useful here for a different reason than persuasion: it lets you see how sensitive the final number is to each assumption before opposing counsel does. Moving the baseline growth rate, the causation window, or the avoided-cost ratio one at a time shows which input the whole claim actually rests on.