Lost-profits damages are often presented as a mathematical exercise: forecast the revenue a business expected to earn, subtract the revenue it actually earned, and calculate the difference. That framing is incomplete. The most important work usually occurs before any formula is entered into a spreadsheet.
A credible analysis must construct a supportable “but-for” world—the economic position the plaintiff would likely have occupied absent the alleged wrongful conduct. That requires more than an optimistic forecast. It requires a clear theory of causation, reliable evidence, and assumptions that can withstand examination from opposing counsel, the court, and other experts.
Causation comes before calculation
An expert may be able to quantify a loss, but the calculation is only meaningful if it measures losses connected to the conduct at issue. A decline in sales occurring after an alleged breach does not necessarily mean the breach caused the entire decline. Competitor activity, customer attrition, supply constraints, pricing decisions, regulatory developments, seasonality, or a broader economic slowdown may have contributed.
For that reason, a sound damages model should distinguish correlation from causation. The analysis should identify when the alleged harm began, which customers or product lines were affected, and whether unaffected parts of the business provide a useful benchmark. Contemporaneous budgets, customer correspondence, sales-pipeline records, contracts, operating data, and testimony can be more persuasive than a forecast prepared after litigation begins.
The transaction or opportunity must be supportable
Claims involving a lost contract, prospective customer, licensing arrangement, or new business opportunity require careful attention to whether the opportunity was reasonably likely to occur. A damages model cannot convert an uncertain negotiation into a completed transaction merely by applying a discount rate.
The Arizona Supreme Court’s 2025 decision in McAlister v. Loeb & Loeb illustrates the point. The court rejected a lost-profits theory tied to prospective licensing transactions because material deal terms remained unresolved. The lesson extends beyond that case: experts should examine the evidence supporting both the existence of the opportunity and the amount of profit allegedly lost.
Revenue is not profit
Another common error is treating lost sales as lost profits. If sales would have increased in the but-for world, the company may also have incurred additional labor, materials, commissions, shipping, marketing, working-capital needs, or other incremental expenses. Those avoided costs must be considered.
The classification of costs deserves particular care. An expense labeled “fixed” in the general ledger may become variable over a longer damages period. A business near capacity might have needed another shift, additional equipment, or expanded facilities to generate the projected revenue. Conversely, some overhead may not change at all. The analysis should reflect economic behavior, not simply accounting labels.
Mitigation and post-event evidence
A plaintiff generally cannot ignore reasonable opportunities to reduce its loss. Replacement sales, new customers, substitute suppliers, insurance recoveries, and operational changes may affect the damages calculation. Post-event results can also test whether earlier assumptions were realistic, although hindsight should not be used selectively.
A useful model often presents more than one scenario. Sensitivity testing can show how damages change when assumptions about growth, customer retention, margins, or the loss period change. This does not weaken the opinion. It helps decision-makers understand which assumptions drive the result.
A defensible story
The strongest lost-profits analyses align the legal theory, factual record, and financial model. Each major assumption should have an identifiable source, and the path from the alleged conduct to the claimed loss should be understandable without relying on spreadsheet complexity.
Ultimately, precision is not the same as reliability. A model can calculate damages to the dollar and still rest on an unsupported premise. The goal is not to build the most elaborate forecast. It is to present a transparent, evidence-based reconstruction of what most likely would have happened.