Before-and-after projects the plaintiff's results before the breach; the yardstick measures what a comparable business or market did. The best models combine them and are tested on the months just before the breach.
In the worked dispute, a growth rate of 6.31 percent from the plaintiff's own history predicted the two months before the breach at $3,555,187; actual sales were $3,251,000, 8.6 percent lower. A market-and-share model built on an industry index predicted $3,240,597, within 0.3 percent.
Before-and-after: project the plaintiff's results from a base period before the breach through the damages period, and compare with actual results.
Example 1. Sales of $21,600,000 in 2022 and $24,410,000 in 2024 give a compound growth rate of 6.31 percent. Applied to 2024, it projects $25,950,271 for the year of the breach. It is simple and rests on the plaintiff's own record, which courts like. It also assumes the market would have stayed as kind as it was in the base years.
Yardstick: measure the but-for world from something comparable the breach did not touch: a similar business, the same business in another region, or the market.
Example 2. Divide the plaintiff's sales by an industry index to get its sales a point of the index: $216,000 in 2022, $227,070 in 2024, a share trend of 2.53 percent a year. In the year of the breach the index fell from 107.5 to 101.6. But-for sales are 101.6 × $227,070 × 1.0253 = $23,653,991, $2,296,280 less than the growth-rate projection.
Worked in full in The Lost Profits Damages Handbook by Julian R. Sterling.Get the DRM-free PDF, $19.99 →or the paperback
Free: the lost profits damages calculator in Excel, from lost sales to damages at the trial date, with its checks.
The months just before the breach are the only months in which the but-for world and the actual world are the same. A model that cannot predict them has no claim on the months after. In the example the growth-rate model missed by 8.6 percent; the market-and-share model by 0.3 percent. A back-test on an earlier calm year could not tell them apart: both came within 0.5 percent of 2024.
Over the whole claim, the growth-rate model finds $18,831,336 of lost sales and the market-and-share model $10,702,829. Holding the share flat instead of continuing its trend would lower the supported damages by $1,620,925: the share trend is the model's soft spot, and the evidence for it (accounts won before the breach, a documented pipeline) deserves a page of the report.
A yardstick must cover the plaintiff's market, be measured the same way (dollars against dollars, or a price adjustment), and be shown to move with the plaintiff before the breach. Without that showing, a court may find an analytical gap between the data and the opinion and exclude it under Rule 702.
The trap: a growth rate measured in rising years projected through a falling market. It attributes the market's decline to the defendant.
Figures checked on 18 November 2026. The dispute in the examples is fictional and its figures are teaching assumptions, not market norms or rules of law. Not legal advice.
None by rule. Courts look for a method grounded in evidence and reliably applied. Before-and-after is common when the plaintiff has a stable history; a yardstick helps when the market moved or the business is new.
Yes. A market-and-share model uses the plaintiff's own history for its share and an outside index for the market, which keeps the persuasive part of before-and-after and adds the fact no plaintiff controls.
Long enough to show the normal pattern and recent enough to reflect the current business, often three to five years. Test the result with one year added or removed.
This article is one step from The Lost Profits Damages Handbook, a practitioner’s guide to measuring commercial damages: the but-for world, the yardstick, incremental margin, mitigation, cover, prejudgment interest, discounting and the expert report under Rule 702, worked on one dispute from the demand letter to the trial date.
DRM-free PDF, $19.99 · Paperback, $32.99, shipped worldwide · Free Excel calculator