When a utility patent is infringed, federal law guarantees the patent owner at least a reasonable royalty for the infringer’s use of the invention, and lets the owner instead recover its own lost profits whenever it can prove, to a reasonable degree of certainty, that it would have earned those profits but for the infringement. Those are the two financial measures, and the central job of the damages accountant is to quantify each one cleanly so the trier of fact can choose the measure the evidence supports.
I am Joey Friedman, a CPA, Accredited in Business Valuation (ABV), and forensic accountant, and I work as the economic-damages expert in commercial-litigation matters. The article below explains how the dollars get calculated in a patent case. I want to be precise about my lane from the outset: whether a patent is valid, whether it was infringed, whether the conduct was willful, and whether attorney fees should be shifted are all legal questions that belong to counsel and ultimately to the court. My work begins after those questions are framed. I take the liability theory as the assignment hands it to me, assume for calculation purposes that the patent is valid and infringed, and then measure the economic harm that flows from that assumption.
The Statutory Framework: A Floor, Not a Ceiling
The damages provision for utility patents sets a compensation standard with a built-in minimum. The court is to award the patent owner damages adequate to compensate for the infringement, and that amount can never drop below a reasonable royalty for the use the infringer made of the invention. The statute also provides for interest and costs.
Two ideas inside that sentence drive everything else.
First, “adequate to compensate” is a make-whole standard, not a punishment. The goal is to restore the patent owner to the financial position it would have occupied had the infringement never happened — no better, no worse. Where I sometimes see overreach is an analysis that quietly tries to do more than restore; my discipline is to keep the number tethered to actual economic loss.
Second, the reasonable royalty operates as a floor. Even an owner who cannot prove a single lost sale is still entitled to be paid for the use of its technology. That floor is why, in practice, a royalty analysis is almost always present in a patent case — either as the primary measure or as the fallback when a lost-profits theory does not carry the full sales volume.
A point worth stating plainly: enhanced damages for willfulness (the statute allows the court to multiply the award) and any fee-shifting are determinations for the court, not figures the accountant supplies. I calculate the compensatory base. What the court does with it on the willfulness question is outside my analysis.
For the broader principles that run through every damages discipline, see my overview of economic damages.
Lost Profits: Proving What Would Have Happened
Lost profits answer a counterfactual question: if the infringing product had never reached the market, how much additional profit would the patent owner have earned? That is the same but-for spine that runs through commercial lost-profits damages generally, and the same evidentiary bar — the loss must be shown to a reasonable degree of certainty, not left to speculation.
There is one feature that distinguishes the patent setting. Because the patent grant is a right to exclude, proof of infringement carries a presumption that some damage occurred — the violation of the right to exclude is itself the injury. That presumption gets the owner past the threshold of “was there harm.” It does not, however, hand the owner a number. The amount still has to be built and defended.
A patent owner generally needs to be selling something it lost sales on. The owner does not have to be selling a product that practices the patent claims, but it must have been competing in the market with a product on which the infringing sales caused real erosion. An owner who sits entirely on the sidelines, licensing rather than selling, is typically in royalty territory rather than lost-profits territory.
The Panduit Framework
The most common way to establish but-for causation for lost profits is a four-part test drawn from the Panduit decision. It is not the only accepted method, but it is the workhorse, and each element maps to a question the accountant and the market evidence have to answer:
- Demand for the patented product. Were customers actually buying because of what the patented technology delivers? Demand has to attach to the patented advantage, not merely to the general product category.
- No acceptable non-infringing substitutes. If buyers had a comparable, lawful alternative they would have been content to purchase, the owner cannot assume the infringer’s sales would have flowed to it.
- Capacity to meet the demand. The owner must show it could have actually made, marketed, distributed, and serviced the additional volume — not just that demand existed.
- The profit that would have been earned. This is the dollar calculation: incremental revenue on the recaptured sales, less the incremental costs of producing them.
I walk through each element below, because in my experience cases are won or lost on how rigorously these are supported rather than on how confidently they are asserted.
Factor One: Demand
The first factor sounds simple and is the one most often glossed over. The question is not whether the product sold — it is whether the patented feature is what motivated the purchase. This requires me to separate the value the customer was paying for: how much of the buying decision traces to the patented advantage, and how much to brand, price, service, unrelated features, or other technologies inside the same product.
That distinction matters enormously when the patent covers a single feature of a multi-feature product. If the patented feature is the reason customers showed up, the demand factor is strong. If the patented feature is an obscure detail that buyers did not know about and did not weigh, the owner may not clear the first hurdle at all, even in a market it largely shares with the infringer.
Demand also assumes the two products genuinely compete — that they are close enough substitutes for the same customers that the infringer’s buyers would have migrated to the owner’s product in the but-for world. When the owner sells at the premium end and the infringer sells at the budget end, that assumption breaks down. Buyers shopping the low end would likely have stayed at the low end, not traded up to the owner’s premium offering.
Factor Two: No Acceptable Non-Infringing Substitutes
This factor is really an extension of the demand analysis. It asks whether the market offered buyers a comparable, lawful alternative. A substitute is “acceptable” when it delivers the benefits customers were after, in a similar price range. A product that lacks the very benefit the customers demanded is not an acceptable substitute, and a product priced or specified far apart from the patented product typically does not compete in the same segment at all.
How wide I draw the market boundary affects this directly. Define the market broadly and substitutes proliferate; define it as a narrow niche and they may vanish. The honest approach is to let the actual competitive facts — pricing tiers, feature sets, the reasons customers chose among offerings — set the boundary, rather than drawing the line to reach a desired result.
One related nuance: the analysis can extend to what the infringer itself could have done. An infringer rarely surrenders its whole market position when confronted with a patent; if it had a non-infringing design it could have switched to within a reasonable time and cost, that available alternative belongs in the picture. The economic value of a patent is fundamentally the advantage it offers over the next-best lawful alternative, so an alternative the infringer realistically could have deployed compresses the owner’s recoverable loss.
Factor Three: Capacity
Demand and the absence of substitutes mean nothing if the owner could not have produced and delivered the recaptured volume. The third factor asks exactly that. The larger the lost-sales volume relative to the owner’s historical output, the harder it is to credibly claim the owner could have absorbed it.
Capacity is broader than the factory floor. It includes the ability to distribute, market, service, and finance the additional production. I have seen analyses that confirm a plant could physically build more units while ignoring whether the owner had the sales channel and geographic reach to actually place those units with the infringer’s customers. That distribution question is a genuine constraint, and I treat it as one.
Factor Four: Calculating the Amount
The fourth factor is the arithmetic, and it rests on incremental profit. Lost profit equals the revenue from the recaptured sales minus the incremental costs of generating those sales. The key word is incremental. Fixed and sunk costs — product development already paid for, plant and overhead already in place — do not increase when an established line produces additional units, so they fall out of the calculation. Only the variable costs that genuinely rise with volume — materials, direct labor, and similar — get subtracted.
This incremental-cost discipline is the same one that governs lost-profits work across every case type, and getting it right is where a forensic accountant earns the engagement. Loading fixed costs into the deduction understates the loss; ignoring real variable costs overstates it. The number has to reflect the margin the additional sales would have actually thrown off.
Shortcuts and Add-Ons to Lost Profits
The Two-Supplier Market
When the relevant market has only the patent owner and the infringer in it, the causation analysis simplifies. With no one else to absorb the demand, a court may presume the owner would have captured the sales the infringer made. The presumption does not erase the other requirements — the owner still has to show capacity and still has to calculate the profit — but it eases the burden of proving where the lost sales would have gone. A tightly drawn niche can create a two-supplier market even inside a larger industry, where many competitors exist overall but only two serve a particular high-performance or specialized segment.
Market-Share Damages
When proving the complete absence of acceptable substitutes is too heavy a lift, the owner can sometimes fall back on a market-share approach. Instead of claiming all of the infringer’s sales, the owner claims its proportional share of them — the slice it would have captured given the share each competitor held. This blends the two measures: lost profits on the share the owner would have won, and a royalty on the remainder that other competitors would have taken. It is a pragmatic middle path, but it carries a watch-out: structured carelessly, it can leave the owner better off than the no-infringement world would have, by awarding royalties on sales the owner never would have touched. I test for that.
Price Erosion
Infringement does not only divert units — it can drag down price. If the owner had to cut its price to compete with the infringing product, the difference between the price it would have charged but for the infringement and the lower price it actually realized is recoverable as price erosion. The measure applies to the owner’s own continuing sales, not just the diverted ones.
Price erosion is genuinely hard to prove and, in practice, accounts for a small share of patent awards. The reason is the basic relationship between price and quantity: raise the hypothetical price and, in most markets, some volume disappears. A credible price-erosion claim cannot simply assert the owner would have charged more — it has to account for the sales that would have been lost at that higher price. Markets differ in how sensitive volume is to price (their elasticity), and defendants routinely argue that forces unrelated to the infringement — ordinary market fluctuation, the owner’s own pricing behavior — drove the price down. A defensible analysis confronts those alternatives head-on.
Convoyed and Collateral Sales
Sometimes the patented product pulls along sales of related, unpatented items — accessories sold with it, or follow-on parts and supplies sold after it. These convoyed or collateral sales can be included in lost profits, but only where there is a genuine functional relationship: the unpatented item works together with the patented one as part of a single working unit or result. Items that merely ride along for convenience, or that have a fully independent use and their own standalone market, do not qualify. The discipline is to include what truly travels with the invention and exclude what just happened to be in the same shopping cart.
The Reasonable Royalty: Paying for the Use of the Invention
Where lost profits compensate the owner for sales it would have made, the reasonable royalty compensates it for the value of what the infringer took — the use of the patented technology itself. This is the measure most frequently awarded, and it is the floor the statute guarantees.
If an established royalty exists — a rate the owner has consistently charged others for comparable use, set before the dispute and not under the shadow of litigation — that rate is usually the best evidence, because it removes the guesswork. In practice, true established royalties are rare; the criteria are stringent, and a single license or a litigation-driven settlement generally does not qualify. Most cases therefore turn to the hypothetical negotiation.
The Hypothetical Negotiation
The standard framework imagines a negotiation, struck just before the infringement began, between a willing licensor on one side and a willing licensee on the other, both assuming the patent is valid and infringed, both genuinely trying to reach a deal. The output is the royalty they would have agreed to — one that leaves the licensee a reasonable profit while fairly compensating the patentee.
A few features of this construct shape my work:
- The timing is fixed at the eve of first infringement. That is the date of the imagined bargain.
- But later facts can inform it. Courts allow the analysis to look at how things actually turned out — sometimes called the “book of wisdom” — because real-world results can correct what the parties could only have guessed at the negotiation date. The infringer’s actual sales success, for instance, can be relevant evidence.
- The infringer’s profit margin is not a hard ceiling. A reasonable royalty is not automatically capped at the infringer’s net margin; in some cases a royalty can exceed it, on the theory that the infringer could have raised its price or that the owner’s own profit expectations matter.
The Royalty Base and the Royalty Rate
A running royalty has two moving parts, and both have to be supported:
- The royalty base is the dollar volume (or unit count) the rate gets applied to.
- The royalty rate is the percentage (or per-unit amount) applied to that base.
The two are linked — a base-and-rate combination has to make sense together — and a reliable conclusion requires evidence on each. This is where apportionment, the subject of the next section, does much of its work.
The Georgia-Pacific Factors
The structured checklist most experts use to build a reasonable royalty comes from the Georgia-Pacific decision, which lays out fifteen considerations. I will not recite all fifteen here, because reciting them is not the point — the courts have been explicit that an expert should analyze the factors that matter and tie them to the rate, not parade through the whole list. The factors group into a few practical categories:
- Market evidence — the owner’s own licenses on this patent, the infringer’s licenses on comparable patents, and customary royalties in the industry. These correspond to a market approach to value, and comparability has to be real rather than loose.
- Profitability and value of the invention — the product’s established profitability and, critically, the share of profit attributable to the patented invention as distinct from unpatented elements, the manufacturing process, business risk, and improvements the infringer added. This is the income-approach heart of the analysis.
- The commercial relationship and licensing context — whether the parties are direct competitors, the exclusivity and scope of the hypothetical license, the patent’s remaining life, and the owner’s licensing policy.
- The nature and advantages of the technology — the benefit the invention offers over the prior approaches, and the extent of the infringer’s actual use.
The fifteenth factor is the hypothetical negotiation itself, which pulls the rest together into a single agreed rate. Good practice is to do the quantitative work first — comparable licenses, profitability, the value the invention adds — and then let the qualitative factors refine, rather than treating the list as a scoreboard.
Apportionment and the Entire-Market-Value Rule
This is, in my experience, the single most contested area in modern royalty analysis, and the place where a sloppy number gets dismantled on cross-examination.
The default rule is apportionment. When a patent covers one feature inside a multi-feature product, the royalty must be tied to the value of that feature — not to the full price of the whole product. You generally cannot use total product revenue as the base just because the patented feature is somewhere inside. The royalty base is meant to approximate the smallest saleable unit that practices the patent, and even that unit may need further apportionment if it still bundles in unrelated, unpatented value.
The narrow exception is the entire-market-value rule. Only when the patented feature is what creates customer demand for the whole product — when it is what drives buyers to purchase the entire thing — may the royalty rest on the full product’s value. That is a demanding showing, and it is the same demand concept that anchors the first lost-profits factor. Absent that proof, apportionment governs.
There is more than one way to apportion. The value of the patented feature can be isolated through careful selection of the royalty base, or by adjusting the royalty rate downward to discount the unpatented features, or by a combination. What the law requires is that, however it is done, the final product of base times rate reflects only the value attributable to the patented invention — and no more. When the patent is genuinely directed at the product as a whole rather than one buried component, using the full product as the base can be appropriate, provided the rate then accounts for the conventional, unpatented elements bundled into the claim.
This focus on isolating the value of what was actually taken connects patent royalties to the broader family of intellectual-property and gain-based remedies. Where the measure is the infringer’s profit rather than the owner’s loss, the mechanics resemble disgorgement of profits / unjust enrichment, and the apportionment discipline carries over directly to trade-secret damages, where courts likewise insist that recovery reflect the contribution of the protected technology rather than the full value of the end product.
A Hypothetical Illustration: Lost Profits vs. a Royalty Fallback
The figures below are entirely hypothetical. I invented them to show how the two measures move on the same set of facts; they are not drawn from any actual case, client, or engagement, and they are not a benchmark for any real matter.
Assume an infringer sold 10,000 units of an infringing product. The patent owner sells a competing product at $500 per unit, with incremental (variable) cost of $300 per unit, for an incremental margin of $200 per unit. Assume the owner can prove demand, the absence of acceptable substitutes, and the capacity to make the additional units — so under the Panduit framework it can recapture the full 10,000 units. As a royalty fallback on the same volume, assume the evidence supports a 6% running rate applied to a properly apportioned base of $250 per unit (the portion of the $500 price attributable to the patented feature).
| Measure | Calculation | Result |
|---|---|---|
| Lost profits (all 10,000 units recaptured) | 10,000 units × $200 incremental margin | $2,000,000 |
| Lost profits (only 6,000 units provable; 4,000 fall to royalty) | (6,000 × $200) + (4,000 × $250 × 6%) | $1,260,000 |
| Reasonable royalty only (statutory floor, all 10,000 units) | 10,000 units × $250 base × 6% | $150,000 |
The illustration makes three points. First, a fully supported lost-profits case is typically the larger recovery, which is why owners pursue it when the proof is there. Second, when the owner can only prove it would have captured part of the volume, the remainder does not disappear — it shifts to a royalty, so the two measures combine. Third, the reasonable royalty is the floor: even with no provable lost sales at all, the owner is still paid for the infringer’s use of the invention. The realism of every figure here depends on the underlying support — the margin, the recapture percentage, the apportioned base, and the rate — which is exactly where the forensic work lives.
How I Approach a Patent Damages Engagement
My role is to build the financial measures and to make them defensible, not to opine on infringement or validity. In practice that means I work to understand the patented technology in business terms — what advantage it actually delivers and how that advantage translates into margin — because both the demand analysis and the apportionment analysis depend on it. I tie every number to a verifiable source: unit volumes, prices, cost behavior, comparable licenses, and the share of value the invention contributes. And I test the other side’s story, including the alternatives an infringer realistically could have turned to, rather than assuming the most favorable counterfactual.
This work overlaps heavily with the methods I use across commercial damages — the but-for projection, the before-and-after and yardstick approaches for modeling the world without the wrong, the causation analysis that connects the conduct to the loss, and the duty to mitigate where it applies. If you want to understand how I serve as a testifying expert on these issues, my page on expert witness and litigation support describes the engagement.
FAQ
What is the difference between lost profits and a reasonable royalty in a patent case?
Lost profits compensate the patent owner for the additional profit it would have earned on sales it lost because of the infringement — a but-for measure of the owner’s own loss. A reasonable royalty compensates the owner for the value of the infringer’s use of the technology, measured through a hypothetical license negotiation. Lost profits are usually the larger recovery when the proof supports them; the reasonable royalty is the statutory floor that applies even when lost sales cannot be proven.
Can a patent owner recover both lost profits and a royalty?
Yes. A common structure is lost profits on the slice of infringing sales the owner can prove it would have captured, plus a reasonable royalty on the remaining infringing sales it could not prove it would have made. The market-share approach is one formal way this combination arises.
What are the Panduit factors?
They are the four-part test most often used to establish but-for causation for lost profits: demand for the patented product, the absence of acceptable non-infringing substitutes, the owner’s capacity to meet the demand, and the amount of profit the owner would have earned. It is the leading framework but not the only accepted method of proving causation.
What does apportionment mean, and why does it matter so much?
Apportionment requires that the royalty capture only the value the patented feature itself contributes, not the full value of a multi-feature product. It matters because using whole-product revenue as the royalty base, when the patent covers just one component, overstates damages and is a frequent target on cross-examination. The narrow exception — the entire-market-value rule — applies only when the patented feature is what drives customer demand for the entire product.
Does the accountant decide whether infringement was willful or whether fees should be awarded?
No. Willfulness, enhanced damages, and fee-shifting are legal determinations for the court. My analysis supplies the compensatory measure — the lost profits or the reasonable royalty. I take validity and infringement as assumptions handed to me by the liability case and quantify the economic harm; the legal questions belong to counsel and the court.
How do I reach you to discuss a patent damages matter?
You can reach me, Joey Friedman, CPA, at 954-282-9615 to discuss a commercial-litigation engagement involving patent or other intellectual-property damages. I am based in Florida and work as the financial-damages expert alongside trial counsel, who handle the legal questions of validity, infringement, and liability.
About the Author
Joey Friedman is a CPA, Accredited in Business Valuation (ABV), and forensic accountant who holds a Master of Accounting and a Master of International Business and is a member of the AICPA and the Association of Certified Fraud Examiners. He also holds a Florida real estate license. Beyond those credentials, he has personally owned and operated more than a dozen of his own businesses across industries including marketing, printing, transportation, restaurants, hospitality and entertainment, and event planning — so he quantifies what an infringement actually cost the patent owner, or earned the infringer, with both a forensic accountant’s rigor and an owner-operator’s grasp of how a single product feature translates into real margin.