Patent damages compensate the patent owner for infringement, and the Patent Act sets their floor: damages adequate to compensate for the infringement, but never less than a reasonable royalty for the use the infringer made of the invention, together with interest and costs. A patentee that would have made the infringer's sales can claim the profits it lost on them, including price erosion and certain related sales, if it proves it would have made them; every patentee can claim a reasonable royalty, the license fee a willing licensor and a willing licensee would have agreed to in a hypothetical negotiation on the date the infringement began. Both measures turn on apportionment, because the damages must reflect the value of the patented invention and not the unpatented features of the product it sits in. This guide explains where patent cases are heard and how the damages period is set, how lost profits and the reasonable royalty are built, how design patents differ, what a patent damages expert does for the patentee or the accused infringer, and where the analysis is contested.
Section 284 of the Patent Act directs the court to award a patent owner who proves infringement damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer, together with interest and costs as fixed by the court. The measure is the patentee's loss: the difference between its position after the infringement and the position it would have been in had the infringer not infringed. For a utility patent, which protects how a product or process works, the infringer's own profits are not a separate remedy; Congress amended the statute in 1946 so that only damages are recoverable, and the infringer's gains matter as evidence of what the patentee lost or of what a license would have been worth. A design patent, which protects how an article looks, carries an additional remedy, the infringer's total profit, which the design patent section below explains.
Within that floor the patentee chooses its theory, and the two measures can be combined. A patentee that competes in the market can claim the profits it lost on the sales it would have made but for the infringement and a reasonable royalty on the infringing sales it cannot show it would have made; a patentee that licenses rather than sells, or that cannot prove lost sales, ordinarily recovers a reasonable royalty on all of them. The court may increase the damages up to three times the amount found, but enhancement is a discretionary sanction the Supreme Court has described as generally reserved for egregious infringement behavior, decided by the court and separate from the economic measure of the loss. The economist measures the compensatory amount and the prejudgment interest on it, at the rate and compounding the court adopts, and the statute itself provides that the court may receive expert testimony as an aid to the determination of damages or of what royalty would be reasonable.
A claim for patent infringement arises under federal law, and the federal courts have exclusive jurisdiction over it: no court of any state, the District of Columbia, or a territory may hear a claim for relief arising under the patent laws, and the same is true of copyright claims. A claim between private parties is heard in a federal district court, while the owner's remedy for an invention used or manufactured by or for the United States without a license is an action against the United States in the Court of Federal Claims for reasonable and entire compensation. Appeals go to the U.S. Court of Appeals for the Federal Circuit rather than to the regional circuit that covers the district, from the territorial district courts as well as the others, so the Federal Circuit's damages decisions govern patent cases in every district. A dispute over the royalties owed under a license is different: it is usually a contract claim governed by state law, which a state court can hear, and the economist measures it under the license's terms rather than under the patent statute, although once an infringement claim is joined with it the case belongs in federal court.
Three rules fix the period the damages can cover. No recovery is had for infringement committed more than six years before the complaint or counterclaim for infringement was filed. Where the patentee or its licensees sold patented articles without marking them with the patent number, or with an internet address that associates the article with the number, damages run only from the date the infringer was notified of the infringement and continued to infringe, and filing the lawsuit is itself notice. And the reasonable royalty is negotiated, hypothetically, as of the date the infringement began, which can fall before the period for which damages can be collected: the six-year limit and the marking rule restrict what is recoverable without moving the date on which the license would have been agreed. The economist builds the schedule of infringing sales for the recoverable period from the infringer's records, by product and by period, and keeps the start of the infringement, the notice date, and the six-year limit visible in it, since each is often disputed.
To recover lost profits, the patentee must show that but for the infringement it would have made some or all of the infringer's sales, or made its own sales at a higher price, and the loss must have been reasonably foreseeable. The usual proof follows four factors from the Panduit decision: demand for the patented product, the absence of acceptable noninfringing substitutes, the patentee's manufacturing and marketing capability to exploit the demand, and the amount of profit it would have made. The factors are a common route rather than the only one. Where acceptable substitutes exist and several competitors share the market, the Federal Circuit has accepted a market share approach, under which the patentee claims the share of the infringing sales that matches its share of the market and a reasonable royalty on the rest.
The but-for market has to be reconstructed as it would have been, not as either side would like it. An acceptable noninfringing alternative that the infringer could have adopted counts against the claim even if it was not on the market during the infringement, provided it was available, because a rational competitor facing the patent would have switched to it rather than leave the market. The lost sales must fit within the patentee's capacity to make and sell them, and the profit on them is incremental: the added revenue less the costs that would have risen with the added volume, such as materials, direct labor, commissions, and freight, with the fixed costs that would not have changed left out. Lost sales of the patentee's own competing product can be recovered even if that product does not practice the patent, where the infringement caused them and the loss was foreseeable, while unpatented items sold with the patented product count only if they function together with it as a unit, not merely because customers buy them together.
Price erosion is the profit the patentee lost because it charged less than it would have without the infringer's competition. The claim is an economic counterfactual: at the higher but-for price the patentee would have sold fewer units, so the analysis has to estimate how demand in the market responds to price, from the sales history, customer evidence, or pricing studies, rather than apply the higher price to every unit actually sold. The same discipline applies to accelerated market entry, lost follow-on sales, and other projected losses, and the lost profits but-for analysis page describes the before-and-after and yardstick methods that commercial damages share.
Every patentee is entitled to at least a reasonable royalty, and the courts measure it through a hypothetical negotiation: the license a willing licensor and a willing licensee would have agreed to just before the infringement began, had both been reasonably and voluntarily trying to reach an agreement. The negotiation is a construct with fixed premises. The parties are assumed to know that the patent is valid and infringed, so the discount a real licensee would demand for the risk that the patent fails is absent; they are presumed to know the facts surrounding the infringement as of its start; and the licensee is expected to keep a reasonable profit after paying the royalty. The courts also recognize an analytical method, which takes the infringer's own projections of profit for the infringing product at the time the infringement began, subtracts the infringer's usual or acceptable net profit from that projection, and treats the remainder as the royalty.
The framework courts commonly use to organize the evidence is the list of fifteen factors from the Georgia-Pacific decision. They include the royalties the patentee has received for licensing the patent in suit; the rates the licensee has paid for comparable patents; the nature and scope of the license, exclusive or not and restricted or not; the licensor's policy of licensing or of keeping the invention to itself; the commercial relationship between the parties, such as whether they compete; the effect of the invention on sales of other products; the duration of the patent and the term of the license; the profitability, commercial success, and popularity of products made under the patent; the advantages of the invention over earlier methods; the nature of the invention and the benefits to those who have used it; the extent of the infringer's use; the share of profit or price customary in the business for the use of comparable inventions; the portion of the profit attributable to the invention as distinguished from unpatented elements, the manufacturing process, business risks, or features the infringer added; the opinion testimony of qualified experts; and, finally, the amount the parties would have agreed upon in the negotiation itself. The factors are not a formula and carry no fixed weights. A report that recites all fifteen and labels each as raising or lowering the rate without tying it to evidence explains little; one that identifies the factors the record actually informs, usually the comparable licenses, the profit attributable to the invention, the parties' competitive relationship, and the alternatives available to the infringer, and quantifies each, can be tested.
The economics of the negotiation set a range. The licensor would not accept less than what licensing the infringer would cost it, which for a patentee that competes includes the profit it expects to lose to a licensed competitor; the licensee would not pay more than the invention is worth to it over its next-best alternative, often the cost and delay of designing around the patent or of switching to an acceptable noninfringing substitute. Where the infringer could have switched cheaply, the royalty is constrained by that cost; where no acceptable alternative existed, it can approach the incremental profit the invention made possible. The royalty can be a running rate per unit or a percentage of revenue, or a lump sum for the whole period, and the form matters when licenses are compared, since a lump sum allocates the risk of future sales between the parties differently than a running royalty does.
The royalty has to compensate for the patented invention and nothing else. When the accused product has many features and the patent covers one of them, applying a rate to the price of the whole product risks paying the patentee for features it did not invent, so the royalty is often built on the smallest salable patent-practicing unit, and where that unit itself carries valuable unpatented features, the analysis still has to separate their value, because the damages must be tied to the invention's footprint in the market. The base and the rate apportion together: the entire value of a multi-component product can be claimed without further apportionment only under the entire market value rule, a narrow exception that applies when the patented feature drives the demand for the whole product, and otherwise a whole-product base is acceptable only where the rate does the apportioning, as with comparable licenses priced on the same base. The Federal Circuit has rejected a rule that every damages model must begin with the smallest salable patent-practicing unit, and it has held that using the accused product as the royalty base and apportioning through the rate is an acceptable methodology, so long as the apportionment is adequate and reliable; a small rate applied to a large base without evidence tying the rate to the invention does not cure the problem. The Federal Circuit has also held that the 25 percent rule of thumb, which assigned the patentee a fixed share of the infringer's expected profit, is a fundamentally flawed tool whose results are inadmissible because it does not tie the royalty to the facts of the case.
Apportionment is done with evidence. The record may support it through the comparable licenses themselves, if they priced a similar technology in a similar product; through the infringer's own documents, such as feature-level cost, pricing, or engineering analyses; through consumer surveys, often a conjoint study designed by a survey expert, that estimate how much buyers value the patented feature relative to others; or through the price premium of products with and without the feature. Each method has weaknesses the other side will press, and the economist states which inputs come from other experts, such as the technical expert who testifies that a feature practices the claims or that an alternative was available and acceptable.
Comparable licenses are often the strongest evidence of the rate and the most contested. A license is probative only to the extent it is comparable in the technology, the rights granted, and the economic circumstances, and the report has to account for the differences: a portfolio license covering many patents, a cross license in which each side pays partly with its own patents, a lump sum that has to be converted into a rate with assumptions about the licensee's expected sales, or a license granted to settle litigation, whose terms can reflect the cost and risk of the case as much as the value of the patent. Sitting en banc in 2025, the Federal Circuit held that a damages expert's testimony that earlier lump-sum licenses reflected an agreed per-unit rate was not based on sufficient facts or data, because the licenses did not show that the licensees had agreed to that rate, and it described the court's reliability determination under Rule 702 as an essential prerequisite to the jury's weighing of the testimony. The license agreements post explains what each license in the record can and cannot show.
A design patent protects the ornamental design of an article of manufacture rather than how the article works, and the Patent Act gives its owner a remedy a utility patentee does not have: whoever applies the patented design, or a colorable imitation of it, to an article of manufacture for sale, or sells such an article, is liable to the owner to the extent of its total profit. The calculation has two steps: identify the article of manufacture to which the design was applied, and compute the infringer's total profit made on that article. The Supreme Court has held that in a multi-component product the article can be the end product sold to the consumer or a component of it, so the identification of the article is often the central economic dispute, followed by which costs may be deducted in computing the profit. The owner keeps the general damages remedy as well, including a reasonable royalty or lost profits, but cannot recover the same profit twice, and the economist commonly computes each available measure so the choice can be made on the record.
A patent damages expert builds the measure from the record, and the work follows the same sequence for the patentee and for the accused infringer. It begins with the products: which of the infringer's products are accused, which of the patentee's products practice the patent or compete with the accused products, and how each is sold. The infringer's sales, pricing, and cost data for the accused products, by unit, period, and customer, establish the royalty base and the volumes; the patentee's financial and capacity records establish what it could have sold and at what incremental profit; and the licenses each side has granted or taken, the files from any licensing negotiations, and the infringer's internal analyses of the technology's value supply the evidence for the rate. The expert relies on the technical expert for what the claims cover, which features practice them, and whether an alternative was acceptable and available, and says so in the report. The sales and cost data are usually produced under a protective order that limits who may see them, which shapes how the report presents them.
The expert then computes the measures the case supports, often lost profits on the sales the patentee can show it would have made and a royalty on the rest, carries them through the damages period with the notice and limitation dates applied, adds prejudgment interest on the basis the court adopts, and tests how sensitive the result is to the inputs the other side will contest. The disclosure rules require a written report that states every opinion and its basis, and in a patent case the schedule usually follows the court's construction of the claims, since the meaning of the claims decides which products infringe and which alternatives were noninfringing. Much of the work is responsive: the opposing expert has made the same choices differently, and the report has to show what the result is under the other side's base, rate, and alternatives, and why each was accepted or rejected. The expert disclosure guide and the federal versus state court guide describe the rules that govern the report and its admission.
The disagreements between patent damages experts fall into a short list, and nearly all of them are visible on the face of the reports. The base: a royalty applied to the whole product when the patent covers one feature and nothing in the rate apportions it, or a smallest salable patent-practicing unit that still carries valuable unpatented features. The rate: licenses offered as comparable that differ in technology or terms, a lump sum converted to a running rate on unsupported volume assumptions, settlement licenses treated as ordinary market transactions, or a rate taken from a rule of thumb rather than the facts. The alternatives: a design-around the record does not show was available or acceptable, or, on the other side, a claimed absence of substitutes that ignores products customers actually bought. The lost sales: a market defined to exclude the substitutes, capacity assumed rather than shown, or profit computed on average rather than incremental costs. The period: damages counted before notice where the patentee's products were not marked, or beyond the six-year limit.
Each of these can be tested against the record, and the Federal Circuit's decisions make many of them questions of admissibility as well as weight. The rebuttal guide lists the questions that expose an unsupported input in any economic report, the report review service describes how an opposing analysis is examined one choice at a time, and the lost profits versus reasonable royalty comparison sets the two measures side by side.
The earlier the intellectual property damages economist reads the record, the more the damages theory can shape discovery rather than react to it. The first requests are usually the infringer's unit sales, revenue, and cost data for the accused products and the patentee's for its own; the licenses each side has signed for the patent in suit and for comparable technology, with the negotiation files; the marking and notice records; the parties' business plans, forecasts, and market studies from around the date the infringement began; the documents on any design-around the infringer considered or adopted; and the technical expert's views on the alternatives. The reasonable royalty methodology page lays out the steps the royalty analysis follows, and the lost profits and commercial damages service describes the related commercial work.
Ordinarily not, because lost profits require proof of sales the owner would have made but for the infringement. A licensing company, or an owner that neither practices the patent nor sells a competing product, recovers a reasonable royalty, which the statute sets as the floor for every patentee and which is measured by the license the parties would have agreed to when the infringement began.
Not necessarily. Earlier offers and negotiations are evidence, but the hypothetical negotiation assumes what real negotiators did not know: that the patent is valid and infringed. A rate offered while validity was uncertain, or in the shadow of litigation, can understate or overstate what that assumption produces, so the report explains how each real negotiation bears on the rate.
No further than six years before the complaint or counterclaim was filed, and where patented products were sold unmarked, only from the date the infringer was notified, which the lawsuit itself provides. The date of the hypothetical negotiation can still be earlier, at the start of the infringement, because the limits restrict what is collected, not when the license would have been struck.
Request a consultation on Patent Damages Explained: Reasonable Royalty, Lost Profits, and Apportionment or call (201) 343-0700. Plaintiff and defense counsel.