In brief. Federal civil litigation—and patent litigation in particular—is expensive, slow, and consequential. For a small business, a single infringement suit can put a core product, and sometimes the company itself, at risk. This guide explains how the process actually works, stage by stage, so an owner can weigh the costs and risks intelligently, communicate with counsel, and decide whether to fight, settle, or walk away. It is written to be understood by a judge, a seasoned litigator, and a first-time founder alike. It is not legal advice; patent cases turn on facts and on law that shifts, sometimes abruptly, so consult counsel before acting.

Introduction: a lawsuit as a capital decision

Here is a number worth sitting with. By the time a contested patent case reaches a jury, each side will typically have spent between $600,000 and several million dollars—and that is the median, not the nightmare scenario. For a company with twelve employees and a single flagship product, that is not a line item; it is a bet-the-company event dressed up in the language of "civil procedure."

This is the central paradox of patent litigation for small businesses. The patent system exists, in significant part, to give the small inventor a fighting chance against the large incumbent. Yet the machinery built to vindicate that right is so costly and slow that the small player can be ground down by the very process meant to protect it. Knowing where the leverage points are, where the money goes, and where a smaller party can punch above its weight is the difference between using the system and being used by it.

To keep the abstractions grounded, this guide follows one fictional company through the entire arc of a case. Meet Brightline Sensors, Inc., a twelve-person Ohio startup that designs a low-power wireless moisture sensor used in commercial greenhouses and cold-storage warehouses. Brightline holds a single U.S. patent on the way its sensor calibrates itself to humidity drift. One morning, its founder discovers that Vortex Industrial, LLC, a national equipment supplier with two thousand employees, has begun selling a competing sensor that looks suspiciously like Brightline's design. Depending on the facts, Brightline could be a plaintiff (suing for infringement) or, if Vortex strikes first, a defendant (accused of infringing a Vortex patent). We use Brightline to illustrate each decision point, because the same procedural machinery looks very different depending on which side of the "v." you sit on. (Brightline and Vortex are invented; any resemblance to real companies is coincidental.)

This guide does not substitute for experienced counsel—patent litigation is among the most specialized fields of American law, and procedural missteps can be fatal to an otherwise strong case—but understanding each stage lets an owner make informed strategic choices, talk to their attorneys as an engaged client rather than a passenger, and weigh cost and risk at each juncture. If you read only one companion piece alongside this one, make it our guide to evaluating and assessing a civil lawsuit, which supplies the case-valuation discipline that underlies every decision below.

We proceed chronologically: pre-litigation strategy and cost assessment; venue selection; pleadings and early motions; discovery; claim construction; inter partes review; patent eligibility under Section 101; summary judgment, pre-trial, and trial; damages and injunctive relief; post-trial motions and appeal; and enforcement. Each section addresses both the general framework that governs all federal civil cases and the patent-specific features that make this area uniquely demanding. Throughout, keep one anatomical fact in mind: a federal lawsuit has a skeleton, and that skeleton is the Federal Rules of Civil Procedure. Rules 8, 12, 16, 26, and 56 recur in nearly every section, setting the rhythm of pleading, dismissal, scheduling, discovery, and summary judgment that every case—patent or otherwise—must move through.

Part I: Pre-Litigation Strategy

The most important decisions in a patent case are often made before anything is filed. A clear-eyed pre-litigation assessment is the difference between a disciplined campaign and an expensive improvisation—and the smartest version happens around a conference table with a teardown of the accused product, a copy of the patent, and a frank conversation about money, long before a dollar of filing fees is spent.

Assessing the case

Before committing, a small business must rigorously assess whether the case justifies the investment. That involves several distinct analyses, each of which should be completed before a complaint is filed—not discovered halfway through discovery.

Infringement or invalidity analysis. The threshold question is whether the would-be plaintiff has a good-faith basis to assert that the defendant's product or process infringes the patent claims. This requires a careful, claim-by-claim, element-by-element comparison of the patent claims—as a court is likely to construe them—against the accused product. The claims are the numbered sentences at the end of the patent; they define the legal boundaries of the invention the way a deed defines a parcel of land. Infringement exists only if the accused product contains every element (or "limitation") of at least one claim. Miss a single limitation and there is no literal infringement. This is the "all-elements rule," and it is unforgiving: a patent is not a vague aura over a general idea but a precisely fenced parcel, and the accused product is either inside the fence or it is not.

Before Brightline sues, then, its counsel will line up the language of Claim 1 against a teardown of the Vortex sensor. Suppose Claim 1 requires "a calibration module that adjusts a baseline reading in response to a measured rate of humidity change." If the Vortex device recalibrates on a fixed timer rather than in response to a measured rate of change, Brightline may have no literal infringement case at all, no matter how similar the two products look on a store shelf. That is the kind of conclusion you want to reach in a conference room for a few thousand dollars, not on the eve of trial for a few million. Good counsel will often build a preliminary claim chart—each claim limitation on the left, the corresponding feature of the accused product on the right—because a chart with an empty cell tells you to think hard before filing.

For a prospective defendant, the mirror-image analysis asks whether viable defenses exist: non-infringement, invalidity over prior art, unenforceability for inequitable conduct, or ineligibility under 35 U.S.C. § 101. Filing—or defending—without this work invites sanctions under Rule 11, which requires that every claim and defense be "warranted by existing law or by a nonfrivolous argument" and that factual contentions "have evidentiary support." Rule 11 is not a dead letter in patent cases; the Federal Circuit has made clear that a patentee must at least apply the claims to the accused device before filing, and courts have sanctioned parties who asserted patents with no pre-suit comparison at all. The broader doctrinal map—what direct, induced, and contributory infringement require, and how the standard defenses line up—appears in what constitutes patent infringement: claims and defenses.

Prior art investigation. "Prior art" is the universe of earlier patents, publications, products, and public uses that existed before the patent's effective filing date. It matters because a patent can be invalidated if the claimed invention was already known (anticipation under 35 U.S.C. § 102) or was an obvious variation on what was known (obviousness under § 103). For a plaintiff, prior art is a threat to neutralize: a defendant will almost always argue invalidity, so the plaintiff needs to know in advance whether the patent can survive that attack. For a defendant, strong prior art is gold—it can support summary judgment, provide settlement leverage, or anchor an inter partes review petition (Part VI). If Vortex's litigators unearth a Japanese sensor patent from 2009 that already taught humidity-rate calibration, Brightline's lawsuit may collapse before it reaches a jury. The obviousness inquiry is its own discipline; our deep dive on overcoming obviousness rejections under Section 103 explains the Graham v. John Deere factors a validity fight turns on.

Prosecution history review. A patent's "prosecution history" (the file wrapper) is the written record of the back-and-forth between the inventor and the USPTO examiner while the application was pending. It is a critical resource, because statements the applicant made to win allowance can come back to haunt the patent owner. Under prosecution history estoppel, a patent holder generally cannot recapture claim scope it surrendered to get the patent granted—a principle the Supreme Court reaffirmed in Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722 (2002). If, to overcome a rejection, Brightline's inventor told the examiner her invention works "only with a measured rate of change, never with a fixed timer," she will struggle later to argue that a fixed-timer device infringes under the doctrine of equivalents. The file wrapper is also where the seeds of an inequitable conduct defense may be buried: if the inventor withheld a material reference with intent to deceive, the entire patent can be rendered unenforceable. Practitioners call it the "atomic bomb" of patent law; we unpack it in finding evidence of inequitable conduct in patent prosecution.

The cost-benefit calculus

Patent litigation is among the most expensive forms of civil litigation in the United States. The AIPLA Report of the Economic Survey—the standard industry benchmark, published biennially by the American Intellectual Property Law Association—reports median costs that should give any small business pause. On a per-patent, per-side basis, the figures look roughly like this.

For cases with less than $1 million at stake, expect about $300,000 through discovery and claim construction, and roughly $600,000 more through trial and appeal. For $1–10 million, plan on about $600,000 through claim construction and another $1 million through trial and appeal. For $10–25 million, roughly $1.5 million through claim construction and $3 million more beyond. For more than $25 million, over $1.5 million through claim construction and $3.6 million or more beyond. For complex technologies—pharmaceuticals, semiconductors, certain software—add up to fifty percent.

Read those numbers slowly, because they reframe the entire decision. A patent case that reaches trial will almost certainly cost each side at least $600,000 in fees, often several million. For Brightline—annual revenue perhaps a few million dollars—the realization that a single lawsuit could cost more than a year's profit is sobering. It is also clarifying: it forces the question whether the recovery (for a plaintiff) or the exposure (for a defendant) justifies the spend. If Vortex's infringing sales are stealing $5 million a year, litigation may be worth it. If Vortex sells only a handful of units in a niche Brightline does not even serve, the math is very different.

Where does the money go? The lion's share—frequently sixty to seventy percent—is consumed by discovery, and within it by document review and expert work. Claim construction (Part V) is a second concentrated spike. Trial itself is often a smaller slice than clients expect, because most cases settle first. The upshot: a small business that resolves a case before full-blown discovery and a Markman hearing has avoided the two most expensive phases entirely. Every off-ramp before those gates is worth real money.

This is also where alternative fee arrangements deserve a hard look. Ask potential counsel about contingency fees, hybrid arrangements (a reduced hourly rate plus a success fee), capped or phased engagements (a fixed budget per phase, with a go/no-go decision before each), and litigation funding. Funding has become common in patent cases—common enough that some courts now require its disclosure—and it can let a cash-poor but patent-rich startup pursue a meritorious claim it could never otherwise afford.

Finally, fee-shifting cuts both ways. The American Rule presumes each side bears its own fees, but patent law carves out an exception: in an "exceptional" case, the prevailing party may recover its attorney fees under 35 U.S.C. § 285. The Supreme Court loosened the standard in Octane Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545 (2014), defining an exceptional case simply as one that "stands out from others" based on the substantive strength of a party's position or the unreasonable manner in which the case was litigated. Its companion, Highmark Inc. v. Allcare Health Management System, Inc., 572 U.S. 559 (2014), held that the finding is reviewed only for abuse of discretion—making fee awards both more available and harder to overturn than before. That is a sword and a shield: a defendant who maintains a frivolous defense, or a plaintiff who brings a baseless claim and litigates it abusively, risks paying the other side's bills. (The Lanham Act borrows the same "exceptional case" language, and Octane Fitness now governs there too—a parallel explored in our note on Lanham Act attorneys' fees under 15 U.S.C. § 1117(a).)

Alternative dispute resolution

Before anyone files anything, counsel should ask whether the dispute can be resolved without a courtroom. Patent disputes are frequently settled through licensing, mediation, or arbitration—and for small businesses, the advantages are substantial. Mediation can preserve a working relationship between competitors who must coexist in the same industry, keeps the dispute confidential (no permanent public record), and usually resolves matters faster and far more cheaply than full litigation. Arbitration yields a binding decision without the procedural weight and expense of federal court; 35 U.S.C. § 294 expressly makes patent disputes arbitrable, many license agreements require it, and even without a clause, parties can agree to arbitrate after a dispute arises.

For a company like Brightline, the calculus often tilts toward resolution. A mediated settlement might cost $50,000 to $100,000 in fees, against the $600,000-plus that trial requires. Perhaps the parties strike a cross-license: Vortex pays a modest royalty and keeps selling, while Brightline avoids a bet-the-company trial and gains a revenue stream. The pivotal question is whether the other side will engage in good faith—which depends on the parties' relative leverage, the strength of the claims, and any ongoing relationship worth protecting. A well-timed demand letter that lays out infringement element by element can itself open a productive negotiation; see our primer on writing a demand letter. For the broader menu, see arbitration, mediation, and choosing a dispute resolution forum.

A note of caution on demand letters. Accusing someone of infringement can itself trigger a lawsuit—against you. Under MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118 (2007), a recipient who reasonably fears suit may race to federal court and file a declaratory judgment action seeking a ruling of non-infringement or invalidity, often in a forum of its choosing. A demand letter is not a free move: Brightline's lawyers must weigh the upside of opening negotiations against the risk of handing Vortex the keys to pick the courthouse.

Preservation obligations

The moment litigation becomes "reasonably anticipated"—whether you expect to sue or to be sued—a legal duty to preserve relevant evidence attaches. In practice, that means issuing a litigation hold: a written instruction to every employee and agent to stop deleting and start preserving documents, communications, and electronic data bearing on the dispute. This is not a formality. Under Federal Rule of Civil Procedure 37(e), as amended in 2015, the failure to take reasonable steps to preserve electronically stored information can lead to serious sanctions; where a party acted "with the intent to deprive another party of the information's use," the court may instruct the jury to presume the lost evidence was unfavorable—an "adverse inference" that can sink a case on its own. Lawyers trace the modern duty to Zubulake v. UBS Warburg LLC, the influential series of opinions that defined it.

For patent matters, the preservation net is wide. It covers technical materials (source code, design specifications, engineering notebooks, test data), communications about the accused product or the patented technology, and financial records tied to sales, revenue, and profit. Imagine that, two weeks after Brightline's founder learns of the Vortex sensor, an engineer routinely wipes an old laptop holding early design files—exactly the documents that would show whether Brightline's invention predated a key prior-art reference. That ordinary act of housekeeping, performed after the duty attached, could expose Brightline to sanctions and undercut its credibility with the court. A prompt, well-documented hold is cheap insurance. The mechanics of building, distributing, and tracking a hold are taken up in our [practical discovery refresher](/documents/a-practical-discovery-refresher---mastering-the-tools-rules- and_pitfalls_of_federal_civil_litigation).

Part II: Venue Selection

Why venue matters

Few early decisions shape a patent case as much as where it is filed. The choice of court can determine how fast the case moves, which judge presides (and how experienced that judge is with patents), the jury pool, the local rules governing discovery and claim construction, and the odds on pivotal motions. Sophisticated plaintiffs spend real money analyzing venue before filing, and for good reason: the same case can be a sprint in one district and a marathon in another, and the difference between a court with mandatory patent local rules and one without can reshape the entire budget.

The patent venue statute

Patent venue is governed by its own statute, 28 U.S.C. § 1400(b), which allows a patent infringement action to be brought either in the district where the defendant "resides" or in a district where the defendant "has committed acts of infringement and has a regular and established place of business." For decades, courts read "resides" expansively, allowing suit almost anywhere a defendant did business. That changed with TC Heartland LLC v. Kraft Foods Group Brands LLC, 581 U.S. 258 (2017), in which the Supreme Court held that, for a domestic corporation, "resides" means only its state of incorporation—rejecting the broad reading that had prevailed since VE Holding Corp. v. Johnson Gas Appliance Co., 917 F.2d 1574 (Fed. Cir. 1990), which had effectively allowed suit anywhere a company was subject to personal jurisdiction.

TC Heartland reshaped the landscape overnight. Before it, a plaintiff could sue a nationally distributed company in nearly any district where its products were sold—which, for many defendants, meant anywhere. After it, venue is limited to the defendant's state of incorporation or a district where the defendant both commits infringement and maintains a "regular and established place of business." The Federal Circuit promptly held that the latter phrase requires (1) a physical place in the district, (2) that is regular and established, and (3) that is the place of the defendant, not merely of an employee. In re Cray Inc., 871 F.3d 1355 (Fed. Cir. 2017). A remote employee's home office, a traveling salesperson, a cloud server—none reliably counts.

For Brightline, this is more than trivia. If Vortex is incorporated in Delaware and has a genuine distribution warehouse in the Eastern District of Texas, Brightline may sue in either place. If Vortex has no physical foothold in Ohio, Brightline generally cannot sue at home, even though that is where it operates and where the harm is felt. Venue follows the defendant's footprint, not the plaintiff's—a counterintuitive rule that surprises founders who assume they can fight on home turf. (Foreign defendants differ: a defendant not resident in the United States may be sued in any district, one reason cross-border infringement so often lands in plaintiff-friendly forums. Serving such a defendant, particularly one based in China, is its own ordeal, addressed in our guide to serving a China-based defendant under the Hague Service Convention.)

The current venue landscape

The map of patent litigation has shifted repeatedly since TC Heartland, and a small business should understand the current terrain rather than rely on the reputation a district had a decade ago.

The Eastern District of Texas, long the country's busiest patent forum, saw filings drop sharply right after TC Heartland, because many defendants lacked the required physical presence in that largely rural district. From roughly 2018 through 2022, the Western District of Texas surged, driven by Judge Alan Albright in the Waco Division, who adopted patent-friendly local rules, ran brisk schedules, and proved reluctant to transfer cases out. At its peak, the Western District drew about a quarter of all national patent filings—concentration that drew scrutiny, including a series of Federal Circuit mandamus decisions ordering transfers under 28 U.S.C. § 1404(a) where Waco had little real connection to the dispute. In July 2022, the district's chief judge ordered Waco patent cases randomly assigned across all the district's judges, ending the near-certainty that a Waco filing would land before Judge Albright; a later order tightened the "related case" practice that had let serial filers steer matters back to a preferred docket, and in early 2025 Judge Albright moved to the Austin Division.

By 2023 the Eastern District of Texas had reclaimed the top spot, and it has held it since. By 2024, roughly one in five patent cases nationwide were filed there. Through the first half of 2025, the Eastern and Western Districts of Texas together accounted for roughly forty percent of all patent filings and a majority of cases brought by non-practicing entities, with Delaware close behind. Delaware—historically a leading venue because so many corporations are incorporated there, which makes venue easy under TC Heartland's "state of incorporation" prong—saw filings dip after its chief judge began requiring disclosure of third-party litigation-funding arrangements, a transparency measure some plaintiffs found unwelcome.

For a small business, the takeaways are concrete. You can only sue where TC Heartland permits, so venue analysis starts with the defendant's corporate home and physical footprint. Among the permissible venues, districts differ markedly in speed, in whether they impose mandatory patent local rules, and in the assigned judge's track record. The landscape moves; what was true in 2020 may be wrong in 2026. And even a properly filed case can be transferred under § 1404(a) to a more convenient district, so a plaintiff who picks an aggressive forum may spend its first months defending the choice. An experienced patent litigator can map the options against the specific judge and district rules, and against your strategic goal—a fast trial to pressure settlement or a slower schedule that lets a smaller party catch its breath.

Part III: Pleadings and Initial Filings

The complaint

Litigation formally begins with a complaint, the document Rule 8(a) requires to contain "a short and plain statement of the claim showing that the pleader is entitled to relief." A patent complaint must identify the patents at issue; allege that the defendant infringes (specifying at least generally the accused products and the manner of infringement); establish subject-matter jurisdiction (exclusively federal under 28 U.S.C. § 1338) and personal jurisdiction; and state the relief sought, whether damages, an injunction, or both.

How much detail a patent complaint must contain has been debated for years. Under the Supreme Court's general pleading standards in Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), and Ashcroft v. Iqbal, 556 U.S. 662 (2009), a complaint must plead enough facts to make the claim "plausible on its face"—more than a label, less than a trial brief. Under Iqbal, a court first sets aside conclusory assertions (the "the defendant infringes" recitals) and then asks whether the remaining well-pleaded facts plausibly entitle the plaintiff to relief. For years, patent plaintiffs enjoyed a safe harbor in the bare-bones complaint form (old Form 18) appended to the Federal Rules, but the abrogation of those forms in 2015 closed that door, and Twombly/Iqbal now governs patent complaints like any other. A bare assertion that "the defendant infringes," with no explanation of how the accused product meets the claim limitations, increasingly draws a motion to dismiss. The safer course—and the one most courts now expect—is a complaint that maps the accused product onto the claim language with enough specificity to show the theory is real. If Brightline simply alleges that "Vortex's sensor infringes the patent," it risks dismissal; if it alleges that "the Vortex Model 7 contains a calibration module that adjusts its baseline reading in response to a measured rate of humidity change, as required by Claim 1," it stands on firmer ground. Many practitioners now attach a claim chart to the complaint for exactly this reason.

If the plaintiff intends to seek enhanced damages for willful infringement—up to triple the award under 35 U.S.C. § 284, as interpreted in Halo Electronics, Inc. v. Pulse Electronics, Inc., 579 U.S. 93 (2016)—the complaint should plead facts showing the defendant knew about the patent and chose to infringe anyway. That knowledge often arises from a pre-suit demand letter. If Brightline wrote to Vortex months earlier identifying the patent and the accused product, and Vortex kept selling, those facts support a willfulness allegation. (Another reason the demand-letter decision in Part I is so consequential: the same letter that creates declaratory-judgment risk also lays the foundation for treble damages.)

The response

Once served, a defendant generally has 21 days to respond under Rule 12(a), or 60 days if it waived formal service under Rule 4(d)—a courtesy that, in exchange for sparing the plaintiff the cost of service, buys time. The response can take several forms.

Motion to dismiss. Defendants frequently move to dismiss under Rule 12(b)(6) (failure to state a claim), Rule 12(b)(2) (lack of personal jurisdiction), or Rule 12(b)(3) (improper venue). After TC Heartland, the venue motion is a standard part of the defense toolkit—but note Rule 12(g) and 12(h): improper venue and lack of personal jurisdiction are waived if not raised in the first responsive motion. Defendants may also move to dismiss on the ground that the patent claims are ineligible under 35 U.S.C. § 101—an early, increasingly common attack, especially against software and business-method patents, that can end a case before the expense of discovery (Part VII). If Vortex believes Brightline's claims merely recite the abstract idea of "adjusting a measurement based on changing conditions," it may file a Section 101 motion at the outset and try to win the whole case in a single brief.

Answer and counterclaims. If the defendant does not move to dismiss—or after such a motion is resolved—it must file an answer responding to each allegation and asserting affirmative defenses. Common patent defenses include non-infringement; invalidity (anticipation under § 102, obviousness under § 103, or ineligibility under § 101); failure to satisfy the written-description, enablement, or definiteness requirements of § 112; unenforceability for inequitable conduct; and equitable defenses such as estoppel, patent misuse, and exhaustion. Crucially, the answer often includes counterclaims for a declaratory judgment of non-infringement and invalidity. These give the defendant its own affirmative claim: if Brightline later tried to drop the suit to dodge an adverse ruling, Vortex's counterclaims would let it press on to a binding judgment that Brightline's patent is invalid or not infringed—a result affecting Brightline's ability to enforce the patent against anyone. A plaintiff cannot simply walk away from a case it is losing once the defendant has planted that flag.

The scheduling conference and Rule 16

Once the pleadings settle, the case enters a phase governed by Rule 16, which directs the court to hold a scheduling conference and enter a scheduling order setting deadlines for amendments, discovery, motions, and trial. In patent-heavy districts, this is where the patent local rules kick in, layering a sequence of infringement and invalidity contentions and a claim-construction schedule on top of the ordinary calendar. Rule 16 may sound like housekeeping, but the scheduling order is the spine of the case—it determines how much runway each side has, when the expensive phases hit, and how much pressure the calendar exerts toward settlement. A fast schedule favors a plaintiff forcing a quick resolution; a slower one can give a cash-strapped small business room to breathe. The full lifecycle of filings is mapped in navigating the paper trail: a comprehensive guide to federal civil litigation filings.

Part IV: Discovery

Overview

Discovery is the formal process by which the parties exchange information—and gather it from third parties—to prepare for trial. In patent litigation, it is usually the longest and most expensive phase, frequently consuming sixty to seventy percent of the total budget. The reasons are structural: the technology is complex, the volume of electronic documents enormous, and the case depends on specialized expert analysis that itself requires extensive underlying data. This is where a well-resourced opponent can try to win by attrition, and where a small business must be most disciplined about cost—the single place where a David-versus-Goliath patent case is most often decided not by the merits but by who can outspend whom.

Initial disclosures and contentions

Discovery opens with mandatory initial disclosures under Rule 26(a)(1): each side must, without waiting for a request, identify likely witnesses, the categories and locations of relevant documents, a computation of each category of damages, and any applicable insurance. Many patent-heavy courts—the Eastern and Western Districts of Texas, the District of Delaware—layer additional, patent-specific disclosure requirements on top through their local patent rules.

Under those rules, the patentee typically must serve detailed infringement contentions early (often within a couple of weeks of the initial case-management conference), spelling out claim-by-claim and element-by-element how the accused product meets each limitation—essentially a formalized version of the claim chart from Part I. The accused infringer must then serve invalidity contentions, identifying the prior-art references and the anticipation and obviousness arguments it will rely on. These requirements do real work: they force both sides to commit to concrete theories early, preventing the "shifting sands" problem in which a party quietly changes its story (amending later usually requires good cause), and they let each side aim its discovery at the theories actually in play rather than fishing in the dark. For Brightline, the discipline of writing out precisely how the Vortex sensor meets each claim element—before discovery—often reveals the strengths and gaps in its own case while there is still time to act.

Discovery tools

After disclosures, the parties use several standard mechanisms, each with its own rule and strategy. Interrogatories are written questions answered under oath, presumptively limited to 25 per party under Rule 33; the most consequential ask the opponent to state its full infringement or invalidity theory. Requests for production under Rule 34 call for documents and electronically stored information—in patent cases, often massive productions of source code, design files, specifications, emails, and financial records. Requests for admission under Rule 36 lock down undisputed facts (a document's authenticity, a sale's date) and narrow what must be proved at trial. Depositions under Rules 30 and 31 are live, sworn, transcribed question-and-answer sessions; in patent cases they typically include inventors, engineers, executives, licensing personnel, and—critically—expert witnesses, and they are often the most expensive and outcome-shaping events in discovery. And subpoenas under Rule 45 reach non-parties for documents or testimony—useful when a former Vortex engineer who designed the accused sensor has since left the company.

Each of these tools rewards craft, in how you use them and how you respond. We treat the techniques in dedicated guides: mastering interrogatories, mastering document discovery and responding to requests for production, strategic responses to requests for admission, and—because depositions so often make or break a patent case—the art and science of depositions in federal civil litigation and its companion on defending depositions in federal court.

Proportionality and e-discovery management

The most important cost-control tool in modern discovery is the proportionality standard. The 2015 amendments to Rule 26(b)(1) restored proportionality to the very definition of the scope of discovery, which now reaches only matter "relevant to any party's claim or defense and proportional to the needs of the case." The rule directs courts to weigh the importance of the issues, the amount in controversy, the parties' relative access to information, their resources, the importance of the discovery to the issues, and whether its burden outweighs its likely benefit. For a small business facing a larger opponent, proportionality is a shield—and the "parties' resources" factor invites the court to consider precisely the lopsided economics a Goliath might hope to exploit. If Vortex serves sweeping demands plainly designed to bury a twelve-person company in document review, Brightline can invoke proportionality—pointing to the asymmetric resources and the marginal relevance of the requests—and ask the court to narrow them or shift some of the cost.

Technology can also help. Technology-assisted review (TAR, or predictive coding) uses machine learning to prioritize likely-relevant documents, dramatically reducing review costs in large productions; courts have broadly accepted it as defensible since Da Silva Moore v. Publicis Groupe first blessed the approach. Where volume is large, ask counsel whether TAR—or even a tightly negotiated set of search terms and custodians—can bring review costs down to something survivable. The parties' Rule 26(f) conference, held before the scheduling order, is the moment to negotiate the protocols that govern the whole production; getting them right early prevents the most expensive disputes later.

Protective orders

Patent discovery routinely surfaces a company's most sensitive material—source code, trade secrets, manufacturing know-how, customer lists, pricing, and financials. To allow this exchange without destroying the information's value, courts enter protective orders under Rule 26(c) that create tiers of confidentiality, most commonly "Confidential" and the more restrictive "Attorneys' Eyes Only" (AEO), the latter limiting access to outside counsel and retained experts so competitors never see one another's crown jewels directly. Source code is often subject to an even tighter regime—reviewed only on a standalone, non-networked computer, with printouts limited and logged.

For Brightline, the protective order is what makes it safe to turn over the very calibration source code that embodies its edge: Vortex's businesspeople never see it, only its outside lawyers and a court-approved expert, under penalty of contempt. It is also where patent litigation and trade-secret protection intersect: the same know-how Brightline guards as a trade secret may have to be produced, and only the protective order stands between production and competitive ruin. Companies that build a disciplined trade-secret protection program before any dispute arises are far better positioned to defend that information when discovery demands it.

Part V: Claim Construction (the Markman Hearing)

Why claim construction is critical

If a patent case has a single pivot point, it is usually claim construction—the process by which the court decides what the patent claims actually mean. Because infringement requires that the accused product meet every limitation of a claim, and invalidity depends on whether the prior art discloses those same limitations, the court's reading of a few disputed words can decide the entire case. A broad construction may make infringement easy to prove but exposes the patent to prior art; a narrow construction may dodge the prior art but lets the accused product slip free. The two sides are, in effect, arguing over the size of the fence around the invention—to the judge, alone, in a proceeding that often determines who settles and on what terms.

The process takes its name from Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996), where the Supreme Court held that claim construction is a question of law for the judge, not a question of fact for the jury. The consequences are enormous: the judge decides what the patent means; that decision can dispose of the case at summary judgment; and it can be reviewed on appeal without deference to any jury.

The claim construction framework

The governing framework comes from the Federal Circuit's en banc decision in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005). Under Phillips, claim terms receive their "ordinary and customary meaning" as understood by a person of ordinary skill in the relevant art at the time of the invention. Courts find that meaning by working outward through a hierarchy of evidence. The intrinsic record comes first and controls: the claim language itself; then the specification—the written description and drawings, which Phillips called "the single best guide to the meaning of a disputed term" and which may expressly define a term or disclaim scope; and then the prosecution history. Only when the intrinsic record leaves genuine ambiguity may a court turn to extrinsic evidence—dictionaries, treatises, and expert testimony—which may inform but can never override the intrinsic record. Phillips cautioned against two recurring temptations: improperly importing limitations from the specification into the claims, and giving extrinsic evidence more weight than the patent's own words.

Consider Brightline's disputed term, "measured rate of humidity change." Brightline will argue it broadly covers any computation of how fast humidity is shifting, including Vortex's approach. Vortex will comb the specification and prosecution history for anything narrowing it—perhaps a statement during prosecution that the rate is "continuously sampled," which Vortex's timer-based device does not do. The judge's choice between those readings may decide whether Vortex infringes at all. This is why claim construction is sometimes called the trial-within-the-trial: a few pages of briefing about a single phrase can be worth more than weeks of testimony about the products.

Strategic considerations

Several strategic threads run through claim construction. First, which terms to dispute is itself a decision; parties usually identify five to fifteen terms, and courts increasingly cap the number, so a small business should husband its resources for the terms that control the outcome rather than fighting over every adjective. Second, timing varies by court. Some judges hold the Markman hearing early, before substantial discovery, which can drive a quick settlement once one side sees the writing on the wall; others delay it until fact discovery is complete. For a cost-conscious small business, an early Markman ruling can be a gift—it may resolve the case before the most expensive discovery is ever incurred.

Third, the standard of review on appeal favors challenging a construction: the Federal Circuit reviews claim construction de novo, although under Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 574 U.S. 318 (2015), any subsidiary findings of fact—what a skilled artisan would have understood a term to mean—are reviewed only for clear error. Because de novo review invites appeal, and reversal and modification rates have historically been meaningful, even a decisive Markman win carries a tail of uncertainty all the way to the Federal Circuit. A favorable construction is enormously valuable, but it is not the final word, and any settlement valuation should discount for the real chance the appellate court reads the claims differently.

Part VI: Inter Partes Review at the PTAB

A strategic alternative

Inter partes review (IPR) has become one of the most consequential tools in patent litigation since the America Invents Act created it in 2011. An IPR is an administrative proceeding before the Patent Trial and Appeal Board (PTAB), a tribunal within the USPTO, in which a challenger asks expert administrative patent judges to cancel an issued patent's claims as unpatentable. It runs in parallel to, or instead of, a district court fight, at a fraction of the cost and time. For a defendant—Brightline accused of infringing a Vortex patent, or Vortex defending against Brightline—it can be the most powerful card in the deck: it lets the challenger attack validity before specialists, under a friendlier standard of proof, often while pressing the court to pause its own proceedings.

Cost and timeline

IPRs are dramatically cheaper than litigation. Industry surveys put the median cost of preparing and filing a petition in the low-to-mid six figures, and carrying an IPR through the oral hearing and final written decision well below the cost of district-court litigation through trial. They are also faster: by statute, the PTAB must issue a final written decision within twelve months of instituting review, extendable to eighteen for good cause—against the two-to-three-plus years a district court case typically takes to reach trial. For a cash-conscious small business, that combination—lower cost, faster answer—is often decisive. A successful IPR that cancels the asserted claims ends the infringement case entirely, because there is no longer a valid claim to infringe.

Key procedural rules

A few features of IPR drive strategy. The petition must be filed within one year of the petitioner (or its privy) being served with a complaint alleging infringement of the challenged patent (35 U.S.C. § 315(b)). That clock forces an early, irreversible decision. The PTAB applies a preponderance of the evidence standard to invalidity—"more likely than not"—meaningfully lower than the clear and convincing evidence standard that governs in district court under Microsoft Corp. v. i4i Ltd. Partnership, 564 U.S. 91 (2011). That lower bar, plus a panel of technically trained judges rather than a lay jury, is a major reason challengers favor the forum. But the IPR's scope is limited: it reaches only anticipation (§ 102) and obviousness (§ 103), and only on patents and printed publications. Other theories—Section 101 ineligibility, Section 112 indefiniteness or lack of enablement, prior public use or sale—cannot be raised in an IPR and must stay in court. A complete invalidity strategy therefore usually runs on two tracks at once.

Discretionary denial: the 2025 overhaul

Even when a petition presents strong invalidity arguments, the PTAB has discretion under 35 U.S.C. § 314(a) to decline to institute review—one of the most volatile areas in patent law. The framework traces to Apple Inc. v. Fintiv, Inc., IPR2020-00019 (PTAB Mar. 20, 2020) (precedential), which directed the Board to weigh several factors when a parallel district court case is pending: whether a stay has been or is likely to be granted, how close the trial date is to the Board's projected decision, how much the parties and court have already invested, the overlap between the IPR and the lawsuit, whether petitioner and defendant are the same, and other circumstances such as the merits.

This area changed substantially in 2025, and any older description is now unreliable. In June 2022, then-Director Kathi Vidal had issued a memorandum that narrowed discretionary denials—making a Sotera stipulation (a binding promise not to re-raise in district court the specific grounds litigated in the IPR) effectively dispositive against denial, and shielding petitions presenting "compelling merits." On February 28, 2025, Acting Director Coke Morgan Stewart rescinded that memorandum. On March 24, 2025, the PTAB's Chief Administrative Patent Judge issued guidance that the Fintiv factors are fully back in play (including alongside a parallel International Trade Commission proceeding), that a Sotera stipulation is "highly relevant" but no longer dispositive, and that strong merits alone no longer foreclose denial. Then, on March 26, 2025, the Acting Director created a novel bifurcated institution process: first, the Director (with at least three PTAB judges) decides whether discretionary denial is appropriate; only if it is not does a three-judge panel reach the merits. The early consequence has been a sharp rise in discretionary denials—by mid-2025, substantially more procedural denials than in the comparable prior period.

For a small business, the lesson is twofold. IPR is no longer the near-automatic option it briefly appeared to be; whether the Board institutes now turns heavily on timing, on the posture of any parallel litigation, and on policy choices that shift with USPTO leadership. And because the rules move quickly, ask counsel about the current state of discretionary-denial practice rather than relying on any guide—including this one—as the last word. If Vortex sues Brightline in a fast Texas court with a trial date twelve months out, the Fintiv factors may push the Board to deny a late-filed IPR, leaving Brightline to litigate invalidity in court under the tougher standard before a jury. The takeaway: file early, and pair the petition with a Sotera stipulation and a stay motion, before the parallel litigation builds the momentum that invites a denial.

Estoppel

The final strategic consideration is estoppel. Under 35 U.S.C. § 315(e)(2), once an IPR reaches a final written decision, the petitioner is barred from later asserting in district court any invalidity ground it "raised or reasonably could have raised" in the IPR. After SAS Institute Inc. v. Iancu, 138 S. Ct. 1348 (2018)—which requires the Board to decide the patentability of all challenged claims if it institutes at all—this estoppel expanded, because the final written decision now covers every claim and ground the petitioner put forward. The petition must therefore lead with the petitioner's best prior art, because anything that reasonably could have been included but was held back is lost in court. A defendant cannot run a half-hearted IPR on weak references and quietly save the strong ones for trial. IPR strategy and the district court defense must be coordinated from the start—ideally by the same team thinking several moves ahead. Get the sequencing wrong and a small business can find itself estopped from making its best invalidity argument in the very forum where it needs it most.

Part VII: Patent Eligibility Under Section 101

The Alice/Mayo framework

Patent eligibility under 35 U.S.C. § 101 has become one of the most significant—and least predictable—issues in patent litigation. Section 101 says what kinds of things can be patented at all (processes, machines, manufactures, and compositions of matter), and the courts have long recognized three implicit exceptions: laws of nature, natural phenomena, and abstract ideas. The modern test comes from Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012), and Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014). At step one, the court asks whether the claim is "directed to" one of those exceptions—an abstract idea, say. If it is, step two asks whether the claim adds an "inventive concept": something more than the idea itself, enough to make it a patent-eligible application rather than a monopoly on the idea. Generic computer implementation—doing an old practice "on a computer"—does not suffice.

The framework has hit software, business-method, and diagnostic patents especially hard. Defendants raise Section 101 routinely, and courts increasingly resolve eligibility on a motion to dismiss or at summary judgment—before the cost of a full trial—because eligibility is ultimately a question of law (albeit one that, after Berkheimer v. HP Inc., 881 F.3d 1360 (Fed. Cir. 2018), can turn on factual disputes about whether a claim element was "well-understood, routine, and conventional," sometimes precluding early dismissal). For Brightline, this is a first-order risk assessment. If its claims could be cast as merely "the abstract idea of adjusting a reading based on a changing condition," implemented on a generic processor, Vortex may attack them under Alice at the outset, and the suit could be dismissed before discovery. Whether a claim survives often turns on how the specification ties the invention to a concrete technical improvement—here, perhaps an unconventional calibration technique that measurably improves sensor accuracy under real-world humidity drift, the kind of advance the Federal Circuit found eligible in Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016). The lesson reaches back to drafting: the best defense against an Alice motion is often written years earlier, in a specification that explains why the invention is a technical advance and not just an old idea wearing a microprocessor. Our dedicated treatment, patent eligibility after Alice, walks through the surviving strategies in detail.

Current developments

The law here remains genuinely unsettled, and a small business with patents in affected fields should track it. The Supreme Court has repeatedly declined to revisit Section 101, despite years of pleas from Federal Circuit judges, practitioners, and industry groups for clearer guidance. In the meantime, the USPTO has tried to supply predictability administratively. Its 2024 guidance on AI inventions clarified that using AI to make an invention does not by itself render the claims ineligible, and that the analysis focuses on what the claims cover rather than how the invention was created. Subsequent guidance directed examiners to issue Section 101 rejections only when a claim is more likely than not ineligible. (Whether an AI-assisted invention can even be patented—and who counts as the inventor—is a related question we take up in AI-generated inventions: who owns what the machine creates.)

On the legislative front, the Patent Eligibility Restoration Act of 2025 (PERA, S. 1546)—introduced by Senators Thom Tillis and Chris Coons—would eliminate the judge-made exceptions entirely and replace them with a closed list of narrow statutory exclusions (a mathematical formula standing alone, a process performed solely in the human mind, an unmodified human gene, unmodified natural material). PERA would, in effect, override Mayo, Myriad, and Alice. It has bipartisan support and drew a Senate Judiciary subcommittee hearing in late 2025, but it faces real opposition—from parts of the technology sector that see Alice/Mayo as a useful filter against overbroad software patents, and from some in the medical and patient-advocacy communities worried about patents on diagnostic correlations and isolated genetic material. With Senator Tillis having announced he will not seek reelection, the bill's path remains uncertain as of early 2026. The takeaway: treat Section 101 as a live, shifting risk, and design both the patent portfolio (specifications that emphasize technical improvement) and litigation strategy (anticipating an early Alice motion) with that volatility in mind.

Part VIII: Summary Judgment, Pre-Trial, and Trial

Summary judgment

After discovery and claim construction, the parties commonly move for summary judgment under Rule 56, arguing that there is no genuine dispute as to any material fact and that the movant is entitled to judgment as a matter of law. The court views the evidence in the light most favorable to the non-movant; summary judgment is granted only when, on that view, no reasonable jury could find for that party—the standard the Supreme Court set in its 1986 "trilogy" of Celotex Corp. v. Catrett, Anderson v. Liberty Lobby, Inc., and Matsushita Electric Industrial Co. v. Zenith Radio Corp. In patent cases, these motions typically attack non-infringement (under the court's construction, the accused product lacks a required limitation), invalidity (the prior art anticipates or renders the claims obvious, or the claims are ineligible under § 101 or indefinite under § 112), or damages (the opponent's theory rests on a legally deficient methodology).

These motions are heavy, supported by deposition excerpts, expert reports, documents, and declarations. Courts are often cautious about granting summary judgment where technical disputes are genuinely contested—those usually go to the jury. But the motion is powerful when the facts simply do not support the other side's legal theory, and even an unsuccessful motion can pay dividends: the briefing narrows the issues and frequently prompts settlement once each party sees how its case reads on paper. If, after the Markman ruling, the court has construed "measured rate of humidity change" to require continuous sampling, Vortex's timer-based device may not infringe as a matter of law, and Vortex may win the case on summary judgment without ever reaching trial. This is the moment a favorable claim construction cashes out—and the reason a Markman loss so often drives the losing side to settle before the summary-judgment briefs are even filed.

Pre-trial proceedings

As trial nears, the parties complete a series of procedural steps. Under Rule 26(a)(3) they make pretrial disclosures—the witnesses and exhibits they expect to present—and prepare a joint pretrial order identifying contested and uncontested facts, witness and exhibit lists, objections, proposed jury instructions, and the issues to be tried; in patent cases it also lays out the asserted patents, accused products, and damages theories. They file motions in limine—pretrial requests to admit or exclude specific evidence—which matter enormously. Frequent subjects include motions to exclude expert testimony that fails the reliability standard of Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), now codified in Federal Rule of Evidence 702 (especially damages experts whose royalty methodology is untethered from the patented feature); motions to keep out revenue from products not accused of infringement (often unfairly prejudicial under Federal Rule of Evidence 403 because it inflates the jury's sense of the damages); motions on the admissibility of the patentee's prior licenses; and motions to bar evidence not timely disclosed, invoking Rule 37(c)(1)'s preclusion of undisclosed evidence. Winning or losing a key motion in limine can quietly decide a trial before the first witness is sworn—exclude the other side's damages expert, and a multimillion-dollar claim can deflate to a nominal one.

Trial

Patent trials are unusual animals. The technology is often genuinely difficult, and the jury must be taught enough to evaluate the evidence before it can decide anything. Trial begins with jury selection (voir dire), where counsel try to seat jurors who can engage with technical material and set aside biases. A good trial team will have spent months figuring out how to make a humidity-calibration algorithm comprehensible to twelve people who have never read a patent—tutorials, animations, analogies, and a focus on the human story.

The patentee's case-in-chief typically addresses three things. Validity: although a patent is presumed valid under 35 U.S.C. § 282 (the statutory source of the heavy "clear and convincing" burden the i4i Court enforced), the patentee defends validity if the accused infringer has put it in issue. Infringement: that the accused product meets every limitation of at least one asserted claim, literally or under the doctrine of equivalents—which can reach products performing substantially the same function in substantially the same way to achieve substantially the same result, subject to the prosecution-history-estoppel limits from Part I. Damages: the economic harm, quantified under the applicable theory (Part IX). The witnesses usually include the inventor, technical experts on infringement and validity, a damages expert, and company witnesses on commercial importance. The accused infringer's case mirrors this—pressing invalidity with prior art and expert testimony, arguing non-infringement on specific limitations, and offering a competing damages calculation. For Brightline, putting its founder-inventor on the stand to explain the late nights that produced the calibration breakthrough can be as persuasive as any expert; juries respond to the human story behind a small company's invention.

Part IX: Damages and Injunctive Relief

Damages theories

Patent damages are governed by 35 U.S.C. § 284, which directs courts to award damages "adequate to compensate for the infringement, but in no event less than a reasonable royalty." Two theories dominate.

Reasonable royalty. This is the most common measure, and the statutory floor. It represents what the patent owner and the infringer would have agreed to pay in a hypothetical, arm's-length negotiation just before infringement began—assuming both sides were willing and the patent valid and infringed. Courts analyze it through the fifteen-factor framework of Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970), which weighs established royalties the patentee has charged, rates paid for comparable patents, the scope and term of the hypothetical license, the commercial relationship between the parties, and—critically—the share of profit attributable to the patented feature rather than the product's many unpatented components. That last point, apportionment, is where many damages cases are won and lost. The Federal Circuit has insisted, in LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51 (Fed. Cir. 2012), and VirnetX, Inc. v. Cisco Systems, Inc., 767 F.3d 1308 (Fed. Cir. 2014), that the royalty be tied to the value of the patented feature, not the entire product, unless that feature drives demand for the whole (the narrow "entire market value rule"). If Brightline's patent covers only the calibration module, its royalty should track the value that module adds, not the entire price of Vortex's sensor—and a damages expert who ignores that distinction invites a Daubert exclusion that can gut the case.

Lost profits. When the patentee actually makes and sells a competing product, it may recover the profits it would have earned "but for" the infringement. Under the Panduit test (Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978)), the patentee must show (1) demand for the patented product, (2) an absence of acceptable non-infringing alternatives, (3) the capacity to have met the demand, and (4) the amount of profit it would have made. Lost-profits awards are usually larger than reasonable royalties but harder to prove—and largely unavailable to patentees who do not sell products, such as many patent-assertion entities. Because Brightline manufactures and sells its sensor, it may argue lost profits: every greenhouse that bought a Vortex unit would have bought a Brightline unit instead. Vortex will counter that acceptable non-infringing alternatives existed, breaking the "but for" chain. Lost profits is an underappreciated advantage a practicing small business holds over a non-practicing entity: it can credibly claim the larger measure of damages and, as we will see, a real shot at an injunction.

Enhanced damages and willful infringement

Under Section 284, a court may increase damages up to three times the compensatory award. Enhanced damages are reserved for willful infringement, and after Halo Electronics, Inc. v. Pulse Electronics, Inc., 579 U.S. 93 (2016)—which discarded the rigid two-part Seagate test—district courts have broad discretion to award them on the totality of the circumstances. Willfulness is, in essence, deliberate or reckless infringement: conduct beyond the typical case, often an infringer who knew of the patent and pressed ahead anyway, "the wanton and malicious pirate" the Halo Court had in mind. This is why pre-suit notice matters so much: if Brightline sent Vortex a detailed letter identifying the patent and the accused product, and Vortex neither stopped nor obtained a competent opinion of counsel, Brightline's willfulness case—and its shot at treble damages—grows considerably stronger. Conversely, an accused infringer who promptly obtains a reasoned non-infringement or invalidity opinion from competent counsel builds a record that cuts hard against willfulness—not a magic shield, but one of the best defenses against an enhancement.

Injunctive relief

For many patentees, stopping the infringement matters more than collecting money. Injunctive relief is governed by eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006), which abolished the old rule that an injunction followed almost automatically from a finding of infringement. Now the patentee must satisfy the traditional four-factor equity test: (1) irreparable harm; (2) that monetary damages are inadequate; (3) that the balance of hardships favors the patentee; and (4) that the public interest would not be disserved. The practical effect depends heavily on who the patentee is. A company that does not make products and only licenses patents will rarely show irreparable harm and usually ends up with ongoing royalties—the result eBay was largely designed to produce. But a small manufacturer that competes head-to-head with the infringer—exactly Brightline's situation—stands a far better chance, because losing sales, customers, and market position to a direct competitor is the kind of harm courts recognize as hard to undo, especially for a young company trying to establish itself. For Brightline, a permanent injunction barring Vortex from selling the infringing sensor could be worth far more than any damages check: it removes a competitor from the field and restores the exclusivity the patent was supposed to provide. The mechanics of remedies and the road from a denied summary-judgment motion through trial to post-trial relief are traced in our comprehensive guide to patent infringement litigation, from summary judgment denial to post-trial.

Part X: Post-Trial Motions and Appeals

Post-trial motions

A verdict is not the end. Several motions commonly follow. A renewed motion for judgment as a matter of law (JMOL) under Rule 50(b)—which must have been preceded by a Rule 50(a) motion before the case went to the jury, or the point is forfeited—argues that no reasonable jury could have reached the verdict on the evidence. It is a high bar, but a real one, and patent verdicts are overturned on JMOL with some regularity, particularly on damages and obviousness. A motion for a new trial under Rule 59 argues that errors (a misstated jury instruction, improperly admitted or excluded evidence, a verdict against the clear weight of the evidence) warrant starting over; a court may also order a remittitur, conditioning denial of a new trial on the plaintiff's acceptance of a reduced award. And the prevailing party may move for attorney fees under Section 285 if the case was "exceptional" under Octane Fitness, while a victorious patentee may move for enhanced damages under Halo. These motions can reshape an outcome substantially: a damages award can be slashed or set aside, and a defense verdict reinforced or unwound—all before anyone reaches the appellate court.

Federal Circuit appeals

Patent appeals go to a single, specialized court: the U.S. Court of Appeals for the Federal Circuit, which Congress created in 1982 and gave nationwide jurisdiction over patent cases to bring uniformity to a body of law that had splintered among the regional circuits. It applies different standards of review to different issues, and understanding them is key to an appeal's odds. Claim construction is reviewed de novo—fresh, with no deference—subject to clear-error review of any subsidiary factual findings under Teva. A jury's findings on factual questions, including infringement and the factual underpinnings of validity, are reviewed for substantial evidence, a deferential standard asking only whether a reasonable jury could have reached the result. Purely legal conclusions, such as Section 101 eligibility and the ultimate question of obviousness, are reviewed de novo, while the Octane Fitness exceptional-case finding and the eBay injunction decision draw only abuse-of-discretion review.

The de novo review of claim construction is the headline. Because construction so often controls the case, and the Federal Circuit has historically reversed or modified constructions at a meaningful rate, a judgment that looks final can be undone if the appellate court reads the claims differently. This is why neither a win nor a loss at trial is ever quite the last word, and why settlement talks often continue through the appeal. A defendant who lost at trial but has a strong construction argument may rationally fight on; a plaintiff sitting on a verdict built atop a shaky construction should think hard about settling before the panel weighs in.

The appeal begins with a notice of appeal, filed within 30 days of the final judgment. The parties then brief the case—opening, response, and reply—and the court may hold oral argument; from notice to decision typically takes twelve to eighteen months. A party unhappy with the result may petition the U.S. Supreme Court for certiorari, but the Court grants very few patent petitions—and, because the Federal Circuit has exclusive jurisdiction, the circuit splits that ordinarily prompt review are rare.

Enforcement of judgment

Winning a judgment is not the same as collecting on it. A money award may have to be enforced under Rule 69 through writs of execution, garnishment, liens, or other state-law mechanisms if the loser will not pay. A permanent injunction requires ongoing compliance monitoring and can lead to civil contempt proceedings if the infringer keeps selling or rolls out a redesign no more than "colorably different" from the enjoined product—the standard the Federal Circuit set in TiVo Inc. v. EchoStar Corp., 646 F.3d 869 (Fed. Cir. 2011) (en banc). Where no injunction issues, the court may set an ongoing royalty for continued use—effectively a compulsory, court-set license, often at a higher rate than the jury's reasonable royalty because the infringement is now adjudicated rather than hypothetical. And enforcement gets materially harder when the infringer sits abroad, requiring navigation of treaties and foreign legal systems that may not honor U.S. patent rights at all. If Brightline wins but Vortex routes its infringing sales through an overseas affiliate, collecting the judgment and policing the injunction can become a second campaign in its own right.

Conclusion: A decision framework for small businesses

Federal civil litigation—and patent litigation in particular—is a high-stakes, resource-intensive process that rewards careful planning and punishes improvisation. The decision to litigate should rest on a clear-eyed assessment of cost, risk, and likely outcome, not on the heat of the moment when an infringer is discovered.

Before committing, a small business should work through a structured set of questions. Assess the merits: is there a good-faith basis for the claim or defense, grounded in a real analysis of the claims, the accused product, and the prior art—robust enough to survive Rule 11 and an early Alice motion? Quantify the cost: using benchmarks like the AIPLA survey, what is the realistic spend through each phase, where do the big spikes fall (discovery and claim construction), and can the business sustain it without starving the operations that make the patent worth fighting for? Evaluate alternatives: can the dispute be resolved through licensing, mediation, or arbitration at a fraction of the cost—and will the other side engage in good faith? Consider the PTAB: if you are the accused infringer, can an IPR eliminate or narrow the asserted claims cheaply and quickly—bearing in mind that discretionary-denial practice tightened in 2025 and continues to shift, and that estoppel forces you to lead with your best art? Weigh the business impact: beyond legal fees, what are the indirect costs—executive distraction, the disclosure of competitive intelligence in discovery, reputational effects, and the disruption of running a small company while fighting a lawsuit?

For Brightline, the answers might point in any direction. If the Vortex sensor clearly infringes a strong patent, Brightline makes a directly competing product, and Vortex has real sales to lose, the case for suing—and seeking an injunction—is compelling, because Brightline holds two cards a non-practicing entity lacks: lost profits and a credible irreparable-harm story under eBay. If the infringement is doubtful, the patent is vulnerable under Section 101 or to strong prior art, and the disputed sales are small, the wiser path may be a license, a design-around, or simply moving on. The point is not that small businesses should never litigate; sometimes a lawsuit is the only way to protect a core asset. It is that the decision should be made deliberately, with experienced counsel, a realistic budget, and a candid view of the risks at each stage. Patent litigation is not a process to enter lightly—but with a sound strategy, the right team, and a clear understanding of the road ahead, it is one a small business can navigate, whether enforcing its own patent rights or defending against the claims of others.

Frequently Asked Questions

How much does a patent lawsuit really cost a small business? More than you think. Median per-side costs run from roughly $300,000 through discovery and claim construction in the smallest cases to several million through trial and appeal in larger ones, per the AIPLA's biennial survey. The two most expensive phases are discovery (often 60–70% of the total) and claim construction. Because most cases settle, many never incur full trial costs—but budget as though they will, and ask counsel about contingency, hybrid, capped-fee, and litigation-funding arrangements that shift risk off your balance sheet.

Can I sue an infringer in my own state where my business is located? Usually not, unless the infringer has a foothold there. After TC Heartland (2017), patent venue under 28 U.S.C. § 1400(b) is limited to (1) the defendant's state of incorporation, or (2) a district where the defendant both commits infringement and maintains a "regular and established place of business." Venue follows the defendant's footprint, not the plaintiff's. (Foreign defendants are an exception and may be sued anywhere.)

What is a Markman hearing, and why does everyone say it decides the case? It is the proceeding where the judge—not the jury—construes the disputed words of the patent claims, a question of law under Markman (1996) and Phillips (2005). Because infringement and invalidity both depend on what the claims mean, the construction of a few key terms frequently dictates who wins on summary judgment—which is why a Markman ruling so often triggers settlement.

If I'm accused of infringement, is inter partes review still a good option? It can be, but it is no longer the near-automatic path it briefly appeared to be. IPR is faster and cheaper than court and applies a lower (preponderance) standard, but it reaches only § 102/§ 103 invalidity based on patents and printed publications, carries a hard one-year deadline after you are served, and triggers estoppel on grounds you raised or reasonably could have raised. After the 2025 overhaul of discretionary-denial practice, institution turns heavily on timing and parallel litigation—so file early, pair the petition with a Sotera stipulation, and ask counsel about the current rules.

Will I be able to stop the infringer, or just collect money? It depends on who you are. Under eBay (2006), an injunction no longer follows automatically from infringement; you must prove irreparable harm and the inadequacy of money, among other factors. A non-practicing entity rarely clears that bar and ends up with ongoing royalties. But a small manufacturer competing directly with the infringer—losing sales and market share—has a genuine shot at a permanent injunction, which can be worth more than any damages award.

Could I end up paying the other side's legal fees? Yes, in an "exceptional" case. Under 35 U.S.C. § 285 and Octane Fitness (2014), a court may shift fees to the prevailing party when the case "stands out" for the weakness of a position or the unreasonable way it was litigated. This cuts both ways: a baseless claim or a frivolous defense can expose the losing party to a substantial fee award—one more reason to do the merits analysis honestly before filing.

What is a litigation hold, and when do I have to issue one? A written instruction telling everyone in the company to stop deleting and start preserving documents and data relevant to a dispute. The duty attaches as soon as litigation is "reasonably anticipated"—which can be before any complaint is filed. Failing to preserve electronically stored information can lead to sanctions under Rule 37(e), up to an adverse-inference instruction that tells the jury to assume the lost evidence was unfavorable. Issue the hold promptly and document it.

Related Articles

This article is for informational purposes only and does not constitute legal advice. The law here changes frequently—sometimes abruptly, as the 2025 shifts in PTAB practice illustrate. Consult experienced counsel and check the current rules, statutes, and case law in your jurisdiction before acting.

Selected Authorities

Statutes: 28 U.S.C. § 1338 (federal patent jurisdiction); 28 U.S.C. § 1400(b) (patent venue); 28 U.S.C. § 1404(a) (transfer of venue); 35 U.S.C. § 101 (patent eligibility); §§ 102–103 (anticipation and obviousness); § 112 (written description, enablement, definiteness); § 282 (presumption of validity); § 284 (damages and enhanced damages); § 285 (attorney fees); § 294 (arbitration of patent disputes); §§ 311–319 (inter partes review), including § 314(a) (institution discretion), § 315(b) (one-year bar), and § 315(e)(2) (estoppel).

Cases: Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996); Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc); Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 574 U.S. 318 (2015); TC Heartland LLC v. Kraft Foods Group Brands LLC, 581 U.S. 258 (2017); In re Cray Inc., 871 F.3d 1355 (Fed. Cir. 2017); Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007); Ashcroft v. Iqbal, 556 U.S. 662 (2009); Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722 (2002); MedImmune, Inc. v. Genentech, Inc., 549 U.S. 118 (2007); Mayo Collaborative Services v. Prometheus Laboratories, Inc., 566 U.S. 66 (2012); Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014); Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016); Berkheimer v. HP Inc., 881 F.3d 1360 (Fed. Cir. 2018); Microsoft Corp. v. i4i Ltd. Partnership, 564 U.S. 91 (2011); SAS Institute Inc. v. Iancu, 138 S. Ct. 1348 (2018); Octane Fitness, LLC v. ICON Health & Fitness, Inc., 572 U.S. 545 (2014); Highmark Inc. v. Allcare Health Management System, Inc., 572 U.S. 559 (2014); Halo Electronics, Inc. v. Pulse Electronics, Inc., 579 U.S. 93 (2016); eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006); Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116 (S.D.N.Y. 1970); Panduit Corp. v. Stahlin Bros. Fibre Works, Inc., 575 F.2d 1152 (6th Cir. 1978); LaserDynamics, Inc. v. Quanta Computer, Inc., 694 F.3d 51 (Fed. Cir. 2012); VirnetX, Inc. v. Cisco Systems, Inc., 767 F.3d 1308 (Fed. Cir. 2014); TiVo Inc. v. EchoStar Corp., 646 F.3d 869 (Fed. Cir. 2011) (en banc); Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993); Apple Inc. v. Fintiv, Inc., IPR2020-00019 (PTAB Mar. 20, 2020) (precedential).

Rules and agency materials: Fed. R. Civ. P. 8, 11, 12, 16, 26, 33, 34, 36, 37(c)(1), 37(e), 50, 56, 59, 69; Fed. R. Evid. 403, 702; USPTO discretionary-denial memoranda and guidance (Feb. 28, 2025 rescission; Mar. 24, 2025 PTAB guidance; Mar. 26, 2025 "Interim Processes for PTAB Workload Management").

Secondary and industry sources: AIPLA Report of the Economic Survey (litigation and IPR cost benchmarks); contemporary patent-litigation venue reporting (2024–2025) on filing concentration in the Eastern and Western Districts of Texas and the District of Delaware; Patent Eligibility Restoration Act of 2025, S. 1546 (119th Cong.).