A patent is a right to exclude, but the state that grants the right does not enforce it. Enforcement is the holder’s burden, carried through litigation the holder funds, and patent litigation is among the most expensive, slowest, and riskiest forms of civil suit. This is the reality that stood behind the small enforcement probability the moat article put in its model and the founder article leaned on. A patent that cannot be enforced protects nothing, and whether a given holder can enforce a given patent is a question of money, time, and nerve at least as much as of law. This article sets out what enforcement actually requires. As with the rest of the series, this is general information rather than legal advice, and it assumes filing in the United States.

A Brief History

Patent enforcement has swung between extremes within living memory. The creation of the United States Court of Appeals for the Federal Circuit in 1982 gave patent appeals a single specialized court and, for two decades, a reputation for strengthening patents. The pendulum then swung back. The Supreme Court’s decision in eBay v. MercExchange in 2006 ended the near-automatic injunction, the America Invents Act of 2011 created a cheaper administrative path to challenge validity, and a wave of assertion by entities that made no products provoked a decade of decisions and reforms aimed at curbing abusive suits. The lesson of the history is that the value of a patent in enforcement is not fixed by the grant. It moves with the courts, the statutes, and the costs, and it has moved a great deal.

The State Does Not Enforce

The most basic fact of enforcement is the one the opening article named. A patent is not self-enforcing. No agency patrols the market for infringers. The holder must notice the infringement, decide to act, and bring a patent infringement suit in federal court, because patents are creatures of federal law. The burden of proof, the cost, and the risk all sit with the holder.

This single fact reorganizes everything that follows. A right that the owner must enforce at the owner’s expense is worth what the owner can afford to spend on it, and against whom. A holder who cannot fund a suit holds a right that a competitor with deeper pockets can infringe in practical safety.

The Cost and the Clock

A patent infringement suit is a multi-year, multi-million-dollar undertaking. A contested case that runs toward trial commonly costs each side in the millions of dollars, and the most valuable cases cost more. The money buys a long sequence. A complaint opens the case. A claim construction hearing, often called a Markman hearing, fixes the meaning of the claims, and because the claims define the right, this step alone frequently decides the outcome. Discovery, the exchange of documents and testimony, consumes much of the cost. A trial follows for the cases that reach one, and an appeal to the Federal Circuit follows many verdicts. The whole arc runs years, often three to five, during which the technology and the market do not stand still.

For a small company the cost and the clock are usually decisive. A startup cannot divert millions of dollars and years of attention into a lawsuit without ceasing to be a startup, which is the concrete reason the founder article treated the enforcement probability as near zero.

Asserting Puts the Patent at Risk

Enforcement is not only expensive. It is dangerous to the patent itself. A patent that is never asserted is never seriously tested, but the moment it is asserted the defendant has every incentive to destroy it. The defendant will argue non-infringement, and, more dangerously, will argue that the patent is invalid over prior art that the examiner never saw.

The America Invents Act sharpened this risk by creating inter partes review, an administrative proceeding before the Patent Trial and Appeal Board in which a challenger attacks the patent’s validity on prior-art grounds, faster and more cheaply than a trial, and with a record of invalidating a large share of the claims it reviews. A holder who asserts a patent therefore wagers the asset. A win brings damages, and perhaps an injunction. A loss can leave the patent narrowed or dead, and a patent struck down is gone against everyone, not only the defendant who killed it. The decision to enforce is a decision to put the patent on trial.

The Wager of Assertion

The risk of asserting a patent can be priced, and the price explains why holders so often decline to sue on patents they believe are strong.

A patent’s value is rarely confined to a single infringer. A strong patent is a threat against every potential infringer in the market, and that threat is the source of the licenses it can command. Call the wider value $W$, the worth of the patent as a credible threat against everyone other than the defendant at hand.

Asserting the patent against one defendant puts $W$ at risk, because if the patent is invalidated it is dead against everyone, not only against the defendant who killed it. Let $p_v$ be the probability the patent survives the validity challenge, $p_i$ the probability infringement is found if it survives, $D$ the damages from this defendant, and $C_p$ the cost of suing. The expected value of asserting is

\[E[\text{assert}] = p_v \, p_i \, D - C_p - (1 - p_v)\, W.\]

The first two terms are the gain the moat article would recognize, the probability-weighted damages less the cost. The third term is new and often dominant. With probability $1 - p_v$ the patent is struck down, and the holder loses not this case alone but the wider value $W$ that the threat was worth.

A worked instance, in millions of dollars. Suppose the patent is worth fifty as a threat across the market, survives validity with probability seven tenths, and would, if it survives, win damages of ten from this defendant with probability eight tenths, at a litigation cost of two. The gain from the suit is

\[(0.7)(0.8)(10) - 2 = 3.6,\]

but the wager against the wider value is

\[(1 - 0.7)(50) = 15,\]

so the expected value of asserting is

\[3.6 - 15 = -11.4,\]

a loss of more than eleven million. The holder is far better off licensing under the threat than testing the patent in court, because a patent worth fifty million untested can be worth nothing the day after a verdict. This is the formal reason a credible threat is often more valuable than a suit, and why the strongest patents are frequently the ones least often litigated.

What a Win Is Worth

The remedies for infringement are narrower than they once were. An injunction, the order that actually stops the infringer, is no longer automatic. Since eBay v. MercExchange a court weighs a four-factor test, and a holder who does not practice the invention often cannot meet it, winning money but not exclusion. Damages are the more reliable remedy. Their floor is a reasonable royalty, the royalty a willing licensor and licensee would have negotiated, and a holder who practices the patent and can prove lost sales may instead recover lost profits. Willful infringement can raise the award, and a court may shift attorney fees in an exceptional case, but these are the exceptions. The ordinary outcome of a successful suit is a reasonable royalty, which is often far less than the headline value of the market at issue.

The Economics of the Settlement

Most patent suits never reach a verdict. They settle, and the logic of the settlement shows where the leverage in enforcement really lies.

Let $D$ be the damages at stake if the plaintiff wins, $q$ the probability the patent is held valid and infringed, which folds the merits into one number, $C_p$ the plaintiff’s cost of litigating to the end, and $C_d$ the defendant’s. A plaintiff who fights to the end expects the expected value $qD - C_p$, and so will not settle for less. A defendant who fights to the end expects to pay $qD + C_d$, and so will not pay more. Settlement is rational for any amount in the range

\[\left[\, qD - C_p,\; qD + C_d \,\right].\]

The width of that range is $C_p + C_d$, the combined cost of fighting, and a settlement exists because both sides prefer to split that saving rather than burn it on lawyers.

Two features of the range explain the reality of enforcement. The center of the range is $qD$, the merits times the stakes, so a holder with a genuinely strong patent, large damages, and the resources to threaten trial credibly can extract real value. But look at what happens when the patent is weak, so that $q$ approaches zero. The plaintiff’s walk-away, $qD - C_p$, goes negative, meaning a rational plaintiff would lose money at trial. The defendant’s walk-away, $qD + C_d$, stays positive at about $C_d$. A defendant will pay up to its own cost of defense to make even a meritless case disappear. This is the nuisance value of a patent suit,

\[S_{\text{nuisance}} \approx C_d,\]

a settlement extracted not from the strength of the patent but from the cost of defending against it.

A worked instance, in millions of dollars. With damages $D = 10$, a merits probability $q = 0.3$, and equal litigation costs $C_p = C_d = 2$, the settlement range is

\[\left[\, (0.3)(10) - 2,\; (0.3)(10) + 2 \,\right] = \left[\, 1,\; 5 \,\right],\]

centered on the expected merits outcome of three. Now let the patent be almost certainly invalid, $q = 0.02$. The plaintiff would lose money at trial, since $qD - C_p = 0.2 - 2 = -1.8$, yet the defendant still faces about two in defense costs, so a settlement near two can be pulled from a case that should never have been brought.

The lesson runs in two directions. A well-resourced holder can wield the cost of defense as a weapon against a smaller rival, and a non-practicing entity that makes nothing and so fears no countersuit can build a business on nuisance settlements alone. The cost of litigation, not the merit of the patent, is the lever, which is why enforcement favors whoever can better afford the fight.

Enforcement Without a Courtroom

Because litigation is so costly, most enforcement happens in its shadow rather than in it. A credible threat to sue is itself the instrument, and most disputes resolve into a license negotiated under that threat. Large companies with deep portfolios practice a mutual version of this, cross-licensing their patents to one another so that each is free to operate and none sues, a détente that smaller players, holding fewer patents, cannot enter on equal terms. The credibility of the threat is everything, and credibility is bought with the resources and the will to litigate. A holder who cannot fight cannot threaten, and a license is only as valuable as the suit behind it.

The Players

The cost structure sorts the participants. Operating companies enforce against competitors, usually reluctantly, because litigation is a distraction from building. Non-practicing entities, which hold patents to assert rather than to make products, turn the cost asymmetry into a model, since they have no operations to countersue and no product to enjoin, and they often fund their suits through litigation finance that supplies the capital in exchange for a share of the award. Defensive aggregators and patent pools form on the other side to blunt these suits. The common thread is that the ability to bear cost and risk, not the quality of the underlying invention, determines who prevails.

The Reality for the Small Holder

The threads gather into a hard conclusion for the founder and the small company. A patent’s protective value is gated by enforceability, and enforceability is mostly out of reach for a holder who cannot fund a multi-million-dollar suit against an opponent who can. This is why the founder article counseled valuing a patent for priority, signaling, and sale rather than for litigation.

The asymmetry cuts both ways, and the second edge is easy to forget. A small company is not only a holder who cannot enforce. It is also a target that cannot easily defend. The same nuisance economics that a startup cannot exploit as a plaintiff can be turned against it as a defendant, when a well-funded holder or a non-practicing entity asserts a patent and offers to settle for less than the cost of defense. For the small company, the cost of patent litigation is a danger from both sides of the caption.

Epistemic State

The settled matters here are that a patent is not self-enforcing, that infringement suits are expensive and slow, that asserting a patent exposes it to invalidation over prior art, that injunctions are no longer automatic after eBay, and that most suits settle. These should hold against independent verification.

The settlement model is a simplification. It folds the merits into a single probability $q$, assumes both sides estimate the odds and the costs alike, and ignores the asymmetries of information, risk tolerance, and strategic reputation that shape real negotiations. It is offered to make the structure of leverage clear, not to predict a particular settlement. The assertion wager is a simplification in the same spirit, compressing the patent’s market-wide threat value into a single number $W$ that is hard to estimate, and treating invalidation as all-or-nothing.

The cost and duration figures are typical orders of magnitude, not quotes, and they vary widely with the stakes, the venue, and the parties. The characterization of non-practicing-entity behavior is a general pattern with honest and abusive examples alike. Throughout, this is general information, United States centric, and not legal advice.

Out of Scope

The detailed procedure of an infringement trial and of an inter partes review, the rules of claim construction, and the methods of computing damages are matters for specialists and counsel. The ongoing policy debate over patent reform, venue, and the proper treatment of non-practicing entities is large and unsettled and is not adjudicated here. The enforcement of patents outside the United States, which varies by jurisdiction and includes systems with faster or cheaper procedures, is left to local counsel.

Conclusion

A patent is worth, in the end, what its holder can enforce, and enforcement is expensive, slow, and dangerous to the patent itself. The state grants the right but leaves the fight to the owner, the fight costs millions and takes years, asserting the patent risks losing it, and the leverage in the fight comes from the ability to bear its cost rather than from the merit of the invention. For most holders, and nearly all small ones, this places real enforcement out of reach, which is the deepest reason that most patents are not moats and that a startup should value a patent for what it signals rather than for what it could win. This closes the patent half of the series. The startup half takes up the question the founder article raised and this one confirmed, namely what it takes to succeed and where moats come from when the patent is not the answer.

References