Patent Assertion Defense Toolkit: Demand Letters, NPEs, and Cost-Effective Response

By Casey Scott McKay · 2024-11-04

Most patent assertions are decided by cost rather than by merit, and the defense that works is the one built around that fact. This toolkit runs a defense from the demand letter through resolution, and routes each stage to the Marksy documents that do the work. It covers letter triage and the internal record that answers willfulness, the complete defenses that end cases in days, indemnity tender and insurance notice, venue and standing challenges, the eligibility and divided-infringement motions worth filing, the prior art search that drives both the district court case and the post-grant decision, joint defense economics, and the exposure model that should govern every subsequent expenditure. It closes with the fee-shifting record, the design-around, the settlement terms that prevent the next assertion, and the reading path, the authorities table, and the forms.

IP and Technology > Patent Litigation | Toolkit | Published 4 November 2024 - Updated 23 May 2026 | Casey Scott McKay - marksy.us

Summary. Most patent assertions are decided by cost rather than by merit, and the defense that works is the one built around that fact. This toolkit runs a defense from the demand letter through resolution, and routes each stage to the Marksy documents that do the work. It covers letter triage and the internal record that answers willfulness, the complete defenses that end cases in days, indemnity tender and insurance notice, venue and standing challenges, the eligibility and divided-infringement motions worth filing, the prior art search that drives both the district court case and the post-grant decision, joint defense economics, and the exposure model that should govern every subsequent expenditure. It closes with the fee-shifting record, the design-around, the settlement terms that prevent the next assertion, and the reading path, the authorities table, and the forms.

Keywords: demand letter triage · non-practicing entity · litigation funding · exposure model · license inventory · exhaustion · standing defect · venue challenge · eligibility motion · divided infringement · prior art search · post-grant petition · stay pending review · joint defense group · customer suit exception · indemnity tender · insurance notice · fee shifting record · design around · settlement terms

Start Here

Ashcombe Industrial receives a letter on a Monday. It names a patent, does not name a product, and offers a licence "to resolve this matter efficiently" at a figure that is meaningful but not enormous.

By Friday four things are true and only one of them is about the patent.

The letter was forwarded to engineering, where a manager replied by email that "we probably do something like that." That email is discoverable.

The company has an indemnity from the supplier of the module that likely implements the accused functionality, with a thirty-day notice requirement. Twenty-six days remain.

The plaintiff is an entity with no products, four employees, and a portfolio acquired from an operating company two years ago.

And nobody has computed what the case is worth, so every decision from here will be made on instinct.

This toolkit answers three questions.

What is this worth, and what does defending cost? The two numbers that govern every subsequent decision, and both are computable in month one.
What ends this cheaply? A licence already held, exhaustion, a standing defect, an expired patent, or a marking problem. All answerable from documents the company has.
What if it does not end cheaply? Venue, early motions, a post-grant petition, and a settlement structured so the next assertion does not follow.

If you read only one thing, read Where Patent Cases Are Fought. It explains the asymmetry the assertion model runs on, and understanding it is what makes the rest of the defense rational.

Triaging the Letter

Read the patent, not the letter. The claims decide it; the letter describes it.

Calendar it. A demand letter may start the damages clock under 35 U.S.C. § 287(a) and it establishes knowledge for indirect liability and willfulness under 35 U.S.C. § 271(b) and 35 U.S.C. § 284. Whether it does depends on whether it identifies the patent and the accused product with specificity — and a letter naming neither starts nothing.

Create the internal record. 35 U.S.C. § 298 means the absence of an opinion of counsel proves nothing on willfulness. The absence of any assessment at all is a different fact and a jury hears about it. A short memorandum recording that the letter was read, the patent reviewed, and a conclusion reached is inexpensive insurance.

Do not forward it to engineering with a question. Ashcombe's manager's email is the most damaging document in the file, written by someone who has never read a claim, and it was created in the first week.

If the company is a customer rather than a manufacturer, tender to the supplier immediately in writing.

Decide whether a declaratory judgment action is available and desirable. 28 U.S.C. § 2201 requires an actual controversy, which a specific letter generally supplies. Filing buys the forum and the initiative; it costs the option of waiting for a case that might never come.

Respond in a way that preserves options. A general denial, a request for claim charts, and a request for the chain of title costs little and produces information. Arguing the merits in detail educates the plaintiff.

The Complete Defenses

Run all of these before spending anything on the merits. Each is answerable in days from documents the company already has.

A licence the client already holds. Through a corporate parent, a subsidiary, an acquisition, a supplier's pass-through rights, a portfolio agreement, or a standards commitment. The cheapest complete defense available and the one most often skipped, because nobody in the legal department knows what the company acquired five years ago.

Exhaustion. An authorized sale exhausts rights in that article, which is a complete answer for a reseller, a refurbisher, or a company whose accused functionality lives in a licensed component.

Standing. Chain of title from each named inventor, whether the plaintiff held the rights at filing, whether accrued causes of action were assigned, and whether a co-owner is missing. 35 U.S.C. § 261; 35 U.S.C. § 281; 35 U.S.C. § 262. Entirely public, and a defect is a complete answer. Portfolios acquired from operating companies — like Ashcombe's plaintiff's — are where the gaps are.

Expiration and lapse. Term under 35 U.S.C. § 154(a), maintenance fees under 35 U.S.C. § 41, and terminal disclaimers under 37 C.F.R. § 1.321.

Marking. 35 U.S.C. § 287(a). Not a complete defense, and it can remove most of the exposure in a week's work by checking the patentee's own products and every licensee's.

The six-year limitation. 35 U.S.C. § 286.

Implied licence and standards commitments, where applicable.

Tender and Notice

Tender to suppliers in writing, within the first week. The obligation is contractual, most agreements require prompt written tender, and delay forfeits it. Ashcombe has twenty-six days and it is the single most valuable action available.

Notify every insurer. Advertising injury under a general liability policy can reach some patent claims depending on the pleading, technology errors and omissions policies sometimes cover intellectual property expressly, and specialty patent coverage exists. Notice provisions are strict and late notice is a coverage defense independent of the merits.

Consider tendering upward where the client is a customer, because a manufacturer stepping forward can support a stay of the customer case under the customer suit exception.

Check the flow-down. A company giving an infringement warranty to its customers should hold matching indemnities from its suppliers, and the gap between what it gives and what it receives is the risk it absorbs.

The Exposure Model

Build it in month one, on one page, and update it at every ruling.

The ceiling. Accused units in the plausible damages period, times an aggressive royalty on a defensible base. Not the plaintiff's number, which is typically built on the whole product and a period nobody has tested.

Defense cost to each milestone. To a venue ruling. To claim construction. Through summary judgment. Through trial. These are estimable and the differences are large.

Probability weights on the venue motion, the construction, and institution at the Board.

The output. A settlement range and a recommendation the client can act on.

The uncomfortable conclusion. Many cases are worth settling regardless of the patent's quality, because proving it bad costs more than the plaintiff will take. That asymmetry is the assertion model, and naming it plainly is more useful than litigating around it.

The exception. Where the client faces a campaign rather than a case, or where the accused feature is core to the product line, cancellation at the Board is worth several times its cost — because a cheap first settlement prices every subsequent assertion and a company known to pay receives more letters.

Venue, Standing, and the Early Motions

Test venue immediately. 28 U.S.C. § 1400(b) permits suit only in the state of incorporation or in a district where the defendant committed infringing acts and has a regular and established place of business — a physical place, regular and established, that is the defendant's own. A remote employee's home office generally is not.

Move to dismiss or transfer. 28 U.S.C. § 1406; 28 U.S.C. § 1404(a); Fed. R. Civ. P. 12. Support transfer with declarations about where the engineers and the records are.

What transfer buys. A different schedule, different local patent rules, and often a different summary judgment culture. A district with a trial date three years out prices very differently from one at eighteen months.

Eligibility. 35 U.S.C. § 101. Worth filing where the claims recite a result implemented on generic components and the specification describes a business advantage; worth skipping where they recite a specific technical mechanism.

Divided infringement. Where an asserted method claim recites steps performed by different actors, this is available on the pleadings and it is frequently the strongest argument in a software or platform case.

Failure to state a claim where the complaint does not identify the accused product or plead the elements of an indirect theory.

Indefiniteness, usually deferred to construction, and available early where a term has no discernible boundary. 35 U.S.C. § 112(b).

File one motion, alone. Three filed together dilute each other, and a court reads the weakest first.

The Prior Art Search and the Post-Grant Decision

Read the file wrapper first, and identify the amendment that produced allowance. That limitation is the point of novelty and it defines the search.

Commission the search against it, not against the field, and prioritize what examiners search poorly — non-patent literature, conference proceedings, standards contributions, foreign patent documents, product manuals, and theses.

Search the client's own history. Products sold before the priority date and internal development records are prior art no search firm can find.

Calendar the one-year bar from service and set the real decision date at month six. 35 U.S.C. § 315(b).

Check whether a privy or real party in interest was served earlier, because that imports the bar.

Ask whether the best invalidity argument is documentary. Only patents and printed publications under sections 102 and 103 are available at the Board. Eligibility, section 112, on-sale, and public use stay in district court. 35 U.S.C. § 311.

Inventory the estoppel cost. 35 U.S.C. § 315(e). A defendant whose invalidity case is entirely documentary bets that case; one with a strong on-sale bar loses little.

Weigh the district court schedule, because a fast trial date makes discretionary denial likely.

Move to stay on filing and again after institution. The second motion is the one that works.

Joint Defense

Form the group where multiple defendants face the same patent. Prior art searching, claim construction briefing, and expert work all divide, and the savings are real.

Paper it properly. Privilege, what happens when a member settles, how positions are coordinated, and how costs are allocated.

Understand the divergence risk. A member whose product infringes on a different theory has an interest in a construction the group may not want.

Understand the estoppel asymmetry. Only the petitioner and its real parties in interest are estopped by a final written decision, which means a non-petitioning member keeps every ground — and which makes funding negotiations within a group difficult.

Understand the settlement dynamic. Early settlers pay less, deliberately, and the plaintiff builds a licence stack it will offer as comparable evidence against the rest. A defendant intending to fight should know that the early settlements are being priced as evidence.

Joinder limits. 35 U.S.C. § 299 prohibits joining unrelated defendants merely because each allegedly infringes the same patent, though consolidation for pretrial purposes achieves much of the same efficiency.

The Fee-Shifting Record

Build it contemporaneously. An exceptional case finding under 35 U.S.C. § 285 is assembled during the litigation or not at all.

Document the weaknesses as they emerge. The construction that destroyed the theory. The art the plaintiff ignored. The contentions that were never supported. Settlement demands pitched below defense cost.

Note the pre-filing inquiry. Fed. R. Civ. P. 11 requires reading the claims and comparing them to the accused product before filing. A complaint that could not have survived that is evidence.

Preserve the correspondence, including any demand campaign aimed at customers.

Check the state bad-faith assertion statutes, several of which regulate demand letters that fail to identify the patent or the accused product or that demand a response in an unreasonably short time.

Understand what it is worth. Fee shifting rarely finances a defense and it is a real check on weak assertions, and the record costs almost nothing to keep.

The Design-Around and Settlement

Price the design-around early. What it would cost to remove the accused functionality and how long it would take. That number caps the royalty in the hypothetical negotiation, caps the settlement value going forward, and answers an injunction threat.

Document it carefully, because the analysis will be produced and a memorandum saying the change is trivial helps while one saying it is impossible does not. Avoid describing it in engineering documents as avoiding a named patent without counsel's involvement.

Implement it where the number is small. A modest engineering change that ends future exposure converts an indefinite liability into a fixed cost.

Time settlement offers to the repricing events. Claim construction, institution at the Board, a marking ruling, and expert exclusion each move the number, and a party that knows which is coming can time its offers.

Get a release and a licence, not a dismissal, covering the accused products, their successors, and customers.

Cover the family. A settlement resolving one patent from a live family invites the next assertion when the continuation issues.

Consider the comparable-evidence problem. A plaintiff building a licence stack will use the client's rate against others, and the client's rate is being priced as evidence.

Understanding the Model

The defense is rational only if the assertion model is understood, and clients consistently misread it as a dispute about technology.

The structure. A non-practicing entity acquires patents, frequently through a shell subsidiary formed for a single campaign, and asserts them against operating companies. It sells nothing, so it faces no infringement counterclaim, has few documents to produce, and cannot be enjoined. Its litigation cost is a fraction of the defendant's.

The asymmetry is the business. A defense through claim construction costs a meaningful sum; through trial, a great deal more. A demand pitched below the cost of the motion that would resolve the case is rational to accept even when the patent is weak. Multiplied across dozens of defendants, modest settlements fund the campaign.

Litigation funding changed the volume. Third-party capital finances campaigns in exchange for a share of recoveries, which decouples the decision to sue from the plaintiff's own balance sheet. Several districts now require disclosure of funding arrangements by standing order, and where disclosure is available it is worth requesting.

The demand campaign. Many campaigns never reach a complaint. Letters go to end users rather than manufacturers, because end users are numerous, unsophisticated about patents, and unaware of the indemnity rights they hold.

Portfolio provenance matters. A portfolio acquired from an operating company — like Ashcombe's plaintiff's — frequently has chain-of-title gaps, missing inventor assignments, and an assignment that transferred the patents without transferring accrued causes of action. 35 U.S.C. § 261. This is where the cheapest defenses live.

What this means for the defense. The correct question is not whether the patent is good. It is what this costs to resolve at each stage, and whether the client's position improves enough at any stage to justify reaching it.

The Customer Campaign

Assertions aimed at a manufacturer's customers are a distinct problem with distinct answers.

Why it works. Customers are numerous, they have no patent expertise, they rarely know their purchase agreements contain indemnities, and each individual exposure is small enough to settle without a fight.

What a customer should do. Tender to the supplier immediately, in writing, under whatever indemnity the purchase agreement contains. That is the entire response required, and it is the step most customers do not know exists.

What a supplier should do. Consider stepping forward. The customer suit exception permits a court to stay customer cases in favor of the manufacturer's, on the reasoning that the manufacturer is the true defendant and its case will resolve the issues. It is discretionary and it depends on the manufacturer having appeared.

Why a supplier might not. Stepping forward converts many small exposures into one large one, and it puts the patent's validity at issue in a case the supplier controls but did not choose. That is frequently the right trade and it is a decision, not a default.

Declaratory judgment. A supplier whose customers are being targeted may have standing to seek a declaration, which consolidates the dispute in a forum it selects. 28 U.S.C. § 2201.

Joinder limits help. 35 U.S.C. § 299 prevents joining unrelated customers in one action, which raises the plaintiff's cost of running a broad campaign — though consolidation for pretrial purposes recovers much of it.

State bad-faith assertion statutes target exactly this practice, regulating letters that fail to identify the patent or the accused product or that demand a response in an unreasonably short period.

Ashcombe's First Week, Rewritten

The same week, done properly.

Monday. The letter arrives. It goes to counsel, not to engineering. Counsel reads the patent's claims, not the letter's characterization, and identifies which claims could plausibly read on which product.

Monday afternoon. The calendar entries: the one-year bar from any future service, a decision date at month six for a possible petition, and the twenty-six-day indemnity notice deadline.

Tuesday. Written tender to the module supplier under the indemnity clause, with a copy of the letter and a request for whatever clearance the supplier performed. Insurers notified.

Tuesday. A short internal memorandum recording that the letter was received, the patent reviewed, and an assessment underway. This is the record that answers willfulness later, and it costs an hour. 35 U.S.C. § 298.

Wednesday. The assignment records pulled. The plaintiff acquired the portfolio from an operating company two years ago, which means the chain runs through named inventors, through the operating company, through the acquisition, and into a shell. Each link is checkable and any gap is a complete answer. 35 U.S.C. § 261.

Wednesday. The contract files searched for a licence — through the parent, through acquisitions, through supplier pass-through rights, and through any portfolio agreement. An afternoon, and it ends cases.

Thursday. The patent's front page checked: term, maintenance fees, terminal disclaimers. The marking analysis begun on the plaintiff's own products, if it has any, and on every licensee it has publicly identified.

Thursday. The claims read for divided performance and for eligibility, both of which are visible from the claim language and both of which are available on the pleadings.

Friday. The exposure ceiling computed. Units, period, an aggressive rate on the module rather than the machine. Against the defense cost to a venue ruling, to construction, and through trial.

Friday. A one-page memorandum to the business: what this is worth, what it costs, and which of four paths the facts support.

What did not happen. Nobody asked engineering whether the product does what the claims say. That email, in the real version of Ashcombe's week, is now the most damaging document in the file.

The Four Paths

Every assertion resolves into one of four, and naming which one applies is the whole of the advice.

Settle now. Where the exposure ceiling is below the cost of reaching claim construction, and where the client faces no campaign risk. The recommendation should include the number and the terms — release, licence, family coverage — rather than a direction to negotiate.

Move and reassess. Where a venue or transfer motion, a standing defect, or an eligibility motion has a real chance of resetting or ending the case. Spend on the one motion, hold everything else, and reassess on the ruling.

Petition and stay. Where the art is materially better than the examiner's, the documentary case is not the whole case, and the district court schedule permits institution. The highest-value path when it fits, and it fits less often than defendants hope.

Litigate to construction. Where a single claim term decides infringement and the intrinsic record supports the client's reading. Budget to the construction ruling and treat it as the decision point rather than as a milestone on the way to trial.

Write it down with the number attached. An assessment that ends in a description of the issues has not done its job. The client is paying for a recommendation, and the four options above are the entire menu.

Controlling the Cost

Defense cost is the variable the client controls, and most of it is spent on work that does not change the outcome.

Resist the document review nobody needs. In a case turning on one claim element, the responsive universe is far narrower than the plaintiff's requests suggest. Negotiate custodians and search terms aggressively and early, and propose a phased production tied to the contested element. Fed. R. Civ. P. 26.

Negotiate the source code provisions before any review. Standalone machine, no network, printing limits, inspection location, and access list. A review conducted under terms nobody read produces unusable notes and a dispute. Fed. R. Civ. P. 34.

Push back on financial discovery until the period is settled. A marking finding that shortens the period by four years shortens the production proportionally.

Sequence the experts after the technical record. An expert engaged early produces an opinion revised later and a client billed twice.

Do not contest every limitation. A response contesting all seven reads as a party with no argument, and it costs seven times what contesting one costs.

Do not propose eleven constructions. The court will cap the number and choose for you.

Use the joint defense group for the prior art search, the construction briefing, and the technical expert. These divide cleanly.

Keep the privilege log defensible, because fee exposure runs both ways under 35 U.S.C. § 285 and most defense-side fee awards rest on discovery conduct.

Report cost against the exposure model monthly. A defense that has spent more than the ceiling is a defense nobody was managing, and it is a common outcome where nobody built the model in month one.

What to Tell the Business

A weak patent is not a cheap case. The asymmetry means a plaintiff with no products has no counterclaim exposure, minimal discovery burden, and no injunction risk, while the defendant carries all three. Proving a bad patent bad costs more than the plaintiff is asking. That is the model, and it should be said once, plainly, with numbers.

Paying may bring more letters. True, and it is the strongest argument for fighting a case whose economics say settle. Where the client faces a campaign, or where the accused feature is core to the product line, cancellation at the Board is worth several times its cost.

Fees are not a financing plan. 35 U.S.C. § 285 is a real check on weak assertions, it takes years, and it requires winning first.

The schedule drives everything. Eighteen months to trial in a fast district, four years in a slow one, and a stay pending a post-grant proceeding adds a year while usually saving money.

We need three things from the business. The contract files, because the cheapest defense is a licence the company forgot it had. The pricing and configuration history, because it caps the damages. And one engineer who will answer questions without guessing.

And one thing we need the business not to do. Discuss the patent in email without counsel. Every message is discoverable, most of them speculate, and the manager's reply in week one is routinely the most damaging document in the case.

Building the Standing Response

A company that receives assertions regularly should stop treating each as a novel event.

A named owner. One lawyer who receives every letter, runs the triage, and maintains the register.

A routing rule. Any communication mentioning a patent goes to that person before it goes anywhere else, and never to engineering with a question attached.

A triage template. Patent, claims, accused product, calendar entries, tender and notice, licence check, chain of title, term and fees, marking, claim screens, and the exposure ceiling. The same ten items every time, on one page.

A maintained licence inventory. Every licence the company holds, through every entity, with the fields and territories. This is the cheapest defense available and it is unusable unless it exists before the letter arrives.

A supplier indemnity register, with notice deadlines, so a tender never lapses for want of a calendar entry.

Standing relationships with a searcher who knows the technology and with counsel who has run petitions.

A decision log. Every assertion, the analysis, the decision, and the outcome. Assertion entities frequently return, and a company that knows what it paid last time negotiates better.

Settlement templates with the terms that prevent the next assertion — family coverage, successor coverage, customer coverage, and a covenant rather than a bare dismissal.

And a briefing for the business, annually, on the two rules that matter and that nobody remembers otherwise: any letter mentioning a patent goes to legal the day it arrives, and nobody discusses it in email. Those two habits prevent more damage than every motion described in this toolkit.

A Suggested Reading Path

In the first week:

If you are considering the Board:

If the case is going the distance:

Primary Authorities

| Authority | Proposition | |---|---| | 28 U.S.C. § 1400(b) | Patent venue | | 28 U.S.C. § 1404(a) | Transfer for convenience | | 28 U.S.C. § 1406 | Improper venue | | 28 U.S.C. § 2201 | Declaratory judgment | | 35 U.S.C. § 101 | Eligibility; early motion | | 35 U.S.C. § 112 | Definiteness | | 35 U.S.C. § 154(a) | Term | | 35 U.S.C. § 261 | Chain of title | | 35 U.S.C. § 262 | Joint owners; missing co-owner | | 35 U.S.C. § 271 | Infringing acts; divided performance | | 35 U.S.C. § 281 | Standing | | 35 U.S.C. § 284 | Damages; enhancement | | 35 U.S.C. § 285 | Fees in exceptional cases | | 35 U.S.C. § 286 | Six-year limitation | | 35 U.S.C. § 287(a) | Marking and notice | | 35 U.S.C. § 298 | Advice of counsel | | 35 U.S.C. § 299 | Joinder of defendants | | 35 U.S.C. § 311 | Inter partes review; scope | | 35 U.S.C. § 315(b) | One-year bar | | 35 U.S.C. § 315(e) | Estoppel | | 35 U.S.C. § 41 | Maintenance fees | | 37 C.F.R. § 1.321 | Terminal disclaimers | | Fed. R. Civ. P. 11 | Pre-filing inquiry | | Fed. R. Civ. P. 12 | Motions to dismiss | | Fed. R. Civ. P. 56 | Summary judgment |

Forms and Templates

The Cease and Desist Template is worth reading from the receiving side, because its specificity determines whether the damages clock started and whether declaratory judgment jurisdiction exists — and a letter that names neither a patent nor a product, like Ashcombe's, accomplishes neither. The License Agreement Template is where most assertions end, and the terms that matter are release scope, family coverage, successor and customer coverage, and the recognition that the rate agreed becomes comparable evidence against others. The Portfolio Inventory Template should already hold the company's own licences and supplier indemnities, because the cheapest defense in this area is a licence the client did not know it had. The Assignment Agreement Template is the instrument a standing challenge examines, and an acquired portfolio with a gap in its chain is a complete answer.

Related Toolkits and Checklists

For the case if it proceeds, the Patent Litigation Toolkit supplies the sequence and the decision points. For the validity challenge, the PTAB Practice Toolkit. For the exposure number that governs every decision, the Patent Damages and Remedies Toolkit. For the clearance work that should have preceded the assertion, the Freedom-to-Operate and Patent Clearance Toolkit. And where the patents are declared essential to a standard, the Standard-Essential Patents and FRAND Toolkit changes the analysis entirely.