Patent Damages Checklist: Marking, Notice, Royalty Base, and Willfulness Evidence

By Casey Scott McKay · 2024-11-05

Patent damages are decided by a handful of documents that already exist and by three analyses that can be finished in the first month. This checklist runs the work in eleven phases: fix the damages period through marking, notice, and the limitation; open financial discovery with requests drafted against the model; choose between lost profits and a reasonable royalty on the substitutes evidence; establish the royalty base; gather apportionment proof; collect and test licenses; run the hypothetical negotiation; engage and constrain the expert; brief exclusion; assemble willfulness evidence; and plan the post-verdict royalty. Each box gives the reason, the authority, and the trap. It covers both chairs. A worked matter runs throughout.

IP and Technology > Patent Litigation | Checklist | Published 5 November 2024 - Updated 21 January 2025 | Casey Scott McKay - marksy.us

Summary. Patent damages are decided by a handful of documents that already exist and by three analyses that can be finished in the first month. This checklist runs the work in eleven phases: fix the damages period through marking, notice, and the limitation; open financial discovery with requests drafted against the model; choose between lost profits and a reasonable royalty on the substitutes evidence; establish the royalty base; gather apportionment proof; collect and test licenses; run the hypothetical negotiation; engage and constrain the expert; brief exclusion; assemble willfulness evidence; and plan the post-verdict royalty. Each box gives the reason, the authority, and the trap. It covers both chairs. A worked matter runs throughout.

Keywords: marking analysis, virtual marking, licensee marking, actual notice, six-year limitation, damages period, financial discovery, incremental profit, lost profits elements, non-infringing substitutes, market share allocation, price erosion, royalty base, smallest salable unit, entire market value rule, apportionment evidence, comparable licenses, damages expert report, willfulness, ongoing royalty

How to use this checklist

| Phase | What it covers | When | |---|---|---| | 1 | Fix the damages period | Month one | | 2 | Financial discovery | Months one to two | | 3 | Lost profits or royalty | Months three to six | | 4 | The royalty base | Months four to eight | | 5 | Apportionment evidence | Months four to nine | | 6 | Licenses | Months two to nine | | 7 | The hypothetical negotiation | Months nine to eleven | | 8 | The expert | After the technical record | | 9 | Exclusion | After reports | | 10 | Willfulness | Throughout | | 11 | Post-verdict | Before trial |

Boxes marked [Gate] should clear before the damages expert is engaged.

The matter. Three patents asserted against an industrial controller. The case was decided by an archived web page and an internal pricing comparison, both available in the first ninety days.

Phase 1. Fix the damages period

[ ] [Gate] Run the marking analysis in month one.
- Authority. 35 U.S.C. § 287(a).
- Trap. Running it at expert reports, after a year of discovery premised on a period that is actually half as long.
[ ] Identify every article that practices any asserted claim.
- Why. The obligation attaches per patent, not per company. A patentee asserting five patents may owe marking on two.
[ ] Photograph or document the product as sold during the period, not as sold today.
[ ] Check virtual marking and its history.
- Why. Two questions: was the web address on the product during the period, and did the page list the asserted patent then? Archived snapshots answer the second, and they frequently show a patent added years late.
[ ] [Gate] Check every licensee's products, not only the patentee's.
- Why. An unmarked licensee product covered by an asserted claim limits the patentee's recovery.
- Trap. The single most common way a damages period collapses.
[ ] Read the marking clauses in the licenses.
- Why. A license omitting a marking obligation is evidence about reasonable efforts.
[ ] Note the method-claim exception.
- Why. Where only method claims are asserted and no article is sold, marking generally does not apply. Claim selection is therefore a damages decision.
[ ] Establish the actual notice date with specificity.
- Why. A communication identifying the patent and the accused activity. A portfolio letter naming neither does not start the clock. Filing suit always does.
[ ] Apply the six-year limitation from the complaint date.
- Authority. 35 U.S.C. § 286.
[ ] Check provisional rights.
- Authority. 35 U.S.C. § 154(d).
- Trap. Assuming they apply; the issued claims must be substantially identical to the published ones, and most applications are amended.
[ ] Check the term, maintenance fees, and any terminal disclaimer.
- Authority. 35 U.S.C. § 154(a); 35 U.S.C. § 41(b).
[ ] Check intervening rights if claims were amended in reissue or reexamination.
- Authority. 35 U.S.C. § 252.
[ ] Deliverable: a one-page memorandum stating the start date, the basis, and the units and revenue inside the period.

Phase 2. Financial discovery

[ ] Request unit and revenue data at full granularity, in native format.
- Trap. Summary spreadsheets, which cannot support apportionment.
[ ] Request cost data sufficient to compute incremental margin.
- Why. Bill of materials, direct labor, variable overhead methodology, and the policy distinguishing fixed from variable.
[ ] Request historical price lists and pricing analyses.
[ ] [Gate] Request by name any document comparing configurations with and without the accused feature.
- Why. In component cases this is the strongest apportionment evidence available and it lives in the defendant's own files.
[ ] Request product management documents.
- Why. Roadmaps, requirements documents, and post-launch reviews describe a feature's commercial importance candidly.
[ ] Request sales and marketing materials by period, including win-loss analyses.
[ ] Request every license in the technical field, both directions, executed and draft, with negotiation correspondence.
[ ] Request any design-around analysis.
- Why. It sets the ceiling in the hypothetical negotiation.
[ ] Request corporate valuation materials from the patentee.
- Why. Board decks and diligence memoranda frequently contradict the litigation position.
[ ] Serve these in month two.
- Trap. Serving at month ten, when the production timeline no longer fits the expert schedule.

Phase 3. Lost profits or reasonable royalty

[ ] [Gate] Answer the substitutes question before building a lost profits case.
- Why. If an acceptable non-infringing alternative was available during the period and acceptable to the customers who actually bought, the sale was not lost.
[ ] Test availability, acceptability, and whether the substitute had the advantages of the invention.
[ ] Establish demand with sales volumes and market data.
[ ] Establish capacity.
- Trap. A patentee at full utilization claiming it would have made an additional hundred thousand units.
[ ] Compute incremental profit, not average margin.
- Trap. Deducting fixed costs that would have been incurred anyway, which understates the claim, or ignoring variable costs, which inflates it and invites exclusion.
[ ] Plan market share allocation as the realistic outcome.
- Why. Lost profits on the share the patentee would have captured, plus a royalty on the balance.
[ ] If claiming price erosion, model elasticity.
- Trap. Assuming the same volume at a higher price.
[ ] If claiming convoyed sales, establish functional relationship.
- Why. Items sold together for business convenience do not qualify.

Phase 4. The royalty base

[ ] [Gate] Start from the smallest salable patent-practicing unit.
- Why. Where the invention resides in a component, the base begins there rather than at the end product.
[ ] Apportion further where that unit contains unclaimed functionality.
[ ] Do not use the whole product unless the patented feature drives demand for it.
- Why. The entire market value rule requires that the feature be the basis for the purchase decision, not merely important or necessary.
[ ] If the base is the whole product, show how the rate carries the apportionment.
- Trap. A rate asserted to be "already apportioned" with no derivation.
[ ] Defense side: spend the budget on the base, not the rate.
- Why. Moving from a five-hundred-dollar product to a ten-dollar component is worth more than any rate argument.
[ ] For design patents, note the different rule.
- Authority. 35 U.S.C. § 289.
- Why. Total profit on the article of manufacture, with the fight over what the article is.

Phase 5. Apportionment evidence

[ ] Find the price differential between versions with and without the feature.
[ ] Gather engineering allocation evidence.
- Why. Design documents, development cost records, and specifications.
[ ] Gather customer decision evidence.
- Why. Requests for proposal, feature comparison charts, and win-loss files.
[ ] Consider survey work, and design it defensibly.
- Trap. Conjoint studies whose attribute sets and question wording drive the answer.
[ ] Use prior versions and successors.
- Why. What the product did before the feature, and what the design-around cost.
[ ] Use usage telemetry where it exists.
- Why. Low actual use of an accused feature is powerful apportionment evidence.

Phase 6. Licenses

[ ] Collect everything, on both sides, before the other party produces it.
- Trap. A favorable license the client forgot about, produced by the adversary.
[ ] Test technological comparability.
[ ] Test economic comparability.
- Why. A license negotiated under litigation pressure reflects litigation economics.
[ ] Unpack portfolio licenses rather than dividing by patent count.
- Trap. Arithmetic offered as analysis.
[ ] Convert rate structures explicitly, with stated volume assumptions.

Phase 7. The hypothetical negotiation

[ ] Fix the date at the first infringing sale or feature introduction.
- Trap. A single blended rate across a long period with multiple product generations.
[ ] Assemble what each party knew at that date.
[ ] Establish the floor from the patentee's licensing history.
[ ] Establish the ceiling from the design-around cost.
- Why. A negotiation whose ceiling is a fifty-thousand-dollar engineering change does not produce a nine-figure royalty.
[ ] Use later evidence to inform the result.
- Why. Actual sales, later licenses, and commercial success are admissible.
[ ] Avoid rules of thumb.
- Trap. The twenty-five percent rule, an even profit split, or an industry-average rate untied to this patent.

Phase 8. The expert

[ ] [Gate] Engage after the technical record, not before.
[ ] Give the expert documents, not a target number.
[ ] Require a forward derivation: base, apportionment, rate, period.
[ ] Confirm the technical premise matches the infringement expert's.
- Trap. Two reports describing the accused feature differently.
[ ] Comply fully with the disclosure requirements.
- Authority. Fed. R. Civ. P. 26.

Phase 9. Exclusion

[ ] Brief the motion on the report and the deposition together.
- Authority. Fed. R. Evid. 702.
[ ] Target the recurring grounds.
- Why. Unsupported entire market value, asserted apportionment, non-comparable licenses, rules of thumb, unsupported technical premise, and a survey built toward its answer.
[ ] Prepare the cross-examination regardless of the ruling.

Phase 10. Willfulness

[ ] Track knowledge from the notice date.
- Authority. 35 U.S.C. § 284.
[ ] Understand that enhancement is the court's decision and is not automatic.
[ ] Note that failure to obtain advice of counsel proves nothing.
- Authority. 35 U.S.C. § 298.
[ ] Assess fee exposure both ways.
- Authority. 35 U.S.C. § 285.
[ ] Defense side: be able to explain what was done on receiving the letter.
- Trap. A detailed notice letter, no response, and no internal record of any assessment.

Phase 11. Post-verdict

[ ] Assess injunction exposure on the equitable factors.
- Authority. 35 U.S.C. § 283.
[ ] Model the ongoing royalty before trial.
- Why. Where an injunction is unlikely, the post-verdict rate governs the client's future and is typically higher than the jury's rate.
[ ] Compute supplemental damages for the gap between the trial cutoff and judgment.
[ ] Calendar prejudgment interest.

Phase 12. The damages model, assembled

Once the phases above are complete, the model assembles from parts that each have evidence behind them. Run this list against the finished model.

[ ] The unit count is tied to production data, not to an estimate.
- Trap. A projection used because the production was incomplete, which the other side will attack as speculation rather than as arithmetic.
[ ] The period start date matches the marking and notice memorandum.
- Why. A model that silently assumes a six-year period after the marking analysis showed two is a model that will be rebuilt under time pressure.
[ ] Every unit in the count was sold, offered, made, used, or imported in the United States.
- Authority. 35 U.S.C. § 271(a).
- Trap. Worldwide sales figures used as the base without a territorial screen.
[ ] Units attributable to indirect theories start at the knowledge date.
- Authority. 35 U.S.C. § 271(b).
- Why. The direct and indirect portions of the period frequently differ, and a single blended period conceals that.
[ ] The base is the same in every ground of the model.
- Trap. A lost profits calculation using product revenue and a royalty calculation using component revenue, presented side by side without explanation.
[ ] The apportionment step has a document behind it.
[ ] The rate has a comparable license or a derivation behind it.
[ ] The model computes cleanly when any single input is changed.
- Why. The court may adopt a different period or a different base, and a model that cannot be recomputed on the stand is a model the jury will not follow.
[ ] Every input traces to a produced document with a Bates number.

Phase 13. What each side is looking for in the other's production

Damages discovery is adversarial in a specific way: each side knows what would hurt it, and both sides under-produce the same categories.

What the defendant is hunting for. An unmarked licensee product. A vague notice letter. A license to the asserted patent at a low lump sum. An internal valuation of the patent that is modest. Evidence that the patentee could not have made the additional units. Competitor products that customers accepted as substitutes. A prior art product the patentee sold itself.

What the patentee is hunting for. A document showing the accused feature was a selling point. A price premium for the configuration containing it. An internal analysis of what a design-around would cost, particularly one showing the cost was high. Evidence the defendant knew about the patent and continued. Marketing that emphasizes the accused functionality. Customer feedback praising it.

What both sides under-produce. Licenses. Draft agreements, term sheets, expired licenses, and cross-licenses embedded in supply agreements are all responsive and all routinely missed by custodial collections that search for the word "license" in email.

The practical instruction. Draft requests that name documents rather than categories, and follow up on the specific absences rather than on the volume. A production of four hundred thousand pages with no pricing analysis in it is a production with a gap, and the gap is the point.

Phase 14. Where the client's own conduct matters

Damages exposure is shaped by decisions made years before the complaint, and a checklist that only addresses litigation misses the part clients can control.

[ ] Mark the products, and require licensees to mark.
- Why. Virtual marking makes compliance cheap: a notation on the product and a maintained page listing the patents.
- Trap. A page that is never updated, so patents issued later are never listed and the period for those patents starts at the complaint.
[ ] Write specific notice letters.
- Why. Identify the patent and the accused product. A letter that starts the clock is worth years of damages; one that does not is worth nothing.
[ ] Keep the licensing file organized and complete.
- Why. Every license will be produced and every one will be used. Knowing what is in them before the case starts is the difference between deploying them and reacting to them.
[ ] Respond to notice letters in writing, internally.
- Why. A defendant that received a detailed letter and has no record of assessing it has nothing to say on willfulness. An internal analysis — even a short one — is the record that answers the question.
- Authority. 35 U.S.C. § 298; note that failure to obtain counsel's advice proves nothing, but the absence of any response at all is a different fact.
[ ] Keep configuration and pricing history.
- Why. The price differential between versions is the strongest apportionment evidence in component cases, and it is only available if the historical price lists were retained.
[ ] Document design-around assessments carefully.
- Why. They cut both ways. A low design-around cost caps the royalty; a high one supports it. Either way the document will be produced.

Phase 15. The damages budget, and proportionality

A full damages program costs real money, and a case whose realistic exposure is below the program's cost should not run the program.

[ ] Compute the ceiling in month one.
- Why. Units in the period times an aggressive rate times a conservative base. If the ceiling is modest, manage the case to that number rather than litigating to it.
[ ] Price the financial discovery.
- Why. It is usually the largest production in a patent case that is not about source code, and the review cost scales with volume rather than with importance.
[ ] Price the expert program honestly.
- Why. A damages expert, possibly a survey, possibly an industry accountant, possibly an economist on market definition — each with a report and a deposition. The exchange runs to hundreds of pages and the cost is predictable.
[ ] Decide early whether a survey is worth it.
- Trap. Commissioning a conjoint study at month eleven because the apportionment evidence turned out to be thin, at a cost that exceeds the difference it will make.
[ ] Consider an early neutral damages assessment.
- Why. A number both sides can negotiate against, produced before either has spent the program's cost, resolves a large share of cases that would otherwise run to expert reports.
[ ] Revisit proportionality at each repricing event.
- Why. Claim construction, institution at the Board, and a marking ruling each change the exposure materially, and the budget should follow.

Phase 16. Reading the outcome for the client

[ ] Explain that compensation is the measure, not the infringer's gain.
- Why. Clients consistently expect disgorgement. It was removed from utility patent law long ago and survives only for design patents.
- Authority. 35 U.S.C. § 289.
[ ] Explain that a royalty is a floor, not a target.
- Authority. 35 U.S.C. § 284.
[ ] Explain that enhancement is rare and discretionary.
- Why. A willfulness verdict does not compel it, and treble is the statutory maximum rather than the usual result.
[ ] Explain that an injunction depends on the equities, not on winning.
- Authority. 35 U.S.C. § 283.
- Why. Competitors often satisfy the factors; entities that license rather than sell generally do not.
[ ] Explain that published verdicts are a biased sample.
- Trap. A client benchmarking exposure against reported awards, which are the small subset of cases that went the distance.
[ ] Explain what happens after the verdict.
- Why. Supplemental damages, prejudgment interest, an ongoing royalty if the injunction is denied, and the appeal. The verdict is not the end of the arithmetic.

Phase 17. Coordination with the rest of the case

[ ] Keep the damages theory consistent with the infringement theory.
- Why. A base built on the whole product sits uneasily with an infringement chart that maps every limitation to one component.
[ ] Keep it consistent with the claim construction positions.
- Why. A narrow construction adopted to survive validity may shrink the accused functionality and with it the apportioned value.
[ ] Account for claims cancelled at the Board.
- Authority. 35 U.S.C. § 282.
- Why. A cancelled claim contributes nothing, and a model built on five claims of which two are cancelled has to be rebuilt.
[ ] Account for the territorial provisions.
- Authority. 35 U.S.C. § 271(f); 35 U.S.C. § 271(g).
- Why. Components supplied abroad and products made abroad by a patented process change which units count.
[ ] Confirm standing supports recovery for the whole period.
- Authority. 35 U.S.C. § 261; 35 U.S.C. § 281.
- Trap. A patent acquired mid-period, where the plaintiff can recover only from the date it acquired the right to sue unless the assignment expressly transferred accrued claims.
[ ] Confirm the exhaustion analysis.
- Why. Authorized sales upstream remove those units from the count entirely.

Phase 18. Presenting the number

[ ] Give the jury one number and one account of where it came from.
- Trap. Two theories presented with equal weight, which reads as uncertainty and invites splitting the difference.
[ ] Anchor on something physical.
- Why. A module the jury can see, a price list, or a two-column product comparison persuades where percentages applied to abstract revenue do not.
[ ] State apportionment in one plain sentence.
- Why. "They should pay for the part they took, not for the whole machine" is the whole concept, and the side that says it first owns it.
[ ] Keep the expert out of report vocabulary.
- Trap. A witness who leans on phrases like smallest salable patent-practicing unit sounds evasive rather than rigorous.
[ ] Answer the other side's number specifically.
- Why. A jury given two numbers and no help splits them. A party that explains precisely why the other number is wrong moves where the middle is.
[ ] Make the willfulness evidence a narrative.
- Why. A letter, a decision not to read it, and an email saying so is a story. The same facts recited as a legal standard are not.
[ ] Prepare the verdict form with counsel who will try the case.
- Why. Whether damages are found per patent, per claim, or in the aggregate affects what survives if any claim falls on appeal.

Phase 19. A one-page assessment memorandum

Every damages case should have a living one-page memorandum, updated at each repricing event and readable by someone who was not in the matter.

[ ] The period. Start date, basis, and the marking and notice findings behind it.
[ ] The units and revenue inside the period, with the source.
[ ] The base, and the evidence for it.
[ ] The apportionment, and the document supporting it.
[ ] The rate range, and the licenses supporting each end.
[ ] Lost profits viability, with the substitutes conclusion in a sentence.
[ ] The ceiling and the floor, as numbers.
[ ] The willfulness posture, in two lines.
[ ] The injunction assessment, and the modeled ongoing royalty if one is unlikely.
[ ] What would change the number, listing the next two events that could move it.

This memorandum is what a client reads before a mediation, what a trial team reads in the week before trial, and what successor counsel reads if the matter changes hands. It takes an hour to update and it prevents the most expensive failure in damages work, which is a team that no longer agrees on what the case is worth.

Phase 20. The failure modes, in one list

Run this against the case at month six and again before expert reports.

[ ] Marking not yet analyzed. The most expensive scheduling error in damages work.
[ ] Licensee products not checked. Where periods collapse.
[ ] Virtual marking page history not pulled. Patents added late start late.
[ ] Notice letter too vague to have started the clock. A self-inflicted truncation.
[ ] Financial requests served too late for the production to arrive before the expert needs it.
[ ] Summary financials accepted in place of transaction-level data.
[ ] The pricing comparison document never requested by name.
[ ] Lost profits pursued without answering the substitutes question.
[ ] Average margin used where incremental margin is required.
[ ] Price erosion claimed without an elasticity adjustment.
[ ] Convoyed sales claimed for items merely sold in the same transaction.
[ ] Whole product used as the base without the demand showing.
[ ] Apportionment asserted by the expert rather than derived from a document.
[ ] Portfolio license divided by patent count and offered as a rate.
[ ] Settlement license used without adjusting for litigation pressure.
[ ] A rule of thumb supplying the rate.
[ ] Damages expert engaged before the technical record was settled.
[ ] Damages report describing the accused feature differently from the infringement report.
[ ] Worldwide units used without a territorial screen.
[ ] Indirect-infringement units counted from before the knowledge date.
[ ] Cancelled claims still in the model.
[ ] Standing gap for the early part of the period, with no assignment of accrued claims.
[ ] Ongoing royalty not modeled before trial.
[ ] No one-page assessment memorandum, so the team no longer agrees on the number.

Each of these is preventable, most of them in the first ninety days, and every one of them has decided a real case.

Outcome. Archived snapshots showed the third patent was added to the virtual marking page eighteen months after issue, and one licensee never marked. A pricing analysis in the defendant's production showed an eleven-dollar difference between the module with the compression feature and its predecessor without it. Those two documents set the period and the ceiling. The case settled at mediation.

Key Authorities at a Glance

| Authority | Proposition | Phase | |---|---|---| | 35 U.S.C. § 284 | Damages; royalty floor; enhancement | 3, 10 | | 35 U.S.C. § 285 | Fees in exceptional cases | 10 | | 35 U.S.C. § 286 | Six-year limitation | 1 | | 35 U.S.C. § 287(a) | Marking and actual notice | 1 | | 35 U.S.C. § 283 | Injunctive relief | 11 | | 35 U.S.C. § 289 | Design patents; total profit | 4 | | 35 U.S.C. § 298 | Advice of counsel not required | 10 | | 35 U.S.C. § 154(a) | Term | 1 | | 35 U.S.C. § 154(d) | Provisional rights | 1 | | 35 U.S.C. § 252 | Intervening rights | 1 | | 35 U.S.C. § 261 | Ownership; standing to recover | 1 | | 35 U.S.C. § 271 | Infringing acts; knowledge for indirect theories | 1, 10 | | 35 U.S.C. § 281 | Civil action | 1 | | 35 U.S.C. § 282 | Defenses; invalid claims yield no damages | 3 | | 35 U.S.C. § 41(b) | Maintenance fees; lapse | 1 | | Fed. R. Civ. P. 26 | Expert report requirements | 8 | | Fed. R. Evid. 702 | Admissibility of expert testimony | 9 | | Fed. R. Civ. P. 56 | Summary judgment on damages issues | 9 |

The five things people get wrong

One: they run marking last. It takes an afternoon, it is answerable from public sources and the licensees' own products, and it decides how many units are in the calculation before anyone argues about the rate. 35 U.S.C. § 287(a).

Two: they concede the base. A defendant that accepts the whole product as the base and argues the rate down from five percent to three has given away most of the money to win a fifth of an argument.

Three: they build lost profits in a market full of substitutes. The substitutes question is answerable from win-loss files and competitor sales data in weeks, and it decides whether the theory exists at all.

Four: they let the expert pick the number first. A rate selected to produce a plausible total and justified backward is the most common route to exclusion, and cross-examination that establishes the order of operations does the same job in front of the jury. Fed. R. Evid. 702.

Five: they ignore the ongoing royalty until after the verdict. Where an injunction is unlikely and the product keeps shipping, the post-verdict rate is the number that governs the client's business, and it is set on a record the parties should have been building all along. See Proving Patent Damages.