How to build a case law research matrix for Patent Infringement Matters

Build a patent-infringement case law research matrix by separating legal propositions, verifying each decision’s holding and procedural posture, and comparing its material facts with the asserted claims and accused technology. Record precedential status, subsequent treatment, contrary authority, and unresolved factual dependencies before recommending reliance. The deliverable is a proposition-level research record showing which authorities support a litigation position, which qualify it, and what counsel must still verify. It is not a collection of case summaries or a substitute for limitation-by-limitation infringement analysis.

When to use this workflow

Use this workflow when preparing a claim-construction brief, evaluating an infringement motion, answering an opponent’s authority, or researching a patent-specific disclosure dispute. Outside counsel can use it to assign and review research; in-house counsel can use it to test whether a proposed position rests on verified authority and established technical facts.

This guide addresses US federal district-court litigation. For each assignment, identify the forum, governing-law question, applicable local rules, and judge’s orders. Research for arbitration, state proceedings, ITC investigations, or PTAB proceedings requires a separately verified framework; do not carry this matrix’s procedural assumptions into those forums.

Set a research cutoff and a reliance date. The assignment’s October 13, 2026 cutoff does not establish that every candidate opinion has been checked through that date. Each row needs its own verification record.

What you need

Assemble a research packet containing:

• Asserted patents, issued claims and dependencies, disputed terms, and relevant prosecution passages.
• Current infringement and invalidity contentions, identifying accused products, versions, and theories.
• The motion, opposing brief, or internal question triggering research, with the requested decision and procedural posture.
• Claim-construction orders, relevant scheduling provisions, and applicable local rules.
• Technical evidence with stable references: Bates pages, source-code locations, drawings, declarations, and testimony lines.
• Candidate opinions, official court copies, and access to a subsequent-treatment research service.

Include the research question’s boundaries: for example, literal infringement only, a particular dependent claim, or adequacy of a specified contention. Separate allegations, established facts, and counsel’s inferences. An engineering hypothesis does not become an established fact because it resembles the technology in a favorable opinion.

Use the template

The companion template is Patent Infringement Matters — Case law research matrix. Copy this table into your working document or spreadsheet. Use one row per proposition; repeat a decision in separate rows when it addresses distinct questions.

In Authority / status, record the case name, court, decision date, official opinion link, precedential designation, and proposed weight in the forum. Keep precedential designation separate from your assessment of whether the decision controls the particular issue. Label rules as rules, not judicial precedent.

Posture / law captures the motion or appeal, question resolved, and governing-law dependency. Holding / pinpoint states the narrow proposition in counsel’s words, with the exact page or paragraph. Use the opinion’s page numbering rather than an unexplained PDF-viewer page.

Patent / factual fit identifies claim numbers, disputed limitations, accused versions, and evidence IDs. Limits / contrary records material differences and opposing reasoning. Treatment / checked identifies the check date, reviewer, later authorities reviewed, and proposition-specific result—not merely “good law.”

Give every row a stable ID, such as R-014. Store longer analysis in a working note keyed to that ID. In the final cell, state the reliance disposition, next action, and responsible lawyer. A named assignment is more useful than “follow up.”

Step by step

1. Define the proposition and the decision it serves

Write a question specific enough to answer: “What technical detail must the claimant disclose for the routing-selector limitation under the applicable contention rule?” is more useful than “Research infringement standards.”

Separate claim construction, infringement, invalidity, damages, and procedural disclosure questions. Within infringement research, distinguish the particular asserted theory rather than combining literal, equivalents, and indirect-infringement questions in one entry. Do not assume an authority addressing one theory answers another.

Identify what research can decide and what requires technical confirmation. A disclosure challenge and an argument that the accused product lacks a limitation need different research propositions.

Record: the row ID, issue category, requested decision, scope exclusions, and responsible lawyer.

2. Establish the governing framework before collecting cases

Identify the applicable rule, order, or substantive question. Determine which body of law governs the particular issue and which authorities counsel should evaluate as controlling or persuasive. Escalate uncertain governing-law questions instead of assuming every issue arising in patent litigation has the same answer.

For a district-specific example, Northern District of California Patent L.R. 3-1 requires identification of asserted claims and accused instrumentalities, limitation charts, and specified infringement theories. Rule 1-2 defines the rules’ utility-patent scope, and Rule 1-3 permits court modification. These provisions are not a nationwide infringement standard. Read the official Patent Local Rules with the matter’s orders.

Record: the framework, applicable version, source location, and missing orders. Keep rule text separate from decisions interpreting its application.

3. Search by proposition, posture, and technical dependency

Construct searches around the exact issue, then vary terminology used in opinions. Add the procedural posture and material distinction: a dependent claim, product version, identified component, or absent technical evidence.

Search for rejection as well as acceptance of the proposed position. Follow citations in candidate opinions, but read each relied-upon authority independently. An opponent’s brief or secondary summary supplies leads, not verified holdings.

Track search boundaries in the linked note, including courts searched and excluded candidates. Explain exclusions briefly: different rule, different procedural stage, or materially different record. Do not reject an unfavorable case simply because its technology differs.

Record: candidate authorities, court and date, search scope, exclusion reasons, and unresolved research branches.

4. Extract the holding without enlarging it

Read the complete opinion and surrounding discussion. Distinguish the court’s reasoning from party arguments, background descriptions, and questions the court declined to decide. Check relevant separate opinions without attributing their reasoning to the majority.

State what the court resolved at that stage. If the outcome depended on a missing expert explanation, particular claim wording, or a stipulated construction, preserve that qualification. Do not turn a record-specific decision into an unconditional rule.

Verify quotations and pinpoints against the court-issued opinion. Save the reviewed version and its source reference so another lawyer can reproduce the check.

Record: the narrow holding, pinpoint, posture, and qualifications. Mark incomplete or unread opinions as candidates, not approved authority.

5. Test factual fit against patent-specific evidence

Compare the opinion’s material facts with the asserted claim language, applicable construction, accused implementation, and evidence in the current matter. Record both the analogy and the difference that could defeat it.

For a dependent claim, identify the inherited limitations relevant to the comparison. For software, distinguish versions and code paths. For indirect-infringement research, identify which factual assumptions the candidate decision actually considered rather than borrowing its conclusion without those assumptions.

For discovery, relevance is not the entire inquiry. Rule 26(b)(1) covers nonprivileged matter relevant to claims or defenses and proportional to the case; discoverability does not require admissibility. That framework does not establish entitlement to a particular source-code request. See the Federal Rules of Civil Procedure, Rule 26(b)(1).

Record: exact claim and evidence references, missing facts, their effect on the analogy, and the assigned technical reviewer.

6. Verify status, treatment, and contrary authority

Check the opinion’s designation, subsequent history, and whether the reviewed version was amended, withdrawn, or superseded. Use a treatment service to locate later decisions, then read the passages affecting the proposition.

Separate subsequent appellate history from later judicial discussion. A treatment signal does not explain whether criticism concerns the holding you need, another issue, or reasoning unnecessary to the result. Likewise, absence of a negative signal is not a substitute for proposition-specific review.

Address contrary authority candidly. Explain the legal or factual distinction rather than writing only “distinguishable.” Refresh material rows before a filing or final recommendation.

Record: the check date, relevant later authority, proposition-specific effect, reviewer, and unresolved status questions.

7. Convert the matrix into a research recommendation

Group rows by the decision counsel must make, not by case name. State the best-supported proposition, strongest qualification, factual dependency, and recommended next step.

Use clear internal dispositions: “rely after technical confirmation,” “background only,” “contrary authority to address,” or “further research required.” These are counsel’s work-product labels, not conclusions supplied by software.

Have a second lawyer test the recommended proposition against its pinpoint and the current record. A row may be ready for an internal discussion but not a filing; identify that difference explicitly.

Record: the approved disposition, reviewing lawyer, permitted use, and any assignment required before reliance.

Worked example

Fictional scenario: Aster Circuit LLC asserts claim 7 of a fictional routing patent against Briar Networks Inc.’s Router R4. All parties, products, claims, and matter documents below are fictional. CT-01, page 12, identifies firmware version 4.0 and maps a “routing selector” limitation to a module without explaining the selector’s operation. EN-08, page 3, describes version 4.2. Claim document CL-01, page 2, shows that claim 7 depends on claim 1.

Counsel is considering a contention-sufficiency challenge and technical discovery covering both versions. Assume a Northern District of California utility-patent action, but the team has not reconciled the governing orders. No judicial opinions or version-4.0 operational evidence have been supplied. The following are completed research-assignment entries, not completed precedent analysis. The real rules provide procedural starting points; no fictional judicial holdings are presented.

The matrix changes the immediate recommendation. Counsel cannot use EN-08 as established evidence of version 4.0’s operation. Nor does the rule alone resolve whether CT-01 supplies sufficient detail. Research should first locate decisions applying the relevant disclosure obligation, while technical review resolves the version relationship and inherited limitations.

The discovery row also requires evidence about collection burden and the versions’ relationship. Until those assignments are completed, none of these entries supports a merits conclusion or a finished case-law recommendation.

Common mistakes

• Treating contention decisions as merits decisions. Preserve the posture and limit the proposition to the issue resolved.
• Ignoring claim dependencies or product versions. Record inherited limitations and the precise accused implementation before accepting an analogy.
• Copying a headnote as the holding. Read the court’s discussion and verify the proposition against its pinpoint.
• Recording only favorable reasoning. Include qualifications, opposing decisions, and the strongest contrary interpretation.
• Using a treatment signal as the conclusion. Read later authority and explain its effect on the specific proposition.

Final quality check

• Every recommended proposition has a verified court-issued source and pinpoint.
• Names, courts, dates, opinion versions, and precedential designations match their sources.
• Proposed authority weight is checked for the particular issue and forum.
• Holdings remain distinct from arguments, assumptions, and unresolved questions.
• Claim dependencies, constructions, accused versions, and evidence references match the record.
• Material contrary authority and subsequent treatment have been reviewed through a recorded date.
• Open issues are resolved or assigned; rows blocked from reliance are clearly marked.

Complete the workflow with DecoverAI

DecoverAI is the Legal Intelligence Platform for litigation and transactions. Its litigation preparation capabilities include document review and evidence analysis that can support examination of the factual record used to test a decision’s fit. Tabular review supports work across multiple documents. The matrix remains counsel’s work product: lawyers must verify holdings, authority weight, treatment, and governing law. Book a demo to discuss reviewing the evidence supporting your research matrix.

Frequently asked questions

Should the matrix include rules as well as cases?

Yes, when rules establish the framework. Label them separately and create case rows for disputed interpretation. A matrix containing only rules is a research starting point, not completed case-law analysis.

How many cases belong in each issue group?

Use enough to explain the governing question, material qualifications, and contrary positions. Avoid a numerical target. Remove repetitive entries unless they add a distinct posture, factual comparison, or treatment point.

Can one decision support several patent issues?

Potentially. Create separate rows with separate pinpoints for construction, infringement, invalidity, or damages propositions. Verify each independently; a later decision may affect only one proposition.

What if the technical record is incomplete?

Identify the missing fact and explain its effect on the analogy. Assign technical confirmation separately from legal research. Do not replace absent version or operational evidence with the favorable opinion’s factual narrative.

Sources

• Federal Rules of Civil Procedure, Rule 26(b)(1): discovery scope and proportionality.
• Northern District of California Patent Local Rules, Rules 1-2, 1-3, and 3-1: utility-patent scope, court modification, and infringement disclosures.

These procedural sources support the stated framework, not a complete substantive infringement analysis. The research steps are workflow recommendations for professional readers, not advice for a particular matter.

This guide is general information about legal workflows, not legal advice for any matter. It was drafted and reviewed with AI assistance. The worked example is fictional. Contract rights and procedure depend on the governing law, the actual wording, and the facts, so check the cited sources and the rules that apply in your jurisdiction before relying on it.