Most prior art searches underdeliver not because of bad searching, but bad briefing. These 10 questions change that.
Why the Brief Determines the Result
In patent prosecution, IP due diligence, freedom-to-operate analysis, and IPR defense, the prior art search is often where cases win or lose. Yet many attorneys treat the prior art search brief as a checkbox exercise—a perfunctory document sent to a search vendor with minimal context or deliberation.
Here’s the uncomfortable truth: a search analyst can only find what you define. The most skilled searcher in the world cannot compensate for a vague, incomplete, or strategically misaligned brief. If you don’t tell them what you’re really looking for, they’ll deliver exactly what you didn’t ask for.
A poorly framed brief produces:
- Too many irrelevant results (noise that obscures the critical references)
- Missed references because key terminology or concept space wasn’t defined
- Searches anchored to wrong claim interpretations or invalid classification codes
- Results misaligned with your actual use case (prosecution vs. litigation vs. FTO produce very different search strategies)
- Weak claim charts that won’t survive attorney review or cross-examination
The fix is not better searchers. The fix is better briefing. And better briefing starts with asking the right questions—before you commission the search.
The 10 Questions to Ask
- What is the exact claim language we’re searching against?Don’t paraphrase the claims. Provide the exact independent and dependent claims you’re evaluating. Better yet: provide both the broadest reasonable interpretation (BRI) claim language and the narrower dependent claims. This tells the searcher what term set to use, what limitations matter most, and where to focus scope. A search against a paraphrased claim is a search in the dark.
- Are we searching for validity or patentability?This is the fundamental fork. A validity search (used in IPR, reexamination, or litigation) looks backward from the claimed invention’s priority date to find references that anticipate or render obvious the claims as filed. A patentability search (used in prosecution or freedom-to-operate) often looks at the current state of the art, may focus on dependent claims or narrower scope, and is designed to support argument rather than to find the strongest possible rejections. The search strategy, scope, and database mix change dramatically based on this answer.
- What technology classification codes (CPC/IPC) should anchor the search?The Cooperative Patent Classification (CPC) and International Patent Classification (IPC) codes are not optional scaffolding—they are the backbone of any systematic search. Specify which codes the invention falls under, which the searcher should weight most heavily, and which to explore as adjacent art. Many attorneys leave this to the searcher. Don’t. Your claim familiarity should drive the classification roadmap.
- Which jurisdictions matter most for this matter?If this is US prosecution, US prior art is primary. If this is a global freedom-to-operate, you need European, Japanese, and Chinese patents. If this is an IPR petition against a US patent, the scope is narrowed to US references and certain foreign publications. Different jurisdictions have different publication practices, examination strictness, and technology strengths. Make your needs explicit.
- What is the critical date for prior art purposes?For a utility patent, this is typically the earliest priority date or filing date of the application. For an IPR, this is the issue date of the patent under attack. For a design patent, the rules differ again. For FTO analysis, it’s often today’s date. Specify the critical date upfront so the searcher doesn’t waste effort on post-date references or miss the temporal boundaries of what counts.
- Who are the known players in this technology space?Provide a list of known competitors, academic labs, or institutional players working in the field. This helps the searcher calibrate scope and identify likely sources of innovation. It also flags whether you’re looking for foundational art (early and broad) or state-of-the-art references (recent and narrow). Many critical references come from known assignees; give the searcher a head start.
- Are there any non-patent literature (NPL) sources we should prioritize?Academic papers, standards documents, technical books, medical journals, trade journals, conference proceedings, and technical specifications can all be prior art. Different fields have different NPL traditions: software relies on technical whitepapers and GitHub discussions; pharmaceuticals rely on peer-reviewed journals; electrical engineering relies on IEEE and conference papers. Tell the searcher which NPL sources matter and why.
- What level of claim chart detail do you need in the output?Do you want a simple list of references with basic relevance scores? Or do you need claim-by-claim maps showing exactly how each reference reads on each limitation? The former is faster and cheaper; the latter is prosecution-ready and litigation-defensible. Be clear about deliverable format so the searcher allocates time appropriately.
- What is the intended use of the results—prosecution, IPR petition, FTO, or litigation?This determines confidence levels, documentation standards, and the evidentiary weight of the analysis. A search brief for prosecution needs references that support narrow claim language and overcome specific rejections. An IPR petition needs the strongest possible references that read on the broadest claims. An FTO analysis needs to identify all plausible infringement risks. A litigation search needs to withstand cross-examination from opposing counsel with access to different databases. State your true use case.
- What is the timeline, and is a phased search approach appropriate?If you need results in 24 hours, the searcher prioritizes speed and high-confidence databases. If you have 4 weeks, the searcher can perform deeper analysis, explore adjacent technology spaces, and run supplementary searches. Some matters benefit from a phased approach: first phase identifies core references quickly; second phase explores deeper if needed. Misaligned timelines produce rushed analysis or wasted effort. Be explicit about deadline and scope flexibility.
The difference between a prior art search that holds up under scrutiny and one that falls apart in prosecution often comes down to the quality of the brief, not the quality of the searcher.
Three Things That Guarantee a Weak Search
1. Bad Claim Mapping
Many search briefs describe the invention in marketing language: “a revolutionary AI-powered diagnostic tool.” Then the searcher has to reverse-engineer the actual claim language on their own—and often gets it wrong. The search becomes a quest for any reference related to “AI” and “diagnostics,” not a focused hunt for references that anticipate or render obvious the specific claim limitations.
Fix: Provide exact claim language. Highlight the independent claims and key dependent claims. Show which claims are broadest and which are narrowest. If claim language is complex or uses terms of art, define them explicitly.
2. Jurisdiction Blind Spots
Many US practitioners commission “prior art searches” that implicitly mean “US prior art searches.” But if your patent is vulnerable to foreign prior art, or if your client operates globally, missing European, Japanese, or Chinese references is a critical failure. Conversely, if you’re prosecuting an application at the USPTO, you don’t need to pay for searches in JPO databases—that’s unnecessary scope.
Fix: Explicitly state which jurisdictions matter. If global, say so. If US-only, say so. If you’re unsure, ask your search vendor what databases will be used and why.
3. Ignoring Non-Patent Literature
Searches anchored only to patent databases miss entire categories of prior art. In software, key references live in GitHub repositories and technical whitepapers. In pharmaceuticals, clinical trial data and peer-reviewed journals are critical. In electrical engineering, IEEE papers and conference proceedings matter. If your brief doesn’t mention NPL, the searcher will default to patent databases—and miss references that could be fatal to your case.
Fix: Tell the searcher: “We need patent databases AND IEEE Xplore for this search” or “Include PubMed and medical journals.” Make NPL part of your strategic requirements, not an afterthought.
Your Prior Art Search Briefing Checklist
Use this checklist before you send your next search request to a vendor. If you’re missing answers to any of these, send your vendor a follow-up brief.
Critical Briefing Elements
- Exact independent and dependent claims (not paraphrased)
- Claim interpretation guidance (broadest reasonable interpretation vs. narrow)
- CPC/IPC classification codes that anchor the search
- Critical date (priority date, filing date, or issue date)
- Validity search or patentability search (explicit choice)
- Jurisdictions included (US only, European, Japanese, Chinese, global)
- Known competitors and assignees in the field
- NPL sources (academic journals, standards, conference papers, etc.)
- Deliverable format (list with scores, claim charts, detailed write-ups)
- Intended use (prosecution, IPR, FTO, litigation)
How Elpis Analytix Handles Prior Art Search
At Elpis Analytix, we’ve analyzed thousands of prior art searches—many that worked, many that didn’t. We built our prior art search process around the principle that a great search starts with a great brief.
Our Process
Step 1: Intake and Clarification
When you commission a search through Elpis, we don’t accept a vague brief. Our intake team conducts a structured conversation: we review the claims with you, confirm the intended use case, establish scope boundaries, and document assumptions. This conversation often surfaces critical gaps that improve the search before the analysis even begins.
Step 2: Search Strategy Development
We develop a documented search strategy that specifies database selection, CPC/IPC roadmaps, terminology expansions, and phase gates. You review and approve the strategy before we execute. This transparency ensures alignment between what we’re looking for and what you actually need.
Step 3: Systematic Multi-Database Search
We search across 15+ authoritative databases:
- USPTO (US patents and published applications)
- EPO (European patents and publications)
- WIPO (PCT patents)
- JPO (Japanese patents)
- Google Patents (comprehensive multi-jurisdiction index)
- IEEE Xplore (electrical engineering, software, telecommunications)
- PubMed/MEDLINE (medical and pharmaceutical literature)
- Academic databases (including Scholar, ResearchGate, institutional repositories)
- Standards organizations (ANSI, IEC, ISO)
- Industry-specific databases (depends on your field)
Step 4: Analysis and Claim Charting
Each reference is evaluated against your claim language. We produce detailed claim charts showing how each reference reads on each limitation, highlighting gaps and strongest possible combinations. We also identify secondary references that support validity arguments but don’t independently anticipate the claims.
Step 5: Defensible Documentation
Our work is performed under NDA and attorney-client privilege when appropriate. We document our search methodology, terminology choices, and database coverage so the search can withstand scrutiny in prosecution, litigation, or IPR proceedings.