How can two analysts study the same technology and reach different conclusions? The answer often lies in the choices they make before they create the first chart.
In our previous article, Patent Landscape Analysis: From Data to Decisions, we explored how patent landscapes use patent data to examine activity and trends in a technology field. To produce useful insights, a landscape must connect patent evidence to a defined business, technical, or policy question. The reliability of the findings depends on the scope, search strategy, data quality, and interpretation of the evidence.
There is no single methodology for building a patent landscape. The right approach depends on the purpose of the analysis, the technology, the available evidence, and the decision it needs to support. Drawing on our experience across multiple patent landscape projects, this article describes one approach we have found useful. Other approaches can also be appropriate, so you can adapt this approach to your own analysis requirements.
The decision that a patent landscape is intended to support determines what the analysis needs to cover. A team evaluating an emerging technology may want to identify organizations beginning to patent in the field. A licensing team may need to assess patent ownership within a technology field covered by an industry standard. Meanwhile, a corporate development team may want to assess an acquisition target’s patent position and its fit with the company’s strategy.
These objectives require different definitions of technology, datasets, metrics, and levels of detail. Before searching for patents, agree on the boundaries of the analysis:
The time period deserves particular attention because patent portfolios, ownership structures, and legal status change over time. If you want to examine how a field or competitor developed, you need a consistent basis for comparing different points in time. A filing date records when a patent application was filed. A reporting date establishes the point in time for which the patent portfolio is being evaluated. Filtering today’s portfolio by older filing dates, therefore, answers a different question from examining the portfolio at an earlier reporting date. For example, a company’s current portfolio may include patents it acquired after that earlier date and exclude patents it has since sold.
Kyocera’s IP Analytics & Intelligence team relies on this distinction. Reconstructing which patents were active at a past date usually means checking grant, abandonment, and expiry dates across many records. With the Reporting Date in LexisNexis® PatentSight+™ with Protégé™, the team can see how its portfolio has changed over time. Analysts can use reporting dates, technology definitions, owner information, and other criteria to build a patent dataset that matches the agreed scope.
Once the scope is clear, the next step is to translate it into a search strategy. Effective patent landscapes usually combine several different building blocks.
Check that the database covers the jurisdictions, years, and information needed for the study. Differences, e.g., in family-building, full-text availability, language coverage, legal-status data, and update frequency, can affect both retrieval and interpretation.
Known relevant patents or their owners can provide a starting point for the search. Keywords can capture the language used to describe a technology in the title, abstract, claims, and description of patent documents. Patent classifications organize inventions into technical categories and can help identify relevant patents that use different terminology. Following references between patent documents or searching for inventors and patent applicants associated with the technology can reveal additional relevant patent families.
Once these elements are combined into a search query, i.e., the instructions used to search a patent database, an initial set of results can be retrieved and reviewed. Subject matter experts (SMEs) can help evaluate whether the results reflect the intended technology. Their feedback can guide further searches and refinements.
This process involves a recurring trade-off. A broad search may retrieve a larger share of relevant patents, improving recall, but introduce irrelevant documents. A narrow search may return a higher proportion of relevant results, improving precision, but exclude patents that use unfamiliar terminology. The appropriate balance depends on the decision being supported.
As you test and refine the search, document:
This record helps other stakeholders understand how the dataset was created. It also allows analysts to reproduce and update the search when new terminology or unexpected results emerge.
LexisNexis® Protégé™ in PatentSight+™, our purpose-built AI assistant, can help structure a search and make its proposed logic visible. Users can describe their question in plain language and review the generated query. Analysts can then transfer that query into LexisNexis PatentSight+ for further validation, refinement, and analysis.
Figure 1: A search string constructed by Protégé™ in PatentSight+™ to identify patents for a patent landscape analysis.
For technology areas that are difficult to define with keywords or conventional patent classifications, LexisNexis® Classification offers an alternative approach. Teams can define their technology scope, train a machine learning classifier using examples of patents to include and exclude, review the results, and refine the model. Classification creates a reusable method for identifying patents that align with the organization’s definition of the field.
The technology and the question being studied also determine which data sources are needed. For landscapes involving standard essential patents (SEPs), technical standards, patent pools, or contributions to standards development, LexisNexis® IPlytics™ can add standards-specific evidence. For complex telecommunications licensing negotiations, this information can provide an evidence base for discussions about fair, reasonable, and non-discriminatory (FRAND) licensing terms.
Before moving on to analysis, validate the search results against the agreed scope. Check whether known relevant patents are included and review a sample of the results for irrelevant documents. Use these checks, together with feedback from technical experts, to refine the search. Several rounds of testing and adjustment may be needed before the dataset is ready for analysis.
Data preparation helps ensure that the comparisons reflect the intended scope. This includes checking for duplicate records, reviewing owner names and corporate group assignments, and recording how missing or uncertain information is handled. The LexisNexis patent database combines automated harmonization with expert review to improve patent ownership data quality and provide a more reliable basis for comparisons.
With a validated dataset in place, the next step is to examine the patterns that help answer the original business, technical, or policy question. Patent analytics tools can help explore:
Patent family counts can indicate the scale of activity, but a simple count assigns the same weight to each patent family. A large portfolio does not necessarily indicate a strong position in a technology field. To assess that position, analysts can complement patent counts with indicators of patent quality. The Patent Asset Index methodology provides an additional perspective for measuring overall portfolio strength.
Within this methodology, each patent family receives a Competitive Impact based on the product of its Technology Relevance and Market Coverage. Technology Relevance assesses how frequently later patents cite an invention, adjusting for patent age, technology field, and differences in patent office citation practices. Market Coverage uses gross national income data to weight the jurisdictions in which a patent family has in-force patents or pending applications, also with adjustments to enable meaningful comparisons. The Patent Asset Index adds up the Competitive Impact values across the selected portfolio, combining portfolio size with the strength of its patent families. Together, these indicators help analysts compare portfolios beyond their size alone.
The choice of indicators should follow the question being asked. Using several complementary indicators can provide a more robust basis for interpretation than relying on a single measure, as they can capture different aspects of patenting activity or strength. Testing whether the main conclusions hold when reasonable alternative filters or indicators are used can strengthen confidence in the findings. Even when several indicators point to a strong patent position, however, assessing commercial value requires additional evidence, such as market demand and the feasibility of bringing the invention to market.
Video: This animated chart tracks the patent portfolios of 10 companies in the same technology field over 10 years. The x-axis shows portfolio size, the y-axis shows average Competitive Impact, and bubble size represents each portfolio’s Patent Asset Index.
Finally, explain the findings for the intended audience. Business leaders rarely need every search string or chart. They need to understand:
We recommend including the following details in stakeholder reports to make the insights transparent and easy to understand: an executive summary, annotated charts, recommendations, and a methodological appendix.
Clear visualizations can help explain complex ownership structures, technology trends, or relationships between patent portfolios. Use charts and tables that address the original question, explain the key assumptions, and link to the supporting evidence. In PatentSight+ with Protégé, you can drill down from a chart to review the underlying patent families. Drilling down from charts to the underlying patent families helps reviewers investigate the exact patent families that are the drivers of a result.
Some companies take this further and bring patent analytics to investors. ASICS, for example, tracks its Patent Asset Index alongside revenue in its integrated report and earnings materials. “PatentSight+ gives us credibility and objectivity,” says Takeshi Horikomi, Managing Executive Officer and CAO. In its 2024 Integrated Report, DENSO uses the Patent Asset Index to benchmark patent quality in its environmental and safety domains. Merck KGaA and Siemens include PatentSight+ analysis in their sustainability reports, Asahi Kasei and InterDigital uses these in their Investor Day presentations. See how companies use patent analytics in investor and sustainability reporting to bring IP into the boardroom.
A patent landscape reflects the evidence available at a particular point in time. New patent applications are published, ownership changes, and patent rights expire or lapse. If the landscape supports an ongoing decision, agree on when to update it. Account for publication delays when interpreting recent trends. Many patent applications become public around 18 months after their earliest priority date, or their filing date if no priority is claimed, so an apparent decline in recent filings may reflect incomplete data rather than reduced activity. Mark recent periods as incomplete where appropriate and allow for delays between publication and inclusion in the database.
Refinement may also be needed during the analysis itself. New findings often expose gaps in the original search or raise questions that require a different analytical view. For example, a chart may reveal an unexpected owner. Reviewing that owner’s patent families could uncover a new keyword, an acquisition, or an adjacent technology category. Analysts can then assess whether the search needs adjusting, test any changes, and record their effect on the dataset.
Validate the analysis with patent-information specialists, technical experts, and the people responsible for the decision. Together, they can assess the search and data structure, technical relevance, and usefulness of the findings.
Once the team has validated the analysis, document the components needed to repeat it. These usually include search queries, workbooks, classifiers, filters, selected metrics, methodological notes, and validated examples. Also retain the final list of patent identifiers, the retrieval and reporting dates, and a record of any manual additions, exclusions, or ownership adjustments. A saved query alone may return different results later as the database changes. This record allows others to understand, reproduce, and update the analysis. It also provides a starting point for similar studies, while leaving room to adapt the scope and methods to a new question. For figures used in reporting, retain the published chart, the underlying data export, and the settings used to generate it. Later database updates can change results for the same reporting date.
A reliable patent landscape connects a clear question with relevant evidence and transparent methods. Its value lies in helping readers understand the findings, their limitations, and what they mean for the decision at hand.
For your next patent landscape, a useful starting point is writing down the decision it needs to support. Agree with the people who will use the findings on what they need to learn, which evidence would help, and what the analysis can realistically answer. Those choices will guide every step that follows.
Ready to put this approach into practice? Get a demo of PatentSight+ with Protégé to explore how it can support a patent landscape focused on your technology and business question.
Watch how GOTTSCHALD Patentanwälte, a law firm that works with global technology leaders, applies patent analytics and AI to strategic client questions in our webinar Turning Patent Data into Business Decisions.
What could patent data reveal about your next business decision? Request a demo of LexisNexis PatentSight+ with Protégé.
Your AI assistant for fast, decision-ready patent insights that drive strategic business decisions.
Senior Project Leader
Carsten holds degrees in Business Administration and Economics from the University of Kiel and earned his doctoral degree from Friedrich-Alexander University Erlangen-Nuremberg. His research focuses on patent and innovation analytics, patent valuation, and the role of intellectual property in investment decision-making. He serves as an editorial board member of Invention Disclosure and World Patent Information.