Turn Patent Rankings into a Technology Research Shortlist

Decide what the shortlist is for

Choose the immediate purpose: investigate alternative process routes, understand a competitor, find technical expertise or identify possible collaborators. These purposes can produce different lists from the same ranking.

For example, a team studying advanced oxidation for water reuse might need to understand how oxidation is combined with filtration. That question directs attention toward treatment sequence and integration. A team studying catalyst materials would inspect different parts of the same field.

Define the application before sorting candidates. What water stream, contaminant, treatment objective and operating constraints matter? If those inputs are still open, treat the shortlist as exploratory and retain the questions that would narrow it.

Understand what the ranking counts

For a concrete starting point, Patsnap’s Advanced Oxidation Process Patent Landscape, published July 13, 2026, reports 1,543 patent families within its search-defined, multi-jurisdiction AOP corpus. The first five entries in its applicant ranking are:

Source: Patsnap Eureka; Air Liquide’s name is shortened here. This ranking compares applicant activity by patent-family count within the report’s search sample, rather than ranking technical quality.

A publication is a document; a patent family groups related applications through priority relationships. Simple and extended families can group records differently. A company’s publication count and its family count therefore measure different things.

Keep the unit beside the number. Do not divide technology-tagged publications by a family total to calculate a portfolio share. If a record carries several classifications, it can appear in more than one technology category; those categories need not form mutually exclusive slices.

Also establish the company boundary. Counts for a corporate group can differ from counts for one legal entity. Joint applicants may appear in more than one applicant total. If the grouping or treatment of joint applications is unspecified, avoid calculations that assume it is known.

Dates matter for the interpretation. A priority-year view, a filing-year view and a publication-year view organize activity differently. Record which is used, the search cutoff and the coverage window. Do not interpret a recent decline until you know whether the period is sufficiently complete.

Finally, examine the search boundary. Keywords, classifications and exclusions decide which records enter the ranking. A broad treatment-system search and a narrow catalyst search can produce very different leaders without either contradicting the other.

Move from a company name to a technical approach

Open records selected for relevance to your application. A highly ranked company may have most of its activity in a different subsystem; a lower-ranked specialist may address your precise problem.

A separate patent example below shows how to assess technical relevance at document level.

A concrete example is US20130104606A1, “Water purification system and washing machine having the same”, which lists Samsung Electronics as the original applicant. Its description combines physical filtration with a photocatalytic advanced-oxidation stage for washing or rinsing water. Figures 3 and 4 and their accompanying description show water passing through the filter and AOP units in sequence.

For a water-reuse investigation, this makes the document a lead on process integration. It does not, by itself, establish suitability for an industrial effluent stream. The next technical questions concern the water composition, treatment conditions and evidence needed for that different application.

This is how an individual document changes a shortlist entry. “Large applicant in water treatment” becomes “source describing a filter-plus-oxidation sequence, worth inspecting for integration ideas.” The latter tells another engineer why to spend time on it.

Read the relevant description, drawing and reported results together. Separate the configuration described from the performance demonstrated. Preserve the publication number and exact location so another reviewer can inspect the same evidence. This technical reading does not establish current patent rights or permission to commercialize a design.

Build a source-linked shortlist

Write each entry around the work you want to do next. Here is a completed example based on the document above:

Keep the initial candidate list separate from entries whose sources you have actually reviewed. An uninspected high-ranking company remains a lead. An inspected candidate earns its place through relevance, even if it is lower in the ranking.

When the shortlist contains several approaches, compare them using your project constraints. Identify which source check would most change the decision. The final handoff should make it clear what to read, what to compare and where experimental evidence is still needed.

Develop your shortlist with Patsnap Eureka

Patsnap Eureka’s competitor research workflow combines company profiles, focused technical research and landscape context. Use the landscape view to identify players, then take the most relevant technical question into a source-linked report. The report outline can be adjusted before generation, helping keep the research focused on your application. Patsnap Eureka competitor technology research.

Start with your technology boundary and shortlist purpose. Ask which players describe the mechanism you need, then inspect the supporting patents before narrowing the list. Preserve the source identifiers and unresolved application questions in the result you share with your team.