Table of Contents
Patent competitive intelligence is the discipline of reading a rival’s research strategy from the one place it becomes public before the product does: the patent record. Every competitor tells the world what it is building eighteen months ahead of launch, in the claims, classifications and filing dates of its applications. This page explains what a patent competitive intelligence program actually delivers, the public data it is built on, and how PerspireIP turns millions of filings a year into four decisions a strategy team can act on this quarter.
What Patent Competitive Intelligence Actually Tells You
Patent competitive intelligence answers a question no sales deck will: what are your competitors actually spending their R&D budget on, and how fast? A patent application is a dated, classified, public disclosure of a technical bet, typically published eighteen months after filing and long before the product reaches a shelf. Read across a field, those disclosures form a map of intent — which problems each rival is trying to solve, which they have abandoned, and which they have not touched at all.
This is not the same as counting patents. A raw filing count tells you who is busy; competitive intelligence tells you what the filings mean — the direction of a program, the pace of investment, and the defensive walls going up around a technology. The value is in the reading, not the retrieval, and the reading is what separates a strategy input from a spreadsheet.
The Raw Material: 3.7 Million Filings a Year
The public record competitive intelligence draws on is vast and growing. According to WIPO’s World Intellectual Property Indicators 2025, applicants filed an estimated 3.7 million patent applications worldwide in 2024, up 4.9% and the fifth straight year of growth. Around 2.1 million patents were granted that year, and roughly 19.7 million patents were in force across 142 jurisdictions. Each of those documents is a searchable, dated, classified data point.
The filing base is also concentrated. China’s CNIPA alone received about 1.8 million applications — 49.1% of the world total — ahead of the USPTO (603,194), Japan (306,855), Korea (246,245) and the EPO (199,402). For a competitive intelligence read, that concentration matters: a rival’s true filing footprint often shows up first in its home office, so a program that only watches the USPTO misses half the picture.
The tooling has kept pace. The EPO’s Espacenet now provides free access to more than 150 million patent documents dating back to 1836, which means the barrier to competitive intelligence is no longer access to data — it is the analysis that turns 150 million documents into an answer about four rivals.
Reading Competitor Filing Trajectories
The first deliverable is a filing-trend analysis: a rival-by-rival curve of applications over time, split by technology area. Trajectory is more informative than volume. A competitor whose filings in a sub-field doubled over three years is telling you where its next products are aimed; one whose filings flatlined has either solved the problem or quietly killed the program.
Timing is the edge. Because applications publish about eighteen months after filing, a trajectory read surfaces a competitor’s direction well before the market sees a product. Pair the curve with priority dates and you can often date the start of a rival program to the quarter — intelligence that a strategy team can act on while there is still time to respond, rather than after a launch has already reset the field.
Technology-Cluster Maps: Where the Bets Are
A technology-cluster map groups a field’s patents by what they claim rather than who filed them, using the International Patent Classification and citation links to reveal the natural clusters of a technology. Overlay each competitor’s portfolio on that map and the strategic picture resolves: dense clusters are the contested ground everyone is fighting over; sparse regions are either dead ends or genuine white space.
The EPO’s Patent Index 2024 shows why the map, not the count, is what matters. In 2024 computer technology became the top field at the EPO for the first time with 16,815 applications, just ahead of electrical machinery, apparatus and energy at 16,142 — which grew 8.9% — while digital communication slipped 6.3% to 15,983. Three fields separated by a rounding error, all moving in different directions: a headline count would flatten that, but a cluster map makes each rival’s shift legible.
Assignee Benchmarking: Who Leads Each Field
Assignee benchmarking ranks the players in a field and positions your own portfolio against them. It answers the board-level questions directly: are we in the top tier of filers in the technologies that matter to us, or are we being out-invested by a competitor we had not been watching? At the EPO in 2024, Samsung led all applicants with 5,107 filings, overtaking Huawei (4,322), with LG (3,623), Qualcomm (3,015) and RTX (2,061) completing the top five.
Benchmarking is most useful when it is narrowed to your exact technology cluster rather than the whole office. A firm ranked twentieth overall may be first in the one sub-field where you compete — and that local leadership, invisible in an office-wide table, is precisely what a patent competitive intelligence read is built to surface.
Finding the White Space Between Rivals
The most valuable output of a competitive intelligence program is often what is not there. White-space analysis reads the same cluster map for the gaps — technical problems with commercial pull that no incumbent has yet claimed. A defensible gap is an R&D thesis and a filing opportunity at once: a place to build where you will not immediately read on a competitor’s wall.
White space has to be read carefully, because not every empty region is opportunity — some gaps are empty because the technology does not work or the market does not exist. That is why we cross-check apparent white space against filing trajectories and the wider market: a gap that rivals are visibly circling is a race, while a gap they have all avoided deserves a hard question before you commit budget. Our dedicated white-space analysis service takes this step further for teams planning where to file next.
Patent Competitive Intelligence vs Patent Landscape Analysis
The two terms are often used interchangeably, but they answer different questions. A patent landscape maps a whole technology — the state of the art, the major players, the trends — to orient a team new to a field. Competitive intelligence is narrower and more adversarial: it tracks named rivals over time to inform a specific decision, such as whether to enter a market, how to position around a competitor, or where a rival is about to attack.
In practice a program often starts with a landscape to establish the map, then runs competitive intelligence on top of it as a standing watch. If you are new to the field, begin with our guide to competitive patent landscape analysis; if you already know the players and need to track them, our competitor patent analysis walkthrough covers the recurring read. The engagement adds the analysis and the standing cadence a one-off guide cannot.
Where Competitive Intelligence Reads Go Wrong
Most flawed competitive intelligence reads fail in the same few ways, and each one is avoidable. The first is name fragmentation: a single competitor files under subsidiaries, acquired entities and romanised name variants, so a search on the parent alone undercounts its real footprint. Before any trend is drawn, the assignee names have to be normalised and the corporate tree reconstructed, or a rival’s program looks half its true size.
The second is watching one office. Because filings are concentrated at home offices — CNIPA took 49.1% of the world’s applications in 2024 — a read limited to the USPTO or EPO can miss where a competitor’s earliest and densest filings actually sit. A serious program pulls each rival’s home jurisdiction alongside your target markets, then de-duplicates by patent family so the same invention filed in five countries is counted once, not five times.
The third is mistaking volume for direction. A spike in a competitor’s filings can signal a genuine new program or simply a change in filing policy, and only reading the claims tells you which. The fourth is treating white space as opportunity by default: a gap can be empty because the technology fails or the market is not there. Cross-checking each apparent gap against filing trajectories and market signals is what keeps a competitive intelligence readout from pointing a strategy team at a dead end.
How PerspireIP Runs a Competitive Intelligence Program
We start from your commercial question, not from a keyword. “Are we losing the solid-state race?” and “where should we file next in power electronics?” need different searches, different assignee sets and different time windows, so we scope the program to the decision it will feed. From there we build the competitor set, pull the filings across the offices that matter — including each rival’s home office — and normalise the assignee names, because a single competitor may file under a dozen legal entities.
The analysis then produces the four deliverables above: trajectory curves, a technology-cluster map, an assignee benchmark, and a white-space readout, each tied back to the decision that prompted it. Because filings publish continuously, we can run the program as a one-time snapshot before a strategic decision or as a standing quarterly watch that flags a rival’s new cluster the quarter it appears — which, in a fast-filing field, is the difference between responding and reacting.
The Unitary Patent has added a wrinkle worth watching. In 2024 roughly 28,000 granted European patents — about a quarter of all EPO grants — carried a Unitary Patent request, changing how rivals build coverage across Europe. A current competitive intelligence read has to account for that shift, because a competitor’s European wall now looks different than it did three years ago.
What You Receive
Every engagement ends in a decision-ready readout, not a raw export of patent numbers:
- Competitor filing-trend analysis — who is accelerating, who is pulling back
- Technology-cluster map showing where each rival is concentrating R&D
- Assignee benchmarking against the leaders in your field
- Strategic white-space readout — the gaps no incumbent has claimed yet
Data Sources & References
This analysis draws on primary patent and market data:
- WIPO — World Intellectual Property Indicators 2025 (Patents) — 3.7 million applications filed worldwide in 2024 (+4.9%); 19.7 million patents in force across 142 jurisdictions; CNIPA 49.1% of filings.
- EPO — Patent Index 2024 — Computer technology the top EPO field for the first time; Samsung the leading applicant; ~28,000 Unitary Patent requests.
- EPO — Espacenet patent search — Free access to more than 150 million patent documents from 1836 to today — the raw material of a competitive intelligence read.
Request a Competitive Intelligence Readout
Tell us the rivals and the technology you need read, and we will scope a patent competitive intelligence program around the decision it has to feed.
Request a Competitive Intelligence Readout
Related PerspireIP work: Competitive patent landscape analysis guide · Competitor patent analysis walkthrough · White-space analysis service.
Frequently Asked Questions
What is the difference between patent competitive intelligence and a patent landscape?
A patent landscape maps a whole technology to orient a team new to a field. Patent competitive intelligence is narrower and tracks named rivals over time to inform a specific decision, such as whether to enter a market or where a competitor is about to file next. Programs often run competitive intelligence on top of an initial landscape.
How current can competitive intelligence be if patents publish late?
Applications typically publish about eighteen months after filing, so a trajectory read surfaces a competitor’s direction well before its product reaches the market. Pairing published applications with priority dates lets us date the start of a rival program, often to the quarter — earlier warning than any product signal.
Which patent offices should a competitive intelligence program watch?
At least each competitor’s home office plus your target markets. Filings are concentrated — CNIPA received 49.1% of the world’s applications in 2024 per WIPO — so a rival’s true footprint often appears first at home. A program limited to the USPTO can miss half of a competitor’s portfolio.
Is competitive intelligence just counting competitors’ patents?
No. A raw count tells you who is busy; competitive intelligence tells you what the filings mean — the direction and pace of each rival’s program, the technology clusters they are building walls around, and the white space none of them has claimed. The value is in the reading, not the retrieval.
Should a competitive intelligence program be one-time or ongoing?
Both have their place. A one-time snapshot suits a discrete decision such as a market-entry or acquisition. A standing quarterly watch suits a fast-filing field, flagging a rival’s new technology cluster the quarter it publishes — the difference between responding early and reacting after a launch.