2.19Lab to company: IP, tech transfer and spinouts

Patent landscape mapping

Stage: Minus-oneDeeptech

1. Core idea and who it's for#

A patent landscape is a map of who is patenting what in your technology area: the main players, how their portfolios are clustered, the filing trends over time, where the white space is, and which families matter most. It is not a legal opinion. It is a strategy tool, and at pre-seed it answers questions that otherwise get answered badly from instinct:

  • Where should we file? Find the white space and the claim directions that are still open, to steer your own filings (2.20).
  • Who are we really up against? Corporate R&D labs, other universities and quiet startups often show up in patents years before they show up in the press.
  • Who might buy, license or partner with us? The companies filing next to you are your likely acquirers and licensees.
  • Where is the risk? Spot dense, aggressively enforced clusters early, then commission a proper freedom-to-operate search on the specific product (2.06).
  • What do we tell investors? A one-page map of the landscape, with your position in it, answers "what about the incumbents' patents?" before anyone asks.

The WIPO guidelines distinguish between a landscape (the broad picture), a map , a watch or alert (ongoing monitoring), a freedom-to-operate or clearance search , a patentability or prior-art search and a validity search . Each answers a different question with different rigour (WIPO Guidelines for Preparing Patent Landscape Reports). Founders can, and should, do a first-pass landscape themselves with free tools. Paid professional work is for the patentability and FTO questions where money or a legal decision depends on the answer.

Who this is for: pre-seed founders in any patent-heavy field, especially before your first non-provisional or PCT filing decision, before a fundraise, and before choosing a beachhead product.

2. Evidence: the tools and what professional work costs#

Free databases are good enough for a first landscape. Four databases between them cover most of the world's patents:

  • Espacenet (EPO), the oldest;
  • Google Patents ;
  • The Lens , run by the non-profit Cambia;

PATENTSCOPE (WIPO) (PTRCA).

Each has strengths.

  • PATENTSCOPE has about 124 million documents, including every PCT application on the day it is published, and cross-lingual search in 13 languages. A free account lets you save searches, set alerts and download up to 10,000 results to Excel.
  • Espacenet is better for legal status, which tells you whether a patent is still in force. Google Patents is fastest, and the best for exploring citations (Intepat on PATENTSCOPE).
  • The Lens links patents to scholarly literature and citations, which suits academic founders (PatSeer). It also has a dedicated search for biological sequences (Maastricht University Library).

For India, InPASS , the Indian Patent Office's public search, gives full-text search of Indian patents and applications, with wildcards and Boolean operators (Intepat). IISc's IP cell, for example, recommends researchers use these databases (IISc IPTeL).

What professional work costs. Prices vary widely, so treat these as rough ranges:

  • Fixed-price landscape products. One UK firm advertises entry-level landscape reports at a fixed £1,500–£2,000 excluding VAT, and notes they are not FTO opinions (Potter Clarkson).
  • Attorney-led searches. Commonly $1,000–$3,000 for a patentability search (UpCounsel).
  • FTO. $3,000–$10,000 and up for an FTO search (LeanLaw), and $10,000–$30,000 or more for a full FTO opinion with claim analysis (UnitedLex).
  • Indian search firms often charge less than this for search and analytics work.

The implication for pre-seed. Spend founder time, not cash, on the landscape. Use the result to scope a narrow, product-specific FTO and patentability search, which is where money is well spent. A phased approach is common: a broad preliminary landscape first, then detailed analysis once the product design stabilises (Patsnap).

3. Method: a founder-run landscape in about two weeks#

  • Step 1. Frame the questions (half a day). Write two or three questions, for example:
  • Who are the top 15 filers in our technology?
  • Where is nobody filing that matters to customers?
  • Which families could block our first product?

Then define the technology in three to five concept groups. For a sodium-ion battery cathode, the groups might be the cathode composition, the synthesis method, the coating or doping, the cell format, and recycling.

Step 2. Build the search (two to three days).

  • Collect keywords and synonyms for each concept group, including spelling variants.
  • Find the classification codes. Open five highly relevant patents and look at their CPC/IPC codes . Codes catch documents that keywords miss, especially in nonEnglish filings.
  • Combine them, along the lines of (concept A keywords OR codes) AND (concept B keywords OR codes), and test the precision by reading 30 random hits.
  • Run the search in The Lens or PATENTSCOPE , so you can export to a spreadsheet. Restrict it to the last 20 years, and group by family so you aren't counting duplicates.

Step 3. Clean and categorise (three to four days).

  • Export up to a few thousand families. Normalise the applicant names, because "Samsung SDI Co Ltd" and "Samsung SDI" are one filer.
  • Tag each family with your concept groups. Use the abstract and the first claim. You can use AI or keyword rules to speed this up, then check a sample by hand.
  • Record the legal status (alive, pending or lapsed) and the countries where each is filed.

Step 4. Analyse (two days). Make five standard views:

  • filings per year, to see whether the field is growing or maturing;
  • the top applicants;
  • applicants by concept group (a heat map);
  • the geographic spread;
  • the most-cited families.

Then identify white space : cells in the heat map that matter to customers but have few filings. Also identify dense, active clusters near your product, and pass them to your attorney for FTO (2.06).

Step 5. Write the one-pager (one day).

  • the map, with your position on it;
  • the top five players, with one line each;
  • the white space you will file into (2.20);
  • three families flagged for FTO review;
  • likely partners and acquirers.

Use it with your patent attorney and investors.

Step 6. Set up alerts and keep it current. Save the search in PATENTSCOPE or Espacenet with alerts. Review new publications every quarter. The landscape is a living document.

Worked example (illustrative)#

A team spinning out a sodium-ion cathode material runs a landscape:

  • Search. Five concept groups, keywords plus CPC codes for sodium-ion cathodes and their materials, in The Lens, from 2006 to 2026. The search returns 6,400 families. After applicant names are cleaned, five Asian battery makers and two universities hold about 45% of them.
  • Heat map. Layered oxides and Prussian-blue analogues are crowded. Polyanion cathodes made by aqueous, low-temperature synthesis , which is the team's method, have few families, most of them lapsed.
  • Risk. Two live families from one large filer have broad claims on certain dopants near the team's composition. They are flagged for FTO.
  • Decisions:
    • File the provisional on the synthesis route plus composition, aiming at the white space (2.05, 2.20).
    • Commission a focused FTO on the two flagged families, at a cost of about £3,000. Approach one of the top filers as a potential development partner.
  • Fundraising. The one-page landscape goes into the seed deck's IP appendix.

4. India / US / UK differences#

India#

  • InPASS is the Indian Patent Office's public search (ipindiaservices.gov.in/publicsearch). It is essential for Indian filings. Data quality and full-text coverage of older records can be patchy, so cross-check with Espacenet and PATENTSCOPE.
  • Free help: WIPO's Technology and Innovation Support Centres (TISCs) operate in India, often hosted by institutes and CSIR units. IIT and IISc IP cells, and many incubators, also offer searches or training. Indian analytics firms are often far cheaper than UK or US firms.
  • Watch for Chinese and Korean filers , who dominate many hardware fields. Machine translation in PATENTSCOPE and Espacenet makes their filings searchable.

US#

  • USPTO Patent Public Search gives full US text and prosecution histories. Patent and Trademark Resource Centers (PTRCs), usually in libraries, offer free help with searching.
  • For your landscape, note that US prosecution histories can be read through Patent Center. They show what claims the examiner rejected and why, which is useful for judging how broad a competitor's patent really is.
  • For FTO, US litigation and damages risk is the highest in the world. Pay for a professional FTO on US-critical products.

UK#

  • The British Library's Business & IP Centre network offers free access to commercial databases and workshops across UK cities.
  • The UK IPO publishes free patent landscape reports on selected technologies, which are worth checking before you start.
  • The European Patent Register and Espacenet give EP legal status, including oppositions, which matters because European patents can be challenged centrally within nine months of grant.

Scripts#

Briefing your patent attorney:

"We've mapped about 6,000 families in our area, and here's the one-page summary. We believe the white space is X, so we'd like our claims to aim there. Separately, these three families are close to our first product. Could you quote for a focused FTO on just these, for the US and Europe?"

In an investor meeting:

"We've mapped the landscape. The incumbents cluster in layered oxides, and our synthesis route sits in a sparsely patented area. We've filed there, and counsel is reviewing the two closest third-party families."

To a potential partner who appears in the landscape:

"Your filings in X show you've worked on the same problem. Our approach solves the yield issue at lower temperature. Would your team be open to a technical conversation under NDA?"

Common mistakes#

  1. Treating a landscape as FTO. A landscape shows where risk might be. Only claim-byclaim analysis by counsel says whether you infringe.
  2. Keyword-only searches. Without CPC codes you miss non-English filings and documents that use different terminology.
  3. Counting documents rather than families. One invention filed in 10 countries looks like 10 patents.
  4. Ignoring legal status. Lapsed patents are free to use, and live ones are not.
  5. Doing it once. Without alerts, a competitor's new filing goes unnoticed for 18 months.
  6. Paying for a big landscape report before you know the questions. Do the first pass yourself.
  • 2.04 Patent or publish first
  • 2.05 Provisionals and the PCT
  • 2.06 Freedom to operate
  • 2.18 Trade secrets vs patents
  • 2.20 IP budget at pre-seed
  • 1.08 Sizing a market that does not exist yet

Sources#

The worked example is illustrative. The cost figures are indicative market ranges, not quotes. Last checked November 2026.

← 2.18 Trade secrets vs patents for process technology2.20 IP budget: what to file at pre-seed →