Prior art search before filing

What a preliminary search is for

Three purposes, in order of value. First, catching the case where the invention already exists plainly and cheaply, before any fees are committed. Second, learning the vocabulary and classification of your field, so your invention can be described in terms an examiner and a drafter will recognise. Third, producing a documented position: here is the closest art I found, and here is the technical difference.

How to run one

1. Write the invention down first. One sentence: the problem, the mechanism, the boundary. Searching before you can state the invention produces noise.
2. Build a vocabulary list. Your term, the industry term, the academic term, and the terms used in the first patents you find. Fields routinely have three or four names for the same component.
3. Search several databases. Espacenet, Google Patents, WIPO PATENTSCOPE, Lens, and your national office. Coverage and machine translation differ.
4. Follow classifications. When you find one relevant document, use its CPC or IPC classes to browse neighbours. Classification browsing surfaces documents that keyword search misses because they use unfamiliar language.
5. Search non-patent literature. Papers, theses, standards, product manuals, catalogues, teardown videos and marketplace listings are all prior art.
6. Record as you go. Document identifier, date, what it discloses, and how your invention differs. A search you cannot reproduce has little value later.

Common mistakes

• Searching only your own words. Your terminology is often the least used in the literature.
• Searching only granted patents. Published applications, expired patents and abandoned filings are all prior art; expiry affects enforceability, not disclosure.
• Stopping at the first reassuring result. Absence of an identical document is not evidence of novelty.
• Reading only the claims. The description and drawings usually disclose far more than the claims cover.
• Treating a clean search as a green light. A negative preliminary search has real but limited value; it removes a reason to stop, it does not create a reason to proceed.

Where the professional stage begins

Patentability turns on how a claim would be read against the state of the art, on inventive step over combinations of documents, and on jurisdictional practice. That assessment belongs to a qualified patent professional. Your searching does not replace it — it lowers its cost and raises its quality, because the conversation starts from documented ground rather than from a blank page.