DIY patentability search positioning and limitations

A DIY patentability search is a preliminary screening activity. It can reveal obvious close documents, improve the vocabulary used to describe the invention and create a record that a patent professional can review. It should not be used to promise novelty, predict allowance or make an irreversible filing decision.

The USPTO describes a self-conducted search as preliminary and warns that an examiner may reject claims using information not found in that search.¹ That warning applies even when the first pass produces no close result.

When a team needs a more structured AI-assisted route, Eureka IP Search’s Novelty Search extracts core technical features, builds multiple search strategies and compares close prior art feature by feature.² Its source-linked output still requires review and does not guarantee complete coverage or a particular patent-office decision.

What a first pass can do

• Reveal a document that appears to disclose the central technical idea.
• Show how patent documents describe the product, function, structure and materials.
• Create seed documents for classification, citation and family searching.
• Prepare a concise list of similarities, differences and unresolved questions.

When it cannot replace professional review

• The invention is a core platform, a major commercial asset or the basis of an imminent filing.
• Protection is being considered in multiple jurisdictions.
• Several close documents exist and their effect on the proposed claims is unclear.
• The subject involves chemical structures, biological sequences, complex algorithms or another specialist search environment.
• The relevant date, public availability or family history of a reference is uncertain.

Preparation before starting the clock

Prepare enough information to describe the invention consistently. Do not start from a slogan such as “a smarter phone stand.” Explain what the product or method is, what problem it addresses, which components or steps interact and what technical feature appears to distinguish it.

• Concise invention description: a short page or structured note written in plain language.
• Core technical features: a manageable list of features that explain the proposed contribution.
• One-sentence concept: the product or method, the technical action and the distinctive implementation.
• Known products or patents: product names, manufacturers and publication numbers, where available.
• Search record: a document for queries, dates, databases, candidate references and observations.
• Timer: sixty minutes for the exercise, with the option to stop early when a specialist issue appears.

If the source material is fragmented, Eureka IP Drafting’s Invention Disclosure workflow can turn rough engineering notes into a structured disclosure and highlight missing details.³ This improves the search input, but it remains separate from a legal patentability opinion.

One-hour DIY patentability search workflow

First 5 minutes: extract three sets of keywords

Create a small keyword bank organized by concept rather than one long query.

• Product or method terms: phone stand, phone holder, desk mount, portable support.
• Technical function terms: support, position, fold, rotate, adjust height.
• Distinctive implementation terms: magnetic coupling, articulated hinge, telescopic column, detachable base.

Add Chinese and English variants when they are relevant to the field. Include broader functional language as well as the inventor’s preferred terminology. The goal is to test several expressions, not to meet a fixed keyword count.

Minutes 5–30: core search on Google Patents

Google Patents provides free access to patent text, images, PDFs, metadata and citations. Its query behavior should be used deliberately: separate search boxes are ANDed, exact phrases can be placed in quotation marks, and Advanced Search supports Boolean logic and field restrictions.⁴

Round 1, minutes 5–15: broad product-and-function search

Start with the product and its main function. For example:

Do not assume that typing several words merely asks for semantic similarity. Google explains that the default operator is AND, while close synonyms and plural forms may be included.⁴ If results are too broad, add one supported structural term. If they are too narrow, remove a constraint or substitute a broader function.

Open a manageable set of records that appear technically close. Review the title and abstract first, but also inspect drawings and the relevant claims before treating a document as significant.

Round 2, minutes 15–25: language and terminology variants

Run the same concept through relevant Chinese and English expressions. Google states that patent documents available only in non-English text are machine-translated into English and indexed, which can help English-language discovery.⁵ Machine translation is useful for orientation, but important technical and legal wording should be checked against the original publication or a reliable family member.

Coverage is not guaranteed to be complete. Google also advises prior-art searchers to consider non-patent literature in fields where papers, manuals or other technical publications matter.⁵

Round 3, minutes 25–30: distinctive features and classifications

Combine the product concept with the implementation that appears most distinctive:

Use a relevant CPC shown on a close seed document to create another search path. The USPTO multi-step strategy recommends moving from terminology to detailed document review, relevant CPC classifications, cited references and foreign or non-patent sources.⁶

Minutes 30–40: rapid citation and similarity tracking

Choose the closest seed documents based on technical relevance, not applicant size. In Google Patents, inspect cited and citing documents, family information and Similar Documents. Google states that the Similar Documents list is based on text similarity.⁷

Citations and similarity results are expansion routes, not validation labels. A cited document may be background, and a textually similar document may not disclose the required technical relationship. Record why each newly opened document matters.

Minutes 40–55: organize and compare

Stop adding queries long enough to compare the closest records. Use a simple evidence table:

Do not decide relevance from publication date alone. Prior-art effect depends on the applicable law, relevant dates and public availability, so uncertain timing should be escalated for professional review.

A rapid novelty check

Ask whether one earlier disclosure appears to contain every essential feature in the required arrangement. Under the current EPO novelty guidance, separate items of prior art cannot be combined to destroy novelty.⁸ Therefore, a single close record deserves careful review, but the result still depends on what it actually makes available to the public and how the proposed claim is framed.

If the features appear only across several documents, do not label the invention unpatentable. Combination questions belong to inventive-step or obviousness analysis. The EPO’s current guidance considers whether a skilled person would combine disclosures and warns that the need for multiple disclosures may, in some circumstances, indicate an inventive step.⁹

Minutes 55–60: document and assess next steps

Finish by recording:

• the databases, date and language searched;
• each exact query and major filter;
• the closest publication numbers and relevant passages;
• the features that appear disclosed and the differences that remain;
• search paths not completed, including classifications and non-patent literature;
• questions for the inventor or patent professional.

The output is a documented first pass, not a legal rating. Good records let a later reviewer reproduce the search, correct assumptions and continue from the most promising seed documents.

Self-check for the one-hour DIY patentability search

• Did I search the invention as a product or method, a function and an implementation?
• Did I test meaningful terminology in the relevant languages?
• Did I read beyond titles and abstracts when a document appeared close?
• Did I inspect claims, drawings, citations, classifications and family information?
• Did I record the exact queries, database, date and unresolved limitations?
• Did I avoid treating “no result found” as proof of novelty?
• Do any close records, specialist search needs or timing questions require escalation?

Frequently asked questions

What if I find very few patents during the hour?

There is no minimum target. Test broader functional terms, older terminology, alternative component names and relevant classifications. Also check whether the technical field relies heavily on non-patent literature. Few results may reflect the query or database rather than the absence of prior art.

What if I find too many similar-looking patents?

Prioritize technical relevance and the relevant date, not whether the applicant is a competitor, university or large corporation. Use a distinctive implementation feature, CPC classification, claims field or date filter to narrow the result set. Keep the broader query in the record so the filtering choice remains visible.

What if I am not confident reading English patent documents?

Machine translation can support initial relevance screening. However, important claim language, technical relationships and dates should be verified in the original document or a reliable family publication. A translated summary should not carry the final legal conclusion.

Can this workflow handle biological sequences?

Not by text searching alone. NCBI BLAST compares nucleotide or protein sequences with sequence databases and calculates the statistical significance of matches.¹⁰ Sequence matters require appropriate databases, parameters and specialist interpretation.

Key takeaway

A DIY patentability search is valuable when it produces a transparent search record and better questions, not when it creates false certainty. The one-hour workflow moves through keyword preparation, broad and focused Google Patents searches, citation tracking, evidence comparison and documentation.

Escalate when you find a close single disclosure, face a combination or timing question, work in a specialist technical field or cannot explain the limits of the search. A positive first-pass signal means that further review is justified; it does not establish patentability.

Sources and verification

1. USPTO, Applying for Patents: Search for Patents and Prior Art. Accessed July 29, 2026.
2. Patsnap Eureka, IP Search. Accessed July 29, 2026.
3. Patsnap Eureka, IP Drafting. Accessed July 29, 2026.
4. Google Patents, Searching. Accessed July 29, 2026.
5. Google Patents, Coverage. Accessed July 29, 2026.
6. USPTO, Multi-Step U.S. Patent Search Strategy. Accessed July 29, 2026.
7. Google Patents, Result Viewer. Accessed July 29, 2026.
8. EPO, Guidelines for Examination, G-VI, 1: Novelty and State of the Art. Accessed July 29, 2026.
9. EPO, Guidelines for Examination, G-VII, 6: Combining Pieces of Prior Art. Accessed July 29, 2026.
10. NCBI, Basic Local Alignment Search Tool. Accessed July 29, 2026.

This article provides general information, not legal advice or a guarantee of search completeness. Patentability rules and prior-art effects vary by jurisdiction and matter. Verify the current official record and obtain qualified patent advice before filing, abandoning or materially changing an application strategy.