A patentability search is a specific type of patent search designed to answer one question: can this invention be patented? It's the most common search inventors and startups need, and it's often the first step in the patent process.
Not all patent searches are the same. Here's how they differ:
Each serves a different purpose. Most inventors need a patentability search first. A freedom-to-operate search matters when you're about to launch a product — it answers whether you'll get sued, regardless of whether you have your own patent. A validity search comes into play when someone asserts a patent against you and you want to challenge it. And a landscape search is a strategic tool for understanding the broader competitive environment.
The distinction matters because each search type requires different scope and analysis. A patentability search focuses on finding prior art that predates your invention. A freedom-to-operate search focuses on currently active patents with claims that cover your product. Confusing the two can lead you to the wrong conclusion — finding no prior art for a patentability search doesn't mean you're free to operate, and vice versa.
Three criteria matter:
Your invention must be new — as defined by 35 U.S.C. § 102. If any single prior art reference describes every element of your invention, it's not novel. This is called "anticipation" — one reference anticipates your entire invention.
The novelty bar is strict but also narrow. A prior art reference must disclose every single element of your claimed invention to destroy novelty. If your invention has five key features and a prior art patent covers four of them but not the fifth, your invention is still novel. For example, if you've invented a solar panel cleaning drone that uses ultrasonic vibration to remove debris without water, and the closest prior art is a solar panel cleaning drone that uses pressurized water jets, your invention is novel because the specific cleaning mechanism is different.
Keep in mind that prior art for novelty purposes isn't limited to patents. Any public disclosure anywhere in the world counts — published papers, conference presentations, products on sale, YouTube videos, even social media posts. If someone posted a detailed description of your exact invention on a forum in 2018, that's a novelty-destroying reference. The date of the prior art matters too: it must predate your invention's effective filing date.
Even if no single reference matches your invention exactly, it might still be unpatentable under 35 U.S.C. § 103 if combining two or three references would make it obvious to someone skilled in your field. This is the harder standard to meet and where most patent applications fail.
The key question for non-obviousness is: would a person of ordinary skill in your technology area have been motivated to combine these prior art references? It's not enough that the individual elements exist separately — there must be a reason someone would think to put them together. Patent examiners often cite the "teaching, suggestion, or motivation" (TSM) test, though they have broad discretion.
Here's a practical example. Suppose you've invented a bike lock that uses both a fingerprint sensor and GPS tracking. If one prior art reference shows a fingerprint-activated lock (for doors) and another shows GPS tracking on bike locks, an examiner could argue it's obvious to combine them — the motivation is clear (better security for bikes). But if your specific implementation solves a technical problem that the combination creates (say, power management that allows both features to run on a small battery for six months), that solution to the new technical problem could make your invention non-obvious.
Secondary considerations can also support non-obviousness: commercial success of your invention, long-felt but unsolved need in the market, failure of others to solve the same problem, or skepticism from experts that your approach would work. These won't save a truly obvious invention, but they strengthen borderline cases.
Your invention must be useful. This is rarely an issue for practical inventions but can matter for theoretical or abstract concepts. The utility requirement has three components: the invention must have a specific, substantial, and credible utility. A new machine, chemical compound, or manufacturing process almost always meets this bar. Where utility gets challenged is with inventions that claim to do something implausible (a perpetual motion machine) or inventions with no identified practical application (a new chemical compound with no described use). For the vast majority of inventors reading this guide, utility won't be your bottleneck — novelty and non-obviousness are where most applications succeed or fail.
A thorough patentability search examines:
The search should cast a wide net. Patent examiners have access to all of this, and they will use it. You want to know what they'll find before they find it.
Non-patent literature (NPL) deserves special attention because it's where many inventors get surprised. NPL includes academic journal articles, PhD theses and dissertations, conference proceedings (like IEEE, ACM, or medical conferences), technical standards documents, product catalogs and user manuals, white papers, and even archived web pages. Patent examiners regularly cite NPL, especially in software, biotechnology, and academic-adjacent fields. Google Scholar is a good starting point for NPL searches, but domain-specific databases like PubMed (biomedical), IEEE Xplore (electrical/computer engineering), or arXiv (physics and computer science) will surface references that general searches miss.
Published patent applications are another commonly overlooked source. In the US, most patent applications are published 18 months after filing, even if they haven't been granted yet. These published applications count as prior art from their filing date. This means an application filed two years ago but still pending can block your patent — even though it's not a granted patent.
A patentability search gives you three possible outcomes:
Each outcome saves you money. The first gives you confidence to invest. The second helps you write stronger claims. The third prevents a costly mistake.
If you get a clear path, move forward with confidence but don't skip the attorney. A clear search result means you have strong evidence of novelty, which makes the patent prosecution process smoother. Share your search results with your patent attorney — it gives them a head start and can reduce their billable hours significantly. File a provisional application quickly to establish your priority date, then work with your attorney on the full application.
If you find overlap, the search becomes even more valuable. Map exactly which elements of your invention are covered by existing art and which aren't. This gap analysis becomes the foundation for your claim strategy. You may need to narrow your independent claims to focus on what's genuinely new, but that doesn't mean the patent isn't valuable — a well-drafted narrow patent that clearly distinguishes from prior art is far more defensible than a broad patent that an examiner fights you on for three years. Consider whether design-around opportunities exist in the overlapping areas.
If you're blocked, you have several options beyond abandoning the invention. First, verify the blocking reference actually covers every element of your claim — read the claims carefully, not just the abstract. Second, consider whether you can modify your invention to avoid the prior art. Third, check whether the blocking patent is still in force (patents expire after 20 years, and many are abandoned earlier for non-payment of maintenance fees). Finally, consider whether a design patent might protect the ornamental aspects of your invention even if the utility aspects aren't patentable.
Before you talk to a patent attorney. Before you file a provisional application. Before you discuss your invention publicly (which starts a 12-month grace period in the US). The search should be one of your first steps, not an afterthought.
Timing matters more than most inventors realize. The US patent system is first-to-file, meaning the first person to file an application gets priority, regardless of who invented first. Every day you delay is a day someone else might file a similar application. But filing without searching is equally risky — you could spend thousands on an application that's doomed from the start. The sweet spot is to search first, then file quickly once you have confidence in your position.
If your invention is related to your work for an employer, check your employment agreement before searching or filing. Many contracts include invention assignment clauses that give your employer rights to inventions created on company time or using company resources. Sorting out ownership before you invest in a patent search avoids complications later.
Related Posts
A step-by-step guide to searching patents effectively, whether you're an inventor validating an idea or a professional conducting prior art analysis.
Learn how to read and understand patent claims — the most important part of any patent document and the key to evaluating patentability.
How to conduct a patent landscape analysis to understand competitive positioning, identify white space, and make informed R&D decisions.