Why the Best Prior-Art is Often Not in a Patent Database?
Ask most people where to look before filing a patent, and they’ll say the same thing: search the patent databases. It makes sense. Patents are where inventions are written down, numbered, and neatly filed.
But the idea that patent databases hold the best evidence of what is already known is one of the most expensive assumptions in the world of innovation. Some of the strongest evidence has never been inside a patent office at all.
First, what is prior art?
Prior art is anything that was public before a certain date and shows that an idea was already known. If you file a patent on something that was already out there, your patent can be rejected. Worse, it can be knocked down after it’s granted, sometimes years later, when you have already spent real money on it.
Here is the part many people miss. Prior art is not limited to patents. It can be a magazine article, a product on a shelf, a YouTube video, a student thesis, a conference talk, a manual, an old textbook, or a post on a forum. If it was available to the public, it can count.
And most of the world’s knowledge was never patented.
Why patent databases are not enough
Think about how much of what humans know ends up in a patent. A tiny slice. Patents cost money and time, so people only file them for things they think will pay off. Plenty of companies, engineers, farmers, hobbyists, and researchers solve problems every day and never file anything. Some don’t care about patents. Some can’t afford them. Some don’t even know that what they’ve done is new.
So if you only search patent databases, you are searching a small room and assuming it is the whole house.
There’s another problem. Patents are written in a special style. Lawyers draft them to be broad and sometimes confusing. A patent might describe a “fluid communication apparatus” when everyone in the real world calls it a pipe. Even a good search can miss things because the words don’t match.
A story that shows how this plays out: Turmeric
In the mid 1990s, a US patent was granted to two researchers at a university medical center for using turmeric powder to help heal wounds. To a patent office, it looked like a fresh idea.
To millions of people in India, it looked ridiculous. Turmeric has been used on cuts and wounds in Indian households for generations. Grandmothers did it. It’s in old traditional texts.
India’s Council of Scientific and Industrial Research challenged the patent. The evidence they brought was not a patent. It was old Sanskrit writings and published traditional knowledge showing that turmeric had been used this way for a very long time. The US patent office cancelled the patent in 1997.
Nobody could have found that evidence in a patent database, because it was never there. It lived in books, in homes, and in culture.
Another one: Neem
Neem is a tree in India. People have used its extracts for centuries, including as a natural pesticide and fungicide. In the 1990s, the European Patent Office granted a patent connected to using neem oil to control fungi on plants. The patent was held by a US company and the US Department of Agriculture.
It was challenged by a group of opponents, and the strongest argument was that the use of neem this way was already known and practiced in India. The patent was revoked in 2000, and the revocation was upheld on appeal in 2005.
Again, the winning material was not a pile of earlier patents. It was knowledge that had been out in the open, in daily use, for a very long time.
Turning a problem into a system
These cases were part of the reason India built something called the Traditional Knowledge Digital Library, set up in the early 2000s. It takes ancient texts, often in languages like Sanskrit, Urdu, and Tamil, and puts them into a searchable form that patent examiners can read. The idea was simple: if the evidence was invisible to patent offices, make it visible.
That is a helpful lesson for anyone, not just governments. Prior art exists whether or not a search tool can find it. The question is whether you looked in the right places.
Where the good stuff actually hides
Here are some places where strong prior art often sits, far from any patent database.
Products that were already being sold. If someone was selling your invention before you filed, that can destroy your patent rights. In the US, a well known Supreme Court case called Pfaff v. Wells Electronics made this clear in 1998. An inventor had taken orders for a computer chip socket before he filed his patent. The Court said the invention was “on sale” and the patent was invalid. A later case, Helsinn v. Teva in 2019, went even further and held that a sale can count even when the buyer is kept confidential.
For you, this means a product catalog, an old invoice, or a sales brochure can sink a patent. None of those live in a patent database.
Student theses and dissertations. Many people are surprised by this one. In a case called In re Hall, a doctoral thesis sitting in a German university library was treated as public prior art. Why? Because it was catalogued and anyone interested could have found it. Think about how many universities have thousands of theses that nobody outside the department has ever read. They can still count.
Manuals, datasheets, and user guides. Engineers write detailed technical documents that explain exactly how a product works. These are often more detailed than a patent, because they have to be accurate enough for real people to use. If a manual from 2009 describes the same feature you are trying to patent, you have a problem.
Conference talks and slides. Researchers present half-finished ideas at conferences all the time. Slides and recordings often stay online. A talk given a year before a patent filing can be enough to invalidate it.
Open source code and online communities. A huge amount of software knowledge lives on code sharing sites, forums, and Q&A pages. A developer might post a clever solution in 2014, and nobody thinks of it as an “invention.” But it is dated, public, and searchable. In software, this is often where the best prior art sits.
Old magazines and trade publications. Hobbyist and trade magazines are a goldmine. A woodworking magazine from the 1980s or a farming journal from the 1970s might describe the exact trick someone is trying to patent today. These are frequently not well indexed online, which is exactly why they get missed.
Videos and social media. A short clip showing someone building or using a device, posted years ago, is a dated public record. Tools like the Internet Archive’s Wayback Machine can help prove when a web page looked a certain way.
Foreign language material. A paper or a product from Japan, Germany, China, Brazil, or Korea may describe your idea long before anyone in your country thought of it. If your search was only in English, you missed a lot of the world.
Why this matters even if you’re not a lawyer
You might think this is only a concern for patent attorneys. It’s not. It matters to different people for different reasons.
If you are an inventor or a startup founder, doing a thin search can give you false confidence. You pay for a patent, build your plans around it, and then a competitor finds a ten-year old manual that kills it. Better to find out early, when the cost of changing direction is small.
If you are a company facing a patent lawsuit, strong prior art can be the difference between paying a large settlement and walking away. Many defendants win not by arguing about technical details but by finding one old document that proves the idea was already known.
If you work in research or product development, it is worth remembering that your own team’s old documents, internal reports that were later published, or past product launches can be prior art against anyone, including you.
If you are an investor, remember that a granted patent is not a guarantee. A patent that has never been seriously challenged has never been tested. The real test is whether someone with a motive to look hard can find something that wasn’t in the database.
How to search better
You don’t need to be an expert to improve how you look. A few habits go a long way.
Search the way normal people talk. Don’t use only the formal language from patents. Try the simple words that real users would type, and the slang that people in the field use.
Think about who else might have solved this problem. Different industries often solve the same problem in different ways. A trick used in car manufacturing might already exist in medical devices. Look sideways, not just straight ahead.
Look at old and obscure sources. University libraries, archived websites, trade journals, and old catalogs matter. Not everything is a click away, and that’s why it’s valuable.
Search in other languages. Even a basic translation tool can open up a lot of material that most people never see.
Ask people. Talk to veterans of the field, retired engineers, long time customers, and distributors. Someone who has been in an industry for thirty years often remembers things that no database records. A single phone call can beat a week of searching.
Check dates carefully. Prior art only counts if it was public before your filing date, so it’s worth confirming when something was really available. Archive tools, publication records, and invoices can all help.
Work with a professional for high stakes cases. A good patent searcher or attorney knows where to look and how to build a case with the evidence. It’s an investment, and when a lot of money is on the line, it’s usually a smart one.
The bigger lesson
Patent databases are useful. They are organized, searchable, and a necessary starting point. Nobody is saying you should ignore them.
But they only show what people decided to patent. The real history of ideas is much messier. It lives in grandmother’s kitchens, farmers’ fields, dusty library shelves, forgotten forum threads, old brochures, and conference rooms. The turmeric and neem cases show how powerful that kind of evidence can be. They also show that a big, official looking patent can crumble in front of something as simple as a very old book.
So the next time someone says, “I searched the patent databases and found nothing,” treat it as the beginning of the search, not the end. The best prior art is often sitting somewhere nobody thought to look.
This article is for general information and is not legal advice. For decisions about filing or challenging a patent, speak to a qualified patent professional.
Comments