When Not to File a Patent: Don't Pay to Smash Your Own Rice Bowl

What is the most common reason I tell clients not to file?

In my 20 years of practice, the reason I give most often has a blunt Chinese phrase behind it: paying money to smash your own rice bowl. The rice bowl is the technology that feeds the business. The money is the filing and prosecution fees. The smashing is what happens when you publish it.

The rule I apply is simple: technology that cannot be derived by reverse engineering should not be patented. It should be protected as a trade secret.

The reason is built into the patent bargain. In exchange for exclusive rights, the applicant must disclose the invention clearly and completely enough for a skilled person to carry it out. That requirement exists in every major system: 35 U.S.C. §112(a) in the US, Article 83 of the European Patent Convention, and Article 26(3) of China's Patent Law. And patent applications are published. So if your competitors could never learn the technology from your product, the patent is the document that teaches it to them.

How do you tell whether a technology can be reverse engineered?

The practical test is to imagine a capable competitor who buys your product and has a laboratory. Can they find out how it was made or why it works?

• Usually yes: shapes, structures, mechanisms, assemblies, circuits that can be traced. Once the product is on sale, these are effectively public. Secrecy cannot protect them, so a patent is the right tool.
• Often no: process conditions inside your plant, the way a material is prepared, the know-how that produces a property the finished product shows but does not explain. These are candidates for trade secret protection.

When the answer is unclear, I ask the client's own process engineers to play the competitor: given only the finished product and ordinary laboratory analysis, what could they work out? Their answer is usually more honest than the marketing team's, and it is the answer that matters.

There is a second reason to care about this line. If competitors cannot see the technology in your product, you usually cannot see it in theirs either. That makes infringement hard to prove. A patent you cannot enforce has still published your know-how.

My rule: if a competitor could not work out the technology by buying and analysing your product, a patent is the document that would teach it to them. Keep it as a trade secret instead.

What happens when a company files anyway?

Here is a real negative example from my practice. A materials company in Guangdong asked for my view. I advised them not to file. The company wanted to show investors, through patents, that it had real R&D capability, so it filed anyway.

Within two years, a large number of competitors studied its patent documents and made products with highly similar performance. The company's market share was heavily eroded.

Nothing went wrong legally in that story. The patent did exactly what patents do: it published the technology. What went wrong was the decision to publish technology that competitors had no other way to learn.

What if you need patents to show investors your R&D strength?

This is the pressure behind most of these decisions, and it is legitimate. Investors look at patents. But the conclusion I draw is different from the one the Guangdong company drew.

The real purpose of a patent is to protect technology. If the goal is only to show the outside world that you have R&D capability, you can choose peripheral technologies for patent filings: improvements that are real and worth protecting but do not give away the core know-how. Keep the core, the part that cannot be reverse engineered, as a trade secret. Investors see genuine filings; competitors do not get your recipe.

Proving R&D strength and protecting your core technology are different jobs. Use filings on peripheral technologies for the first, and trade secret protection for the core know-how that cannot be reverse engineered.

What does trade secret protection actually require?

A trade secret is only protected if you treat it as a secret. Most legal systems, including the US Defend Trade Secrets Act, the EU Trade Secrets Directive and China's Anti-Unfair Competition Law, require the information to have commercial value from being secret and the owner to take reasonable confidentiality measures. In practice that means:

• confidentiality agreements with employees, suppliers and partners;
• restricting who can access the critical process information;
• splitting knowledge so that no single outsider sees the full process;
• keeping records that show what the secret is and how it is protected.

Know the limits too. A trade secret does not stop a competitor who develops the same technology independently, and it does not stop legitimate reverse engineering. That is exactly why the reverse-engineering test above comes first.

This is not legal advice; consult a qualified patent attorney for your situation.

When is a patent still the better choice?

When the technology is visible in, or can be worked out from, the product. Then secrecy ends the day you start selling, and only a patent can stop copying. The decision is not "patents good" or "patents bad". It is a question about the specific technology: once it is in the market, can competitors see it? Answer that honestly before you spend money on a filing.