Patent Descriptions: Every Word Hurts You

Every Word in a Patent Application Hurts You

The patent application has two parts: the specification and the claims.

Every word in the specification is a trade secret you give away to your competitors.

Every word in the claims narrows the claim and limits your enforceable rights.

Long Specifications Hurt You

Many clients are told by their patent attorneys to put as much information as possible in a patent application.

This is a terrific strategy – for the patent attorney.

The theory is that the patent attorney can “mine” the specification for lots of inventions.  Once one patent issued, we can do a continuation application and file for another invention.

This Used to be True.

Before 1995, Jerome Lemelson became famous for this strategy.  He pioneered the idea of a “submarine” patent, where they would file continuation applications one after the other.  This strategy was brilliant because the later patents were directed at specific infringers and use cases that were not apparent at the time.

In the 1990’s, it was standard practice to put as much information as humanly possible into a patent application.  One of my friends who was a patent attorney at Hewlett Packard back in the 1990’s referred to their Calculator Patent which was well over 1200 pages long.  This was before the advent of PDF documents, and the Calculator Patent would be passed from one patent attorney to the next as a six-inch stack of papers.

The best part about this strategy: the patent attorneys had work for a lifetime, or until the money ran out. Patent attorneys put kids through college on this strategy.

But the Laws Changed.

In 1995, the laws changed.  Before 1995, patents were held in secret at the Patent Office and expired 17 years from the date of issue.  After 1995, patents were published at 18 months, meaning they lose their secrecy.  But the key thing: every patent in the patent family expired no later than 20 years from the earliest filing date.

This law dramatically changed the calculus for patents, but nobody informed the patent attorneys.

For the last 25 years, every patent expires 20 years from the earliest filing date.

The Myth of the “Early Filing Date.”

There is a mantra among patent holders – pushed by the patent attorneys – that an “early filing date” is good. This myth is only partially true.

In my business of loaning against patents, I see lots of portfolios that have some age to them.  Some have been pending for 10, 15 years or even longer.  Because the law changed in 1995, patents will issue out of these patent families with only 5 years before they must expire.

The notion that you should put everything in a patent application used to be a good strategy for a patent owner.  But for the last 25 years, it is only a good strategy for the patent attorney.

From an Advantage to a Boat Anchor.

If you put too much information in your description, the huge specification is a gold mine for a couple years, then turns into a boat anchor. It is surprising how long it takes for a technology to really become useful. It may be a decade or longer.

I see countless startup companies who have been slogging it out in the trenches and finally get a bit of traction. They often take a decade and 3-4 rounds of funding to get there. They show up with a patent portfolio that started with a giant description, and thankfully, they kept the application open with endless continuation applications.

While their patent attorney was lining their pockets, the clock was ticking on the startup. Because they disclosed “all” their ideas early, the only patents that come from that disclosure must expire 20 years from that early filing date.

Now the “advantage” turns into a boat anchor.

Our entrepreneur is strapped with a whole set of patents that are already halfway expired before they even issue. I have seen loan candidates with a pending set of continuations with 15+ years of pendency, meaning that their next patent will issue with less than 5 years of patent life.

What is the Best Strategy Now?

I am a big fan of pacing the patents to match a company’s growth.

The first patent is always the worst one – because we have little data at the beginning.  As a company grows and understands their customer base, they will start to uncover the real customer needs. 

The only meaningful data to support a patent application is whether the customer buys the product.  Everything else is day dreaming and speculation.

I prefer targeted patents that capture the reason why a customer bought the product, at the same time, I prefer patents that reset the 20 year window of the patent life. By resetting that 20 year window, we keep enough protection available for an acquirer – or for going after a competitor.