Data Driven Patents
Updated April 2021.
The patent system requires that we look ahead and guess the future. If we are right, there is a huge reward. This process can be powerfully addictive, as it is the same psychology as lottery tickets.
Patents are bets that the marketplace will adopt a technology. The interesting characteristic is that the bet must be placed today, and we will find out several years later if we were right.
We do not get the ability to steer or change our bet as new information pops up. We are stuck with whatever we had in our original patent filing.
If we did not think of it then (or more specifically, if we did not write it down then), we cannot fix it.
Essentially, we are guessing — prophesying — about the future. This feature is baked into the system but represents one of the biggest risks of a patent.
Prophetic Patents.
Lots of inventors – and even a few companies – do “prophetic patents.” For the most part, patents that come out of research and development departments of Big Companies are the incremental breakthroughs that engineers solve as they advance their product line.
Some companies do “forward patenting sessions.” These are pure brainstorming sessions where the “best and brightest” of the company look forward to guess or “prophesy” where the market is going. Many independent inventors do a similar thing, where they just write patent applications without actually building, testing, and marketing a product.
These patents were prophetic patents. Prophetic in the sense that they prophesied or guessed about technology that would become relevant in the future.
Here’s the rub about patents: once a patent application is filed, you cannot add to them. We are betting that what we invent today will have real value at some point in the future.
This is a key element: we are making a bet today, hoping that the idea will be valuable later.
These “forward patenting sessions” were purely mental exercises of making up what possibly could happen in the future. They were science fiction in the sense that these patents were technology that had not yet been invented or commercialized.
Here’s the key point:
Prophetic patents represent the highest risk possible.
They have high risk because there is little data to support the investment.
Patents at an early stage of a business are just a bunch of “great ideas” that are dutifully documented by a patent attorney and examined by a patent examiner. Once the examiner anoints the idea as a true invention, the invention is enshrined in the annals of history as being an idea that has never before existed. The inventor has created something that has never existed in the entire history of the universe.
However, valuable patents are not just inventions that are novel and non-obvious. Valuable patents are ones that have business value.
How does a patent have business value? It is only when someone is selling a product covered by the patent. Until a product is in the marketplace, patents have prospective or potential value, but not real value. In these types of situations, patents are like out-of-the-money call options. They *might* have value in the future, if everything goes our way.
The problem with prophetic patents is that there is a big investment that needs to be made before the patent has value. The investment is in research and development to build, test, and scale the product (technology risk) as well as an investment in marketing, advertising, and sales (market risk). Until that investment is made, the patent is worthless. Worse than that, the patent is a liability.
Prophetic patents are a liability in two senses. First, the prophetic patent requires capital to keep feeding the patent attorney to get the patent through the patent office and maintain continuation applications. Second, the time is running on the patent. It is not uncommon for a company to take 5 or 10 years to get some traction in the market. If a patent was filed 10 years ago and the product is just getting started in the market, the patent is half-expired.
I am an advocate of a different type of patent: a data-driven patent.
Data-driven patents.
We really want data-driven patents: we need to reduce our risk.
Every business proposition and every new product has two risks: technology risk and market risk.
Technology risk is whether the invention actually works. Does it function? Can it be built? Can it be built for an affordable cost?
Market risk is much more elusive but ultimately much more important. Will someone buy it?
A valuable patent is one that not only forecasts that the technology risk, but one that also guesses right about whether the invention will have commercial success. Ideally, we want as much data as possible before we invest.
The patent system has certain requirements that cause a tug of war between the prophetic nature of the patent and the data we wish we could have at the time we file it.
The patent timeline requires that you file a patent before disclosing it in public or offering the product for sale. In the US, we are given a one-year grace period, but in Europe and other countries, there is no leniency: you must file the patent before disclosing.
Your patent attorney will push you to file as soon as possible. “We want an early filing date” they will say.
Of course, the patent attorney makes money when you file a patent. And they make more money when you have to “fix” your patent application because of new material.
From a business standpoint, we want to wait as long as possible before filing – because we have more data to support the investment.
The pitfalls of prophetic patenting.
In general, patents that come out of “forward patenting” sessions like Intellectual Ventures are a net negative effect on both the company and society as a whole. The biggest negative effect is that they create prior art for a company’s projects that may take a decade of development before they hit the market. The prior art issues are non-trivial and can cripple a company’s later patents.
Creating science fiction patents without actually doing the heavy lifting to solve the engineering problems is also not benefiting society. Patents are a contract with society where you give up trade secrets in exchange for the protection of the patent claims. This forces competitors to design around your patent and leapfrog, rather than merely copy, your technology.
For the company, the forward patenting sessions generate lots of prior art. This prior art will be cited against them years later when the engineering resources are actually committed to work on the problem. Because of prior art created by their own prophetic patents, the company’s engineers will be forced to design around the ideas in those early patents.
Purely prophetic patents (the ones not coupled with ongoing research and development) are essentially bets that someone else will have the same idea as the inventor. Think about the risks of this type of bet. These patents are hoping that someone (eventually) will have the same insight as the inventor.
But the inventor is supposed to be is supposed to be brilliant and insightful – is this a bet that someone else is just as brilliant and insightful? Is this really a bet that the inventor is just putting together pieces in the same way as someone else in the future?
Put another way, these are bets that the invention is obvious, at least to someone. Perversely, forward patenting patents are bets that someone will have the same brilliant insight as the inventor.
One of the requirements for a patent is that it is not obvious. The examiner should reject the patent if it meets the legal definition of obvious. Also, is this type of patent really valuable? The real solutions come from solving the problems that come from the two-pronged heaving lifting of engineering and determining product-market fit. Is anyone really that clairvoyant to be able to anticipate all that? I have never met an entrepreneur or seen any company bring a product to market without a lot of iteration and learning on the way.
Prophetic patents require willful ignorance that we do not know what we do not know.
A data-driven patent is one where we have data to support an investment. The more data we have, the lower the risk.