🙂 His life history sounds lot more interesting than mine.
There are 3 groups of companies:
- pure IP play ones such as Intellectual Ventures. They treat patents purely as a financial asset with legal team as operations and engineering serving the legal team. The engineers are there to see what is in the market and explain the tech to lawyers. Lawyers then read that into the patents and go after those companies. There are some pure trolls that patent whatever comes into their mind and see if it flies in court. The others actually buy patents from companies, bundle them into different portfolios and license/litigate them. This makes sense to sellers of patents too because of the nature of their business, they can't or don't want to monetize their patents.
- R&D companies such as IBM, Bell Labs, etc with substantial R&D budgets where hundreds of Ph.D.s work on future stuff not tied to shipping products. For these companies, patents are an aggressive asset to carve out future roadmaps, leverage through standards, grow licensing revenue, etc. This is a legitimate business that benefits from the patent office and drives innovation. There are quite a few small companies too that follow this model.
- Other companies that file patents as non-core activity to encourage innovation during their product development. For them, these are largely defensive assets so that others dissuaded from filing patent infringement claims against their products. It's usually these companies that sell their patents to the first group because they keep a license to use/modify for themselves when they sell.
There is a lot behind patent filing process. The goal of a patent lawyer is to guarantee the company as much domain as possible for as long as possible with the strongest boundaries and most general claims. Once the patent is filed, the patent office responds after 1-2 years (there are ways to pay more to get earlier response) saying that this idea is already invented, the claims are too general, some claims are too different from others and should be filed separately, the method claims are too many, etc. The lawyers will delete, say, 15 of the original 50 claims, take out another 20 because they are different to submit as separate patent, etc. Now, the engineers are also working and they come up with more ideas on the same topic. The lawyers try to see if they can add these new ideas with the original filing date to get earlier invention date compared to what competitors could be doing. They find common language with the first patent and file continuations and divisionals. So, if the original idea is strong and the company is putting resources into that activity, it is easy to see it grow into a portfolio of patents through combined engineering and lawyering.
Stanford is great at that, with Silicon Valley being the proof. It is not easy to go from a patent to payoff. Patents cost money to file and maintain. An entity generating 100 patents/year probably has 1000-3000 live patents that cost 1M-3M/year to maintain. That is just paper patents. Actual R&D behind it probably cost multiple times that. And taking that R&D to products/revenue will cost multiple times more. To license, the licensee should see value in it because they have to invest R&D and engineering knowing that only a minority of ideas actually succeed. The people who know best are the students that worked on it, and the universities should enable them by starting incubators, etc. and being very generous with how "tightly" they hold onto the IP. This is the easiest way to see IP turn into products and whatever they lose by being generous in terms, they gain by these students living the startup lives and attracting venture funding.