Part 3: The Keep - Your Critical First Patent Sprint

Name Russ Wilcox Partner at Pillar VC

Now that you have a foundation cleared up, it’s time to build your fortress, and at the center of every fortress is the most protected area of all, the Keep. We better explain our castle analogy, and then tell you step by step how to get started.

The Castle Analogy

Patents are like walls because they let you block competitors. So the plan for building a patent defense is exactly like the plan for building a castle defense. 

Look at the big building in the center of the picture below. This is the Keep (labeled 8). If invaders breach the outermost walls, you can fall back to this last reserve. In patent terms, the Keep is your first filings, which have the earliest priority date and directly cover your fundamental invention.

Let’s continue with the castle analogy. As your start-up gets older and develops its products, you will find new aspects of the invention worth patenting, perhaps covering new materials or applications, or alternate methods, or small improvements. Those later patents become the additional walls around your Keep. The more patents you file, the more walls you have to block the enemy. 

As you get further away from the core invention, there is more and more ground to cover around all the possible variations and applications, and since it could take a few years to sort all that out, those later patents may not be early enough to block competitors. So as the years go by, the incremental filings tend to be stretched thin and easier to breach, just like the outermost walls of a castle.

At E Ink, the key claim in our Keep was something like “an electronic display material operating by principle of electrophoresis with two different types of particles in clear microcapsules”. The patents we added later covered alternate principles (instead of two particles, use one particle and a dye), improved materials (brighter white particles), similar but different arrangements (sponge- or waffle-like cavities instead of round microcapsules), software (new drive waveforms), packaging (humidity barriers to protect the material), articles of manufacture (a display cell combining our film with a display backplane); and also lots of novel product applications, such as an idea for “a whiteboard that uses any kind of ePaper”. 

For a rival ePaper company, an E Ink patent on a “whiteboard using any kind of ePaper” would act like a distant outer wall. Alone it would only be a nuisance (by eliminating one particular market segment and reducing their addressable market size), and they could still compete in other segments. When combined with all the other outer-wall patents though, those nuisances can add up to a significant deterrent.

The walls

Your Crucial First Patent Sprint

When you first start your company, start by building the Keep! You do this by quickly filing a lot of patents that will have the earliest possible priority date. 

Didn’t the university patent already cover every possible idea? No!! That’s not really their game.

 The university’s goal is to file one fundamental patent, so that if anyone later develops a product, they can sue for a moderate royalty. In contrast, if your start-up’s goal is to sell an actual working product, there could be two dozen further micro-innovations required to commercialize and scale up, and each of those might become a separate future patent filed by someone who will want a royalty. You need to own ALL those micro-innovation patents too, or you will be unable to ship. If you are slower than your competitor to anticipate those ideas, they may hold you hostage. 

Again: Universities only have to cover one innovation necessary for the product, so they can sue; you have to cover every innovation necessary for the product, so you can sell without getting sued. 

The stakes also differ. When a university files a patent, for them it is one of many in a giant institution. For you, it is the basis of your whole company and key to the next chapter of your career. If you really believe you can build a unicorn company around this technology, you better believe that determined competitors will be ripping every sentence of the university patent apart a few years from now, perhaps sooner if you get a lot of attention and appear to be a winner. They will be seeking any edge where they can make a significant improvement on your idea that you will be forced to license, or any way they can go around the university claims by finding a parallel path that the first inventors did not anticipate, so they can bring out their own competitive product and eat your lunch. You need to batten down all those hatches NOW, before your competitors are even aware of what you are doing.

That is why the first step for a new deep-tech company should be an IP sprint.

This is a big commitment. Once the team has moved into the company offices, ask them to spend 50% of their time on IP for the first three months. They will be groaning now, but thanking you, years later. Your goal is to brainstorm, explore, test, and write – try to file 10, 20, or 30 provisional patents as the basis of your Keep. This can be less expensive than it may seem as we will explain below.

When brainstorming, think both upstream and downstream of your specific invention. Here is a checklist:

  • The raw material chemistries
  • How the raw materials are tested for incoming quality acceptance
  • How the raw materials are produced into parts you need for your invention
  • How the parts you need are tested for incoming quality acceptance
  • Your invention’s physical arrangement, function and composition
  • Alternate techniques and principles to accomplish the same results as your invention
  • How your invention is assembled and inspected for quality
  • How your invention is controlled and any algorithms
  • How your invention is packaged and shipped
  • What subassemblies or articles of sale may incorporate your invention
  • How your invention is incorporated into a finished product
  • All the finished products that may incorporate your invention 
  • How those products may be shipped, installed, and controlled downstream
  • Any software that may be needed to use the product that is relevant to your invention
  • Services and data that arise from the products, e.g. selling electronic books and articles
  • How those products and your invention may be disposed

Because ALL steps in a supply chain and all levels of the stack in a product architecture are required to sell the whole product to the end consumer and earn revenue, if a rival can obtain a chokehold at any ONE layer (e.g. a key patent on how to cost-effectively inspect the product) they can hold up your whole business. 

That is why it is so important to anticipate the entire path to market now. Brainstorm up and down the entire supply chain and product architecture. You can’t stop others from filing future patents on narrow ideas at the different levels, but you do want to make sure you have freedom to operate and to build your business over time. Disclose the whole scope of your vision and the implications of your invention in your earliest filings, and list every example you can think of at each level above, so competitors will have more trouble proving that their idea combined with your invention was non-obvious.

With each patent, bear in mind there are four types of utility patent claims: composition of matter, methods, articles of manufacture, and machines. A good patent will claim an invention in multiple aspects. For example, although the gold standard for a drug patent is the composition of matter (which can easily be enforced by analyzing a competitor’s drug to spot the claimed molecule); a strong drug patent will also claim methods such as a process for manufacturing the molecule, or the use of the drug for a certain indication such as treating cancer; articles of manufacture could be a pill, a liquid, or a cream containing the drug; and they may even claim a machine such as a specially-designed inhaler to administer the drug in aerosol form.


By patenting the invention in every aspect possible, you reduce the chance that someone will find a loophole or close variation that allows them to get around your patent.

While the team brainstorms, you can also come up with a boilerplate introductory text that can be used across all company filings. The template will cover the background to the technology that shows the need for an invention, recite past attempts prior and why they failed, and explain why your core technology is better. Do your homework and expand the prior art list to include every prior patent, article, speech, or other reference that you can find. Including all those citations in your prior art list now will ensure they cannot be used against you later. Go back and expand any lists in the patent language to be exhaustive – if you before you listed 3 possible plastics that could be used for a particle, do your research and figure out the other 40 plastics that could also be used; if before you listed 6 possible applications, add 60 more applications. All of this will become standard boilerplate that you add to the front of every one of your filings from now on, and this helps ensure that whatever patents you do get will be super strong.

Document each new idea from the brainstorm team with diagrams and explanation and examples, ranging from a few pages to a few tens of pages. For speed, clear hand-drawn diagrams are OK.

Take Full Advantage of Provisional Patents to Create Optionality

Send each invention disclosure to your patent agent or attorney, and ask him or her to staple the boilerplate on the front and then file it as a provisional patent. 

Major IP firms will fuss and complain that they want to do a proper job, and that it costs $5-7K to write a “real” patent. Even for a barebones provisional, they will want to draft claims and even under pressure they will want to charge you $1K. You are a cash-poor start-up, and if they help you file a lot of provisionals now that means a lot of business for them later. Try to set a fixed cap like $5-10K for them to draft the up-front boilerplate part with you once and then file the 20-30 provisionals all in a group. If they refuse, go find a patent agent with low overhead who can file these for you. Go back to the big IP firm a year later. Or find a more entrepreneurial IP firm.

Our advice is to ignore claims for now, and to negotiate this fee aggressively.

That is why if you are frugal, filing each extra provisional can cost you almost nothing, and it gets you the earliest possible priority date, plus you get up to twelve more months to go back and extend the patent with more text and experimental data. 

If you file 30 provisional patents during your first three months, then do nine more months of further science and technical development, you may well discover 10 of those patents were dead wrong (you can just abandon those and they expire unpublished), 15 are minor (you can combine those into 1 big application and just prosecute the best idea for now), and 5 turned out to be super important. Next turn those 5 provisionals into 5 full applications, strengthened by all the discoveries you have made during those nine extra months. 

By trimming the numbers with the advantage of an extra year of insight, you will end up paying a lot less in legal fees, and you can give the key filings more attention.

As we said in the intro, those early super-strong patents, combined with the university IP, are your Keep.


Once you have rushed to protect your most fundamental patents, you can turn to the primary focus of your company – enabling, developing, and launching new products and services.

You will want to make sure that whenever that effort yields valuable inventions, you have a system in place to identify, protect, and file on the right intellectual property. Each time you do, you will be building an additional outer wall to protect your core IP. Read about a comprehensive system for managing those processes in the next section.