part 2

The Foundation

(before you incorporate)

In this section, we discuss how to evaluate a technology or science available to license, how to figure out if there is room to start a new company without getting sued by existing patentholders, and how to make a fair deal with a university or government lab licensing officer. 

There are three key steps to take before you launch a deep-tech company:

⚙️
Find the best core technology

🆓
Check freedom to operate

🎯
License the IP for your startup

get to the core technology

Find the
Best Core Technology

Where do you look for deep-tech IP suitable for a start-up? Begin with research universities and government research labs. Licensing offices at these institutions have way more ideas available for license than they have CEOs and investors ready to take them forward. If you are a qualified founder, you can expect a positive reception and introductions to the inventors of any technology that catches your eye. 

Some people believe the best way to start a deep-tech company is “technology-pull”. First identify a problem, then search worldwide for a breakthrough you can license. The wisdom of this approach for the entrepreneur is that you are commercially focused from the outset, and you know who will buy.

In practice, most deeptech companies we know started with “technology-push”. A scientist believes he or she has invented something valuable and decides to start a company and becomes a fierce advocate. The inventor either leaves the lab to become the full-time founder and CEO, or finds an entrepreneur to lead the business. 

Most deep-tech companies
we know
started with “technology-push”.

From the viewpoint of the business cofounder, the downside to technology-push is that the start-up is committed to its founder’s technology, which may or may not work as hoped, and the company may find itself holding “a solution in search of a problem”. The upside is that the inventor is usually a technical genius who is deeply committed to the vision. This proves a crucial asset, because for most deeptech start-ups the vast majority of technical work still lies ahead, and more inventing will be required. So we do often back this approach, because it begins with strong people, and people are everything.

If you start by meeting an academic inventor, it is vital to conduct substantial due diligence. Make sure to consult other experts in the field, because inventors rarely poke holes in their own ideas. 

Also consult the university licensing office. The licensing officer is a professional who has been watching the technology evolve over time. This person can confirm whether the university has filed patents and whether the rights are still available. He or she may also have a good instinct for whether the technology is suitable for a start-up. 

Set up multiple meetings with the academic founder to test your working rapport and go together to see potential users or industry participants to learn how great the invention really is and whether it solves an actual problem. 

Remember that cutting-edge technology that works inside an academic’s lab under controlled conditions quite often does NOT work in the real world.

In the biology space, something like 50% of all discoveries do not actually replicate in a commercial setting. Therefore, try to obtain third-party verification before you license if possible.

If an academic inventor says that the university “does not have any rights”, you need to confirm that and the academic should ask the licensing office to provide a written disclaimer letter. You will need this to show investors that there is no IP ownership dispute with the university.

Check
Freedom to Operate

Before you start a new deep-tech company, it is wise to check for patents from third parties who could use them to block the product you intend to build. If there are none, then you have “freedom to operate”.

Do Your Own Search at First. While you can pay a law firm e.g. $30,000 to do this work, before you raise capital you do not have cash. You will also learn a lot by making an initial patent search yourself. 

Allocate 10 hours of your own time for this before you raise Seed capital. After you raise Seed capital, go ahead and hire an IP firm to do it for you better, and ask them to set up monitoring of key terms.

How do you search? Start at the U.S. patent and trademark office search site, and begin by looking at their issued patent database. 

Search for relevant keywords. Carefully read the claims of any patents you find. 

Then, ask your technical inventor for the names of a few people who are the closest competitors, and search on their names to see what they have filed for patents. 

When you are done, return to the USPTO site and look at the patent application database. These are applications that are not yet granted by the PTO but if they are more than 18 months old, they are “laid open” to the public and you can read them. Note that the claims here are written by the applicant and not yet approved by the PTO, and they usually ask for way more than what the PTO will allow, so don’t panic when you see applications asking for very broad claims.

Now repeat the above and look at foreign patents and applications. You can find them at the World Intellectual Property Organization (WIPO) site.

In the best case, you learn that no one else has a claim that already covers the technology you want to license, the product you want to build, or the application you want to target. You are hoping that if the core technology is an evolution of past work, that the past work was either never patented, or was disclosed by old patents that are now expired and thus do not represent a threat.

How to Handle Prior Patents. What if you discover that other companies hold patent claims that seem to apply to the product you have in mind for your start-up? All is not lost! There are a couple of options:

1. Re-design your product

Carefully read their claims and then re-design your product so you do not infringe. This can work well if the patent is issued and the claims are narrowly worded. Be careful though, if the other party is still prosecuting its patent, or files a continuation of its original patent, they might still be able to amend their claims to catch workarounds.

2. Find the problem patent potentially invalid

Discover if the problem patent is potentially invalid. The best way to do that is to find “prior art” (a public disclosure that predates the patent filing) because that helps prove their invention was not novel. Evidence of prior art may come from an earlier patent, or from a journal, presentation, or poster session. You do need to find prior art that was not already cited by the other company’s patent filing. If a reference was cited as prior art in an issued patent, then it is presumed that the examiner knew what he or she was doing when granting the claims and the claims are usually upheld as novel. If you do find new prior art, also set that aside to cite in your own patent filings, so it cannot be used against your patents.

3. Take a license?

A key question is whether to approach the third party now, or wait until later? It is tempting to hope they do not notice you and just wait until years in the future when you receive a nasty letter, but remember at that time you will be selling a product and they will have you in a vulnerable position and can sue you for damages. Another key issue you need to consider here is that whenever you raise investment capital, the standard investment documents (including SAFE notes) require you to state whether you are aware that the company’s planned products or services will infringe any third-party IP. So if your plan is to wait, then you still have to disclose the issue and then convince all your future investors that you will be able to solve the problem later. That’s a big thorn in your side. So our usual advice is to approach the other parties promptly. After all, before you have incorporated a company, you have nothing to lose! You have no cash for them to try to claim. And if the product you have in mind cannot survive a legal challenge, and you cannot afford the license terms, you need to know that before you invest years of your life.

4. Wait

The patents with the broadest claims tend to be the oldest patents. Patents expire and so you may be able to outwait them. One phenomenon we have noticed is that tech innovation often runs in cycles where ideas that were popular 20 years ago suddenly resurface. That is because avenues that appeared closed off by older patents suddenly become open to use freely.

5. Avoid the geography

When E Ink was starting out, we were worried about a patent that was issued only in Japan. So, we restricted our first products for sale only outside Japan. You will find major enterprise customers will be put off by this because they need products and services that they can adopt and deploy worldwide. However in our case it was fine, because we were starting with pilots in the USA, and we knew that within 2-3 years we would be switching to a next generation of our platform that was different enough to avoid the issue after that.

6. Partner Up

This is not an ideal approach, but it is worth noting that some major companies negotiate “IP peace treaties” with each other. So if you are worried about a patent held by one corporate giant, it may be the case that you can find a corporate rival who is able to practice. You can then safely outsource that step or component. For example, E Ink displays require a thin-film transistor (TFT) backplane to show high resolution graphics, and the IP landscape for TFTs is a complicated mess. However the TFT industry has negotiated enough general truces that we could access backplanes from any one player, without worrying that another player would sue us.

If you find yourself suspecting your new company will start off infringing a third party’s patent, it is well worth reviewing your thoughts with an IP attorney before you incorporate. If you can point at the filings of interest, they will quickly grasp the situation. If you have no money, then it is still worth asking, as an attorney may give you a few minutes of more general advice now, in order to win your business later, or they may be willing to delay billing you until after you raise cash.

As you search, you should never put into writing a negative legal conclusion like “uhoh, looks like we infringe this patent.” You don’t want your first panicky email to haunt you in litigation because if the plaintiff can show you knowingly infringed their patent, they can sue for triple damages. Remember, you are not a lawyer, and claims language can be highly confusing, so you may find yourself changing your opinion later! So no matter how alarmed you feel, keep your conclusions to yourself, and only say “this patent is of current interest” when talking with anyone except your attorney. 

Clear Foundation?
Now License the Core IP

After you take an interest in a particular invention, it could take you a period of weeks or months to decide whether the tech is marketable and has freedom to operate. If so, before you invest your time, ask the licensing office for an Option agreement. They will often give you an exclusive period of 3-12 months to develop your business plan, and assure you that you will have the right to take a license on commercially reasonable terms. You may also be able to pre-agree on the major terms such as royalty rate. An option letter can cover the basics in a few pages, which means you can delay the greater effort of negotiating a full license (often 40-80 pages long) until after you have completed your diligence.

Negotiating the commercial terms of a license is more art than science, because every IP is unique. The two most critical terms are the Scope, which covers the types of products you can make, and the Fees, especially the royalty, which becomes a tax that your future start-up will have to pay for many years to come.

Terms — Scope

What are you licensing, and what does the license enable you to do?

The license “covered IP” ought to cover everything that the inventor’s lab has already created that is needed to practice the technology you intend to commercialize. If the university will continue to develop improvements of the technology, then you can also ask that the license include additional work in the same area or from the same lab or by the same inventors, for a transitional period of 1-3 years. This is called the forward IP or the improvement IP.

You want to make sure exploit is defined to cover activities up and down the value chain such as: research, develop, make, have made, use, have used, sell, offer for sale, have sold, modify, enhance, improve, import, export or commercialize. The point of listing all these activities is to make sure your start-up will be able to outsource as needed to maximize commercial value, without any of your vendors, partners, or customers being asked to pay added royalty. 

A license can be limited in how you are able to “exploit” the intellectual property.

Ideally you want your license “field” to cover every application for the technology, so that your company has numerous ways to make money. For example, if the patent is on a new type of paint, you want to be able to paint any color and any object.

You do need to nail down the field definition early, because any limitation could greatly affect your business plans. When it comes to the field, the university has mixed feelings. They need your company to succeed because they only earn royalties when you sell product. So they want you to have enough opportunity to be a healthy business. But they would rather not license you on a field that you never intend to exploit. Let’s say you build boats and you want to license a slippery paint that resists barnacles. The university may propose to license you for maritime craft, but reserve the application of painting docks to license to someone else. That way they enjoy two chances at revenue.

It’s up to you to get the broadest field you think you could possibly exploit.

Of course, you would like the license to be “exclusive” within your own field, so that you will not have competitors using the same technology.

The term of a license is usually until the end of the patent life, although if trade secrets are involved then the license may run indefinitely. However the university often reserves the right to revoke the license if your start-up seems to be failing at commercializing it. So there are often performance requirements to maintain exclusivity, which could include raising a minimum amount of capital, and reaching a first sale by a certain deadline. 

Terms — Royalties

The most important license fee is a royalty. The royalty is usually a fixed payment plus a running royalty calculated as a percent of sales. It can also be a fixed amount per unit or a percent of gross margin.

Using a royalty approach means the university is only paid after the start-up commercializes the product. So in the early years, you are in the same boat together, and the start-up does not have to pay much up front. However because the royalty is calculated from revenue (“off the top line”), once the business starts operating then the university does not have to worry about whether the management team is running the business to earn a profit. The university also receives no control rights in the business, so this is very much an arms-length relationship.

Every percent of royalty matters a lot.

A good way to think about the proper royalty percentage of sales is to compare the royalty to your profit-before-tax margin. A 1% royalty sounds small, but if you are in a commodity industry that usually earns 5% profit before tax, then the patent is costing you 20% of your profits. That means the royalty holder gets as much income as an investor with a 20% equity stake! In fact their position is far more valuable than that, because they get paid off the top – they get their royalty even in lean years when the business does not earn any profit. When you compare the royalty to your expected profit margin, you will see that every percent of royalty matters a lot.

The most compelling principle I have heard about how to establish a “fair” royalty is the “one quarter of profits rule”. The idea is that if you license a complete design for a product for your business, then the IP ought to be worth about 25% of the profits. The other 75% of profits ought to reward the capital and talent that operates the business. In the example above, that means you would want to negotiate a royalty of 1.25% of sales. 

But when we talk about deep-tech university licensing, you must bear in mind that the university IP is often incomplete.  You and your investors may have to invest millions to continue inventing and developing until there is a full product. And if you infringe a third party’s IP, you may have to give them a slice as well. So in an industry that earns 5% profit-before-tax, if serious hurdles remain for a technology to become viable at commercial scale, then a royalty of 0.25-0.5% for the university may be completely appropriate.

This “percent complete” idea is particularly crucial in biotech. A lab that identifies a novel target for a new drug usually cannot license the target for any royalty, because there are so many ideas for targets published in the literature. If they can find a chemical that hits the target and shows an exciting effect in vitro (cells in a dish), they might ask a 1% royalty. If shown to work in vivo (mice), perhaps a 2% royalty. If brought forward a few years with lots of promising supporting preclinical data and promising toxicology, perhaps a 3-5% royalty, plus a lump sum milestone payment if the drug later enters clinical trial. If the drug makes it through Phase I and II clinical trials, it may command a 10-15% royalty plus milestone payments in the hundreds of millions of dollars. 

Overall, one study states that the median university license in recent years across several universities including MIT and Harvard is 3%. Bear in mind however that this is skewed because “the three most active fields for licensing activity are Pharmaceuticals, Biotechnology, and Healthcare Products and Supplies” which tend to have high margins.

Other typical royalty rates vary by industry. In hardware, universities typically charge 1-3%. Commodity materials might be 1-2%, while exotic materials might justify a royalty of 3-5%, because the company will likely sell them in small quantities at high margins.

In software, universities would feel lucky to get 2%, even though software margins can be high. Why so low? Because nowadays software developers can just put their mostly-finished code into the public domain and then the university gets nothing. They can leave the university and then finish the software later. There are also no patents and no easy way to enforce an infringement. So universities cannot get too greedy here.

The royalty can be set at different levels depending on the basis of the revenue. For example, there might be a lower royalty for printer equipment revenue, and a higher royalty for printer ink. This is reasonable because the expected profitability on equipment is lower than on ink.

It is common to adjust the revenues to exclude small sales from samples, non-recurring development contracts, pass-through cost of precious materials (like gold or platinum), unpaid invoices, and payments for ancillary services such as consulting, installation, design, testing, maintenance and support.

Usually a university license will also give your start-up the right to grant sublicenses to other parties. Perhaps you have a new method for making diet soda and want to give one sublicense to Coca-Cola and another to Pepsi, for example. That means you could earn some “sublicense income” from charging your own royalties instead of product revenue. Typically a university will want to receive 25% of your sublicense income. This is consistent with the 25% share-of-profits rule explained above.

A university license might also give your start-up the right to grant sublicenses

If your IP search has discovered potential prior art that you would have to license in addition to the university IP, then you can negotiate an “anti-stacking” provision to cover a situation where the company has to pay royalties to multiple parties. For example, the university to discount its royalty by half the third-party royalty, up to a maximum of 50% discount.

Having agreed on a royalty rate, the university will typically also demand a fixed annual payment, such as $10-25K (often creditable against each year’s running royalties). This covers their administrative costs and also helps them make sure that your business remains active.

One negotiating point to consider is that as time goes by, your company will likely contribute more and more of its own IP, and that growing investments will be required from your investors to reach high volumes. So it is sometimes agreed that the royalty and sublicense percentage will decline over time, or after you have hit a certain volume level.

Terms — Royalties

There are several other types of fees that universities may seek, in addition to royalty.

Usually, the university will demand that your start-up reimburse them for all past and future patent and legal fees. That can amount to a $20-200K bill at closing, unless you negotiate a delayed payment time. You will also have to pick up the future patent expenses. In return, they will usually allow you to participate in key decisions about how to prosecute the patents.

There can be a request for milestone payments, such as a large fee when a drug begins clinical trials, and another fee when a drug reaches the market.

🥧

We increasingly see universities asking for equity in start-ups.

In our view this is not reasonable if you are paying a full royalty rate, but it can be reasonable if the university agrees to discount the royalty rate, delay the start of royalties, waive some of the prior patent costs, or offer you an option to buy them out for a fixed fee later. We often see universities asking for a few hundred thousand dollars worth of equity, amounting to a few percentage points after the seed round closes for a venture-backed start-up. Our advice is to then ask for a lower royalty rate or a royalty credit to offset their ask.

Finally, universities with large endowments are increasingly asking for a right to invest in future rounds. This is a mild negative for the company, because usually a start-up will want to preserve its ability to decide who can invest. However if the right is kept to single digit percentages, such as the right to buy up to 5% of the Seed or Series A round, then it will not be too disruptive. 

Make it a Win-Win Relationship

As you can see, a license can be complex. The typical university license can take 3-6 months to negotiate and run 40-80 pages long. However, the end goal is not just the contract, the end goal is to develop a strong working relationship.

A good university licensing officer will typically begin a license negotiation by asking the entrepreneur to provide a business plan. The licensing officer can then adjust the scope to cover the company’s intent, and choose a royalty basis to be appropriate for the target industry and the type of revenue. Seeing a good business plan also helps assure that the entrepreneur is competent and can be trusted to make a strong attempt to commercialize, thus helping to justify your request for an exclusive license. Experienced licensing officers have seen many start-ups and will even offer suggestions and introductions to potential investors and corporate partners and customers.

Because you will be working together for many years, it is critical that you use a professional and ideally collaborative negotiating style with your university licensing officer.

A number of issues can come up in the future that require cooperation with the license office:

You will often discuss patent prosecution.

New opportunities may emerge that you did not anticipate.

The university develops an improvement you want to fold into the license.

You may need to resolve who owns a future invention.

Additionally, if the product infringes IP of many parties, then you may need to renegotiate the anti-stacking. 

The good news is that the university earns its income from royalties, so they do not get paid unless you succeed. Also, if the university holds equity or a right to invest, that is further incentive to help the company succeed. 

License or Seed Round,
Which Comes First?

Should you launch the company before the license is signed? As mentioned, the license can take 3-6 months to complete. You may be able to attract investors, open an office, and start product development well ahead of that time. While the technology team remains inside a university, any new inventions are owned by the university, rather than the company. The sooner you raise money, the sooner they can leave.

And after they invest, your negotiating leverage with the university office immediately drops. You can no longer just walk away. It’s also hard to ask for a lower upfront fee if they know you just raised millions of dollars! So it can be to both the founder and the investor’s advantage to refuse to proceed with a round until the license is agreed.

Investors quite reasonably do not want to fund a company before it holds a license.

We have seen the timing dilemma resolved in several reasonable ways:

The company signs a term sheet with the investors that is contingent on the license being completed. The company and university then work rapidly to negotiate the full license and the investment closes immediately after. This compresses the normal 3-6 months into ~6-8 weeks and a signed deal in hand means no investor risk. Moving fast may not yield the best terms though, because the university can afford to be more patient than the founders and investors, who are obviously eager to begin.

The company signs first a term sheet with investors, then a second term sheet with the university. The investors proceed to close the funding in two tranches; a smaller amount to cover the first 3-6 months, and the rest coming after a satisfactory license is signed. That allows 3-6 months for the full license to be negotiated. This gives the company a way to get moving sooner, at a small risk to investors in case the university does not honor the term sheet, a rare occurrence.

The investors close a seed round even before the license terms are agreed, trusting that founder will find a way to make a reasonable deal. Although they invest the full amount upfront, investors may require that if the royalty or university equity exceeds a certain limit, the founders will award investors with additional shares to offset the bad news.

With your license complete, and hopefully, seed round closed, you are ready to start building the company, starting with the innermost Keep of your IP fortress