I spent 12 years managing patent portfolios in-house for a semiconductor IP company and an AI software company. I also hosted the Discussion of Patents meetup for startups for 10 years. This page answers the top questions I get from inventor-entrepreneurs.
I do not have a law firm and do not represent clients.
If you have questions, email patents@probell.com.
These are the main things startups and small companies do with patents.
Signal innovation to investors
Create a barrier to competition
Create a licensing revenue stream
Have motivational wall covering
This is important. A patent gives its owner the right to block or give permission for others to implement a claimed invention. A patent does not give its owner the right to provide products with the claimed invention. Read those sentences again carefully to understand the difference.
The patents you own can block your competitor, and other patents that your competitor owns can block you. For example, if Alice invents wheels and Bob invents wheels with gears attached then Bob can block Alice from selling wheels with gears, but Bob is unfortunate. Alice can block Bob from selling any kind of wheels.
Your patents should claim what other people might want to sell in the future. That might cover your product. It could be something different. A good patent covers more details and variations than you have in your products. This requires brainstorming when preparing the patent application.
You can deduce the answers to many questions about patents by remembering why we have a patent system.
Every powerful country in the world has its own patent system. They are all based on the same principles. Patents are a deal between inventors and the public. The inventor teaches the public about an invention, and, in exchange, the public allows the inventor to control who can implement the invention for a period of years (20 years for the most common type of invention in most countries).
By controlling who can implement the invention, a patent owner controls a monopoly. They can charge any price they want to maximize their profits.
This has two desired effects.
The ability to maximize profit creates an incentive for investors to invest the money needed for expensive research projects that might benefit the public.
Knowledge is shared. In the far future, the public can use the knowledge for free. In the near future, other inventors might be able to invent improvements.
US law 35 USC 101, European Patent Convention Article 52, China patent law Article 25, Japan patent law Article 2, and Korean patent law Article 2 are interpreted by courts as saying that inventions that are merely non-technical, abstract ideas are not eligible for patenting. Another general principal is that anything that could prevent doctors from treating patients or prevent people from managing their money are not patent eligible.
In practice, this means that inventions in these general realms require special attention when deciding whether to pursue a patent and how to write claims for.
Financial methods
Methods of doing business
Computer software implemented inventions that merely receive, transform, and output data
Methods of diagnosis based on measuring naturally occurring molecules in human bodies
It might be possible to receive and enforce patents if there is a specific, technical features and the patent claim on it describes elements with specificity, not merely by the results that can be achieved using the invention.
For patenting inventions in those realms, it is especially important to work with a patent professional who (a) studies the relevant court cases carefully and (b) has a lot of experience in the realm of the invention at hand.
See the section above.
The short answer is that you should to apply for a patent before you bring your ideas to the public or you will lose rights forever.
If you want to patent your invention outside the US, file an application in any patent office (US or other) before allowing the invention to become publicly known.
If you want to patent your invention inside the US , file an application in any patent office (US or other) within 1 year of the invention becoming publicly known and within 1 year of first offering to sell a product or service that implements the invention to anybody (even if the offer is under NDA and the sale does not complete).
A US provisional application can suffice.
Remember that patenting an invention means that it will eventually become publicly known. If you want to keep the invention secret, the deadline for filing for a patent is the day before somebody else discovers the same invention and applies for a patent or makes the invention known.
After filing an application in any country's patent office, if you would like patents in any other country, within 1 year you must either file in all of the other target countries or, to cover countries that are signatories to the Patent Cooperation Treaty, you must file a PCT application within 1 year.
Patenting has a lot of complicated requirements around dates and deadlines. You should work with a professional to avoid screwing up and losing rights.
Patents can be very valuable. It avoids confusion and disputes if it is clear who owns each one.
Many/most countries patent systems envision inventors working for an employer who pays for their living. When an application is filed or a patent is granted, it is owned by the employer. Those patent systems have to deal with exceptions and grey areas such as (a) things invented by somebody who has two jobs, (b) inventions by people whose job description doesn't require them to invent for the employer, and (c) inventions that are not applicable to the employer's line of business. One way some countries address this is by requiring the employer filing for a patent to pay a bonus to the inventors for each invention to make it clear that the company owns the rights. For this reason, it is a good idea for startups with global patenting ambitions to pay some reasonably large bonuses to inventors for each invention, no matter where the inventors live.
The US is different from those other countries. Article 1, section 8, clause 8 of the US constitution is interpreted as granting patent rights to inventors, regardless of who, if anybody, pays them. Therefore, for a company to own the rights to a US patent or application, it must have the inventors assign the rights to the company. This is best done with a good assignment agreement template, signed for each time the first patent application is filed describing an invention. It is a good backup plan to have employment agreements between each employee (including the founders and executives) and the company. The employment agreement should have template language giving the company the rights to some reasonable range of inventions, such as ones within the company's foreseeable line of business.
If the company does not get those assignment agreements, it does not own the patents that its employees invented. The employees could walk away from the company with their patent rights and give it away to any competitors for free, making the patents effectively worthless.
Not owning or having exclusive rights to its patents is one of the most common, expensive, and devastating patenting mistakes startups make!!!
Fixing patent assignment problems is usually simple and inexpensive to do if detected early. After an inventor leaves the company, it gets more difficult and expensive. Fixing assignments in response to due diligence can be even more expensive and slow down an acquisition, IPO, or vital financing. If nobody notices the problem until planning to enforce the patent rights, it can cost the company dearly.
US law makes it very important to record the assignment agreements for US patents in the public USPTO assignments database. Savvy investors will check. Get assignments taken care of early to avoid expensive problems later.
Patent lawyers are one of the most highly paid types of lawyer. Unfortunately, there is no Yelp! for lawyers. It's difficult to know who to choose.
When you are choosing the patent attorney to work with, some things matter and some don't.
Patent attorney is a credential. So is patent agent. Either can have the power of an attorney to apply for patents. Everything on this page applies equally to patent attorneys and patent agents. I will use the word attorney for convenience.
To get patents that meet your business needs, competence is more important than credentials. Anybody with the right knowledge can write your patent no matter where in the world they are or what credential they have (or don't have). A patent attorney needs two kinds of knowledge: technological and legal. Don't rely on business advice from a patent attorney.
Good patent attorneys are quick at studying unfamiliar technologies so that they can ask inventors knowledgeable questions. You know your tech, so you can tell if a patent attorney understands it well enough. However, startup founders, CEOs, and CTOs usually cannot tell whether a patent attorney is legally competent or just good at faking it. Here are the questions you can ask a US patent attorney or agent to see if they know their stuff.
1. Should I search for prior art?
If the patent attorney tells you about "inequitable conduct" or "treble damages", that's a yellow flag. It's okay if they also mention "estoppel" and "extra office actions" and the resulting extra attorney's fees. If not, that's a red flag. They don't understand the scale of different risks, and their advice will harm your patent value and cost you extra money to their benefit. Walk away from any attorney who tries to persuade you to avoid prior art.
2. I have multiple inventions written and ready to patent. How can you file them quickly?
There is no limit on how many inventive ideas can be in a patent specification or how closely they must be related. Some attorneys recommend filing multiple applications, each with one idea and pay for them all to be handled in parallel. However, it is also possible to put everything into one spec and file one application, then file sequential applications in the future all with the benefit of the early date. That can stretch out the costs for years until the company has more capital and revenue. A good attorney will tell you about that option.
3. I want patent coverage in the US and UK. Can I work with a UK attorney first?
The US attorney will probably tell you about "foreign filing licenses". No problem. Those are easy to get. But if they try too hard to dream up other reasons why you must work with a US attorney first, beware. It doesn't have to be a US attorney first for you to get US patent coverage. You can often get better quality work at a lower price in other countries.
4. What are URLs for 5+ recently published US patent applications that you wrote?
Look at the claims. US patent claims should only use two words of legalese, "comprising" and "wherein". You should be able to read the sample claims and understand what the invention does. If the claims are difficult to understand, the patent is weak. You want to hire a patent attorney who writes claims that are easy to understand. The Detailed Description section of the patent should also be clear and free of legalese, except for the last few paragraphs.
5. Notice the words that the patent attorney uses in your conversation.
If a patent attorney talks about "rejection", "allowance", and "examiner", they probably know how to get patents. However, getting bad patents is easy. If the attorney also talks about "valid", "enforceable", "infringement", and "damages", they are showing an understanding of what makes patents that are strong, useful, and valuable.
6. What percentage of the attorney fees that you pay go to the attorney?
Big law firms are like two- or three-tier pyramid schemes. Less experienced associates do the work while more experienced partners take a cut. The cut is often more than 50%. The partners spend their time looking for large clients with a lot of business. Attorneys in small law firms and solo practitioners get to keep all the money you pay. This gives them more incentive and freedom to spend enough time to do a good job for startups. Small firms for small companies, large firms for large companies.
7. Do I need an agent or an attorney?
Patent attorneys can file lawsuits in courts. They tend to bill at significantly higher rates than patent agents. For the purposes of getting patents, attorneys and agents are equally credentialed. If you find an attorney who describes patent agents disrespectfully, look for somebody respectful to work with.
8. What are some of the most important cases you have read?
Doing the wrong things in a patent application can hurt the eventual patent's enforceability. To avoid the pitfalls, your patent attorney should have read and understood at least the most important US case law. They should be able to summarize for you the important takeaway lessons from KSR, Alice, Mayo, and Festo. For software inventions, they should also know Williamson.
9. What do you know about my field of technology?
This is simply to test your patent attorney's preparation. A good patent attorney will take the time, off the clock, to read your website and do at least enough web searching to know what you mean when you use the technical words from your website.
10. Do they ask questions?
Many patent attorneys record discussions with inventors for later review. This is okay. But it's important that a patent attorney follow and understand their entire discussion with the inventors. There is something initially confusing about every great invention. Attorneys who lack self-confidence will not stop to ask a question when they don't understand something. They pretend that they understand, hope that they can get back on track as the discussion continues, and hope to rely on the recording later and look things up. That's bad. You need an attorney who will stop at important points in a discussion of an invention and ask the inventor to provide more detail.
These are some things that are not very important in choosing a patent attorney.
Years of experience
New patent attorneys should learn on large company clients first. A startup should work with a patent attorney who has at least 5 years of experience and have written 60 patent specifications. This is because the patent office takes 3 years to do things. So, it takes 5 years for an attorney to have made a few mistakes and have had a chance to learn from their result.
Beyond 5 years and 60 specifications, it doesn't matter. A lot of old timers follow habits that they learned many years ago. As case law develops, best practices change over time. It is more important to find an attorney who reads case law than one who has been in practice for a very long time.
Billing rate
There is very little correspondence between an attorney's hourly billing rate and their competence. Some who work at a high rate can do a good job in a short amount of time. Some will take longer to do a good job but bill at a lower rate. What messes up incentives is when the attorney doing the work has to share a large portion of their income with partners who are not doing the work. Most attorneys can give you a good idea up front of what it will cost. Many will even offer you a fixed fee for the project.
If it's going to cost you more than $16k USD to get a patent application written and filed, that would be a big yellow flag on price. If it's going to cost less than $8k USD, that would be a big yellow flag on quality. Each office action on merits should cost somewhere in the range of $2.5k USD to $4k USD.
Former examiner
Many former patent examiners leave the patent office and become patent attorneys or agents. That is fine. But patent examiners don't need to know about how lawsuits work. They don't need to know how to get good patents that are enforceable, useful, and valuable. Having been a former patent examiner is nice but not a strong reason why a patent attorney is a good one.
Referrals
Be careful choosing an attorney based on a referral.
It's usually 10+ years from when a patent is written until it might be tested in court (for strength) or the marketplace (for value). Prior clients rarely know, for sure, whether they have been served well. So, be careful with another startup founder's recommendation to distinguish between them just liking their attorney or knowing for sure that their attorney served them well.
If the referral is from another attorney or somebody in a company with a recurring budget for attorneys, be skeptical about their motivations. Who paid for lunch last time they got together with the attorney being referred? Does the attorney being referred send holiday gift baskets?
Personally, I always pay for lunch and give the gift baskets to my inventors. My bias is towards attorneys who have taught me useful things I didn't already know. Those are the ones I refer to startups.
In theory, an attorney must represent a client's best interest. In reality, an attorney's interest in getting paid is in direct opposition to their client's interest in saving money. Even worse, most patent attorneys have no interest in a patent after it's granted.
There are two ways that an attorney's interests could be aligned with their startup client.
A carrot.
The startup pays the attorney in equity. This way, the attorney has an interest in getting a strong, valuable patent so that the value of the stock won't be greatly harmed by a patent failure. An alternative is pay in installments stretching out over many years. Unfortunately, there are many reasons, beyond an attorney's control, that a startup can fail. So, very few attorneys accept equity compensation or installment payments.
A stick.
The attorney provides a warranty that a patent claim is valid. If the warrantied claim is held invalid in court or a post-grant challenge, the attorney pays the startup 10x what the startup paid to get the patent. Entrepreneurs are masters of risk-reward trade-offs. Attorneys are trained to avoid all possible risk. Therefore, good attorneys don't want to provide a warranty on their work. Because the guild keeps attorneys scarce, they don't have to.
It depends.
Attorneys cannot feed their families with stock in high-risk illiquid startup companies. They almost all require money. This leaves their interests unaligned with the company. Startups are cash-poor clients with little recurring business. The best attorneys don't take startups as clients. There are 100 possible mistakes inexperienced attorneys can make to fatally harm patents. Startups must be especially careful in choosing their patent attorney and then not be cheap about paying for quality work.
As a very rough estimate of cash costs of patenting expect to spend around $10k-15k USD soon to get on file, then another $5k for each additional country, then another $10k (per country) to get the application to grant.
Below are some ranges of reasonable cash rates for different kinds of patenting projects.
When you get quotes, be clear about what project the attorney is describing. Be wary of the quality of work of anybody offering to do these kinds of projects for lower rates. Carefully question anybody charging much higher rates. Some cases really do require higher rates. Biological inventions tend to be more specialized. Extra time and cost is needed for especially complicated inventions. Your preparation and behavior also matter. If you provide your attorney too little or too much or poorly formatted information or if you add material to the project while the work is in progress, that will increase costs.
You can decide.
Government agencies are slow. As a rule of thumb, imagine 3 years from filing to grant.
If somebody else will start making money by implementing your invention soon, you can ask the patent office to accelerate your application. In some offices, such as Europe and Japan, that's free. In the US, it requires a hefty fee.
Accelerating an application gives the attorney their next billable bite of your budget sooner. Some attorneys recommend accelerating unnecessarily. Your invention is, by definition, new. It is unusual that somebody else would discover the same invention and start making money from it much faster than the patent office gets around to processing your application.
In some ways, a pending application is more valuable than a granted patent. This is because patent claims usually change in examination. While the application is pending, nobody knows for sure what the final claims will be. The uncertainty to competitors is valuable to you. There is more uncertainty in high tech than in life science patents.
If you want to slow down the process to stretch the budget, you can request a 3 year deferral of the initial examination in the US and 4 or 5 years in some other countries.
This requires a CTO technical analysis and CEO business calculation. It requires answering 3 questions.
What is actually patentable?
You need to figure out what is the broadest valid claim that you can patent in view of the prior art. A professional prior art searcher (who is is usually less expensive than an attorney) might be able to help, but you can probably get a good idea yourself by searching Google Patents, Google Scholar, or the internet. It is easy to overestimate the scope of what about your invention is truly different from other known things. That leads to wasting money chasing unpatentable claims. A diligent and dispassionate prior art search is important.
How much is it worth?
You can only block others from implementing what you claim. That is probably a small sub market of the market you expect to obtain with your products. It's just the products that have every feature in the claim. If you invent a bike with a lever that can both operate the brakes and shift gears, how much more would people pay for such a bike than a bike with separate brake and shift levers? A lot of people don't care about the convenience of having one lever for both functions. Being accurate about estimating the increase in market value is important to avoid wasting money.
What is the opportunity cost?
If you didn't apply for a patent, what would you do with the money? Hire another scientist or engineer to improve the product? Run a new ad campaign to get more customers? Patenting is only worth the money if the value of its market potential is greater than the value of whatever you could do with the money if you didn't apply for a patent.
There are two ways to make money with patents.
Blocking competitors so that you can charge a monopoly price without competition
Licensing out the invention rights to others
For some types of technology in some countries, competitors sell products without regard to infringing the intellectual property of patent owners.
In general, competitors avoid infringing patents for technologies in which products are covered by one or a small number of patents such as pharmaceuticals, chemicals, and simple devices. On the other hand, complex devices and computer-implemented products have so many features that they likely infringe thousands of patents. Makers of such products tend to ignore patents when designing and selling their products. Patent owners must approach and ask the makers to stop or file a lawsuit to ask a court to step in.
The courts in some countries, such as Germany and China are more likely to block the sale of infringing products, even for complex products. In contrast, courts in the United States rarely block the sale of complex products. Instead they force the maker of the infringing product to pay a royalty as if they had licensed the patent rights for the patent owner.
Unfortunately for the owners of patents on complex products in the United States, they probably cannot block a competitor with patents. If the patents are strong, they can probably get a royalty payment. Weak patents tend to be practically worthless. That is why it is so important to have the help of a capable and motivated patent attorney.
With a strong patent and a friendly approach, it might be possible to keep a competitor out of the market for products that implement the patent with just a letter and a few emails and phone calls. A few thousands dollars of help from an attorney might help.
If that fails, you might at least negotiate to license out the patent rights to the competitor for a royalty.
If a competitor holds out and will not negotiate, it might be necessary to file a lawsuit. A lawsuit filed by a prompt settlement agreement costs tens of thousands of dollars.
If it is necessary to go all the way through a trial and court ruling, it would typically costs hundreds of thousands of dollars in most other countries, but millions of dollars in the US.
There are millions of patents, tens of thousands of patent-related agreements between companies each year, and only a few hundreds that go through a full US court trial. It is worth considering the high potential cost of enforcing US patents, but it is also important to consider the low probability of having to go through that.
This depends on the strength and scope of the patent's claims and the country and field of the patent's technology.
Strength - If the competitor's patent is weak, it would not stand up to a review challenge at the patent office. You can probably use the technology in the patent claims without much concern.
Scope - If the patent claims have a narrow scope, you might find an alternative design that is nearly as good and avoids infringing the competitor's patent claims.
Even for a patent that is strong with a broad scope, you might be able to get around it depending on the country where the patent is granted and the field of technology.
In the US, patents covering computer-implemented products and other complex devices, companies tend to license or cross-license each other's patents based on the number and breadth of the patents and the amount of profit the product implementer is deriving from using the claimed inventions. You can view your competitor's patents as a tax on your profits. You can proceed, unblocked, to make and sell your products and wait from the competitor to approach you to collect the tax. If they are much larger or much smaller than you or if competition is not very tight, they might never try to block or tax your profits.
In the US, if you are making or selling products that can be covered by just one or a small number of patents, such as pharmaceuticals, chemicals, or simple mechanical devices, you are at significant risk of being blocked by competitors' patents. Companies in those businesses would be wise to do a patent landscape analysis before entering the market to avoid being blocked.
In other countries such as China and Germany, it is more common to be completely blocked by competitors' patents. It is important to have some leverage, such as your own patents covering their products or be willing to negotiate permission and pay royalties.
To get free general education, read this web page. You might also find helpful information on Quora, Reddit, or StackExchnange.
If you still have a question, email patents@probell.com.
If you want advice that is specific to your invention and business situation, you will not find reliable advice for free.
Most major countries' patent offices are happy to do outreach. You can call your country's patent office. However
They are in the patent business, so they will almost always recommend pursuing patents
Their experience is in examining claims for validity but not writing patent claims that effectively cover products, so they will give you bad claim writing advice
They have no experience with your business, markets, or competitive landscape
The few law schools that have IP courses often make their students available for pro bono patent work. Some law firms provide their inexperience attorneys for pro bono work, too. However
They are in the patent business, so they will almost always recommend pursuing patents
They are the least experience patent professionals and rarely know the best claim drafting techniques for strong patents
They have no experience with your business, markets, or competitive landscape
So, get all the general education you need for free, then pay for an experienced patent attorney or agent that passes the tests in the section of this web page about choosing a patent attorney.
This is tricky and easy to get wrong. Getting it wrong can make some patents unenforceable and therefore essentially worthless.
In most countries, patents are granted to and owned by the employer of the inventors. In the US, patents are granted to the inventors who (usually) sell their rights to their employer in exchange for their salary and (maybe) a bonus. Patents are (potentially) very valuable pieces of property. It would be wrong to deny patent rights to a true inventor by leaving off their name.
Leaving off the name of a true inventor destroys a patent, making it unenforceable and potentially worthless.
It would also be wrong to give rights to valuable patent property to somebody who is not a true inventor. On academic publications, it is common to name long lists of co-authors who contributed in any way. However, patents should only name people who contributed to figuring out how the claimed invention works. Patents should not name somebody who implemented the invention (wrote code, setup the fabrication process, did the lab bench work, etc.) if that person did not contribute to thinking of the invention. Also, patents should not name somebody who identified the problem or assigned the inventors to the task (boss, project manager, market researcher, etc.) if that person did not contribute to thinking of how the invention works.
It is very common in startups that a co-inventor leaves the company before the patent application is filed. Sometimes, the remaining inventors don't want to talk with the inventor who left. That is okay. Unfortunately, it is common that such startups leave off the name of the inventor who left. That is not okay. It can destroy an otherwise valuable patent. It is important to work out an arrangement with the departed inventor early rather than lose the patent protection after building a successful business around it. An attorney can usually help to negotiate with the departed inventor.
If a patent lawyer steals your idea, you can cause them to lose their license, face professional discipline, and potentially be disbarred. Since they learned the idea from you, they are not an inventor so they could not get a patent for themselves, nor could anybody else they tell the idea to. It is highly unlikely that a patent attorney or agent would give up their profession and get nothing of value from it by stealing your idea.
If somebody claims to be a US patent attorney or agent, you can check with a Practitioner search at the USPTO Office of Enrollment and Discipline.
They could, but they probably wouldn't. They could not get a valid patent on the idea since you are the inventor and they only learned the idea from you. They would be starting a business without the background domain knowledge that helped you invent the idea, and they would be behind on any development work that you did before telling the investor the idea.
An investor could give your idea to another company they invest in. But, if they did, that company probably wouldn't take it. Most companies have their own product development plan and aren't going to redesign their product because an investor has an idea.
In any case, since you are the inventor, you can get a valid patent, and nobody who learned it from you could can. It would be a waste of their time to try to build a business on your patented idea.
Most employees sign employment agreements that give the products of their work to their employer in exchange for the company paying a salary. That's pretty reasonable. If you start a startup, you should have everybody sign such agreements.
Inventions can be a product of an employee's work. Most such agreements include giving the employer the rights to inventions that the employee makes. Depending on the country or state, the job description, where and when the employee conceived their invention, and how closely the invention relates to the employer's business, the employer might own the patent or trade secret invention rights.
If you are planning to start a startup, it is ideal to quit your old job before inventing ideas for the startup. If your invention is in your employer's line of business, and especially if you invented it as part of your usual work, it is likely that the employer can claim the invention rights. If you feel like your invention is not something your employer would pursue for their business, you still might want to meet with somebody friendly in the legal department or the executive team and ask them to agree in write (even just an email to your personal email address) that the company makes no claim to the rights to your invention.
For example, if you work for a database software company and you invent a better can opener, it is likely that you can start a new company and own the invention rights. If you work for a database software company and invent a better way to access APIs, there is a higher likelihood that the company would feel that it should own the invention rights.
What you should NOT do is start a startup based on an invention that the employer might feel that it should own. Probably nobody would notice or care for a long time. But you might find that when your startup becomes very successful, the old employer will claim ownership, causing the startup to fail or at least have an expensive lawsuit.
If you can read and think and you have enough time, you can actually look up all the laws yourself without a lawyer. Most countries publish their laws online. You can look up all the US federal laws statutory laws made by Congress here.
https://www.law.cornell.edu/uscode/text/
US Patent laws are in Title 35.
Because the US has a common law legal system, we look to past case law from the courts whenever the statutory law is unclear. You can search case law here.
https://scholar.google.com/scholar?as_sdt=6
By searching the internet, you can find the laws of the US states, other countries, cities, government regulations, and most other stuff that people pay lawyers to look up. If you have time, save yourself some money, and do your own legal research.