Startup Patent Portfolio Roadmap

A Roadmap For A Startup’s Portfolio

A startup’s patent portfolio can be thought of in two distinct patent types: prophetic patents and non-prophetic.  Non-prophetic patents can also be thought of as “data-driven” patents.  Prophetic patents are a necessary evil, but they can be very damaging to a startup when used badly.

Prophetic patents are those that are almost purely forward-looking.  These patents are filed prior to raising funds or at least before going to market.  They are prophetic in the sense that they are guesses about how the technology will work and how the market will adopt the technology.

Non-prophetic or “data-driven” patents cover things for which data exists.  The data may be performance-related test data, or market-related data that comes from customer behavior.  The patents that come from market-related data tend to be some of the most valuable patents in a portfolio.

The idea of “data-driven” patents is that there is substantive data that supports the business proposition of the patent.  The substantive data can be test results, but the most important data are market results.  When the patent captures a deep market insight, such as a customer pain point, that patent has strong commercial value.

In general, data-driven patents will have much more value than prophetic patents for two reasons:

  • There is less guesswork about the technology, which means the claims will be more meaningful.
  • The decision to protect something has real business value, meaning the money is well spent.

The data-driven patents can also be thought of as risk-reduced patents.  The risk of a patent is the technology risk and the market risk.  When data exists on either the technology or market risk, the overall risk of a patent goes down.[1]

A startup needs prophetic patents at the very beginning to get started.  These will protect the basic idea of the business’s product, but it is critical to avoid creating downstream problems with these patents.  As the startup grows, it will understand the markets better, and it will solve technical problems on the way to realizing its goal.

These two elements – understanding the markets and solving problems along the way – are where the real gems of a patent portfolio lie.  However, these assumptions cannot be verified when the first patent is written.  The discussion below provides solutions to handling this dilemma.

Filing The First Patent

The first patent is typically filed prior to entering the market.  This prophetic patent might have several different ways a product may be designed and capture a couple different ways the product may go to market.

At the point that the patent is filed, there are only guesses about how the market will respond.  These guesses may be risk-reduced by doing market research, testing landing page conversions, or just talking with potential customers.  However, the real market response comes when customers break out their credit cards and pay money for the product.

The purpose of the first patent application is to clear some space so that the company can keep competitors away.  This patent application is done with the highest degree of uncertainty about both the technology and the market.  It is critical to note that not only is the entrepreneur just beginning the journey at this point, but so is the patent attorney.  Neither player knows which elements of the invention will turn out to be important.

The first patent will NOT be the most important patent in the portfolio, so do not treat it like it is.

Conceptually, the first patent might be merely a placeholder or framework into which other patents will follow.  Practically, this patent application must capture the one or two points of novelty that differentiate the product – and nothing more.

The claiming strategy for the first patent should cover the points of novelty of the product as it will be introduced, and the claims should be somewhat broader than normal.  The strategy is to get some feedback from the examiner (i.e., a rejection) and get a better lay of the land for what could be patented in this area.[2]

Do NOT put too much information in the first patent.

Far too many inventors and entrepreneurs file gigantic, self-written prophetic provisional applications.  These are a very bad practice.

“Kitchen sink” patents – where everything is thrown in – can cause endless problems down the road.  These patent applications are brain dumps of as much information as possible.  Some people even copy pages from their lab notebooks and stuff them into a provisional application, mistakenly thinking that somehow they are protected.

This is one of the worst practices.

Disclosing too much information prevents getting a patent on the actual invention when the research is actually invested to figure out how to do it.[3]

Everything in a patent application is prior art which will be used against the company down the road.  This includes provisional applications, which are publically available once their non-provisional cousins become public.

When a provisional application states that something is possible but does not explain how to do it, an examiner can still cite it as prior art.

The scenario: An inventor of an artificial sponge product writes and files a “kitchen sink” provisional application and happens to mention that fibers could be included in the product.  The inventor had a guess that fibers could be added, and the concept was on the long list of things to explore.

Three years later, there is a huge need in the market for stronger sponge products.  In fact, several customers were offering to pay an enormous premium to have the stronger products.

So, the inventor tries adding fibers to the product.  It turns out that adding fibers was very difficult, and there was a lot of development to determine how to distribute the fibers in suspension and how to keep the fibers in the product while it cures, and the length of fibers really made a big difference.

Proud of the accomplishment, the company submits a patent application for the fiber, unaware that there was a mere mention of the fibers in the original patent application.

The examiner rejects the claims stating: “The inventor said that it was possible to add fibers to the mix in the provisional application three years ago”.

Now the company is stuck.  They cannot argue with their original patent without weakening it completely, nor can they argue that their new patent application deserves broad coverage.

Even if the company is able to save the second patent by claiming priority to the first patent, the company loses all international rights to the second invention.[4]

The purpose of the first patent is to cover the “big” point of novelty of a business, but the next batch of patents are where the real protection begins.  These are explained below.

Know What Is Protected – And What Is Not

Many CEOs mistakenly tell investors what they think their patent covers, and this can be very misleading.  The easy way to tell is when the CEOs make excessively broad, sweeping statements about what their patents cover.  In many cases, this is not because the CEO is being disingenuous or deceitful; it is because they do not understand this basic concept:

Patent protection comes from the claims, not from the description.

The patent protection is defined completely and totally in the claims of the patent.  Any information in the patent application that does not support or relate to the claims is damaging.

When a patent attorney finishes a draft of a patent application, the inventor reviews it.  The inventor almost never reviews the claims but usually reads the description thoroughly.  At the end of this process, the inventor believes that there is protection for everything contained in the application.

Legally, the only protection comes from the claims, not the description.  In fact, anything extra in the description hurts the company.

How is it damaging (other than the prior art problems described in the previous section)?

Information that is disclosed and not claimed are trade secrets that are being handing over to competitors free of charge.

Remember that the quid pro quo with the government is that the patent applicant discloses trade secrets in exchange for the patent.  The purpose is so that competitors can use the patentee’s trade secrets to improve on the invention.

The “kitchen sink” patent application – especially the ones done at the beginning of a company – often lay out the complete business and technology plans of a company.  This can be gold to a competitor and give them deep insights into the startup’s plans.  Is this a good thing?  No.

Where To Look For The Next Batch Of Patents

The best patents have data to back them up.  While every patent can have some prophetic elements, the best ones are based on real technology or market data.

As a company brings a product to market, the company solves various problems that make the product possible.  The entrepreneur often sees these problems as a nuisance, but they are often the most valuable patents in the portfolio.

Valuable patents are those that solve the same problems competitors will have to solve.

Patents on the best way to solve a problem that competitors will face can protect a business much better than a patent on the end product.  For example, a business may produce toothbrushes that have a pumice polisher.  It may be just as effective – if not more so – to have a patent on a mechanism to attach the pumice polisher than to have a patent on the toothbrush with the pumice polisher.

These inventions pop up when it takes some effort from the engineering staff to solve a problem.  One way to find these inventions is to ask the engineers what was the hardest problem they had to solve on this product?

These patents have the technical data of a problem solved to help justify the patent investment.  In some fields, such as chemistry and biotechnologies, the examiners expect to see actual test data.

From a business standpoint, these patents can be easy to analyze from a design-around alternative.  In the example above, nobody had ever figured out how to attach the pumice polisher to a toothbrush.  A patent covering the attachment mechanism would likely be:

  • easier to get through the patent office;,,
  • broader than the prophetic patent on the toothbrush;,
  • much more effective at preventing competition; and,
  • have outbound licensing

A patent on attaching the pumice polisher could be used for putting pumice polishers not just on toothbrushes, but to other personal care products such as fingernail polishers, foot care products, and the like.  The pumice polishers could be added to scrubbing and cleaning products like pot and dish scrubbers, toilet cleaning brushes, countertop cleaning brushes, floor scrubbers, and many others.  There are probably many industrial uses as well, such as inside grinding and finishing machinery, and so forth.

This patent could be widely licensed into many different industries.  One use might be to trade this patent in an exclusive cross license that would give the company the exclusive right to another company’s technology within the personal health industry in exchange for that company’s exclusive right to the pumice polisher in their industry.

These types of patents that come out of the problems solved in product development often are some of the most valuable technical advances and are often some of the most overlooked.

Valuable patents capture customer value.

Whenever possible, patent claims should focus on the customer’s problem and solutions for it.  These inventions come from the market data – not from the engineers.

Sometimes, identifying the problem is where the invention happens.

For example, an automated paper towel dispenser solves the problem of being exposed to germs after washing your hands.

The patent claims could be written to have motors, sensors, and all the components in a specific configuration – or the claims could be written about a device that produces a paper towel without a user’s touch.  In the first case, patents would have to be filed for every version of an automated towel dispenser and inevitably, someone could design around the claims.  In the second case, a single set of claims could capture the entire market.

The first type of patent – the one with the motors and sensors – is something that would come from the engineers who are focused on the solution.  The second type of patent – the one that described the customer problem (no touch towel dispensing) – comes from marketing.  The second one is problem or market-focused, not technology-focused.

These types of patent claims come from market-related data.  No-touch paper towel dispensing is a marketing message, not a technology message, because it focuses on a problem that consumers may or may not know that they had.  Once the consumers are aware of the problem, all of the competitor’s paper towel dispensers look like dirty, germ-infested petri dishes.

In many cases, the market-related inventions came after extensive trial and error.  Market research and understanding consumer reactions are every bit as difficult – and valuable – as the research performed by engineers who build the products.  When those insights come, they are excellent fodder for patent protection.

Sadly, the marketing and sales professionals are not as familiar with the patent process as the engineers, so getting a patent is not their go-to thing.  Marketing and sales professionals are constantly talking to customers and are at a high risk of disclosing the invention to customers before the patent is filed, which is a big problem.

Building Out The Portfolio In Response To Business Signals

As the startup enters the market, there will be multiple course corrections.  Problems with the technology will be corrected, but the market responses will also cause the product to morph and change.

Entrepreneurs have a vision of what the market wants, but often the market wants something different.  A good entrepreneur will test several different go-to-market strategies, including different ways of reaching customers, different value propositions, different product configurations, and different pricing models.

A good entrepreneur will develop ways of identifying the signal from all the noise.  What are the key elements that resonate with the consumers?  The engineers thought that the pumice polishers in the toothbrushes were the defining feature, but the consumers may have liked the unique colors of the packaging.

An entrepreneur is wrong more times that right, but good entrepreneurs are constantly learning and responding to what the market says is important.  With each data point that can be validated, there is learning, and with learning, there should be some consideration as to whether a patent would help (or harm) business exploitation of that learning.

Patents are not the only tool in the toolbox but can be a very efficient and economical way of making business moves in some circumstances; many of those circumstances are discussed in this book.

For example, a startup company may know that a bigger competitor is already in their space but does not understand the problem as well as the startup, and therefore the big competitor cannot develop as strong a solution as the startup.  These insights can be patented, which makes the startup a much stronger acquisition target but also positions the startup to cross-license technology with the competitor if the big competitor were to assert their patents.

In another example, a company’s products may be receiving a good market response and generating great cash flow.  However, the market trends are towards miniaturization.  One or two early prophetic patents might be useful to address the technical problems of miniaturization, followed by some data-driven patents as those problems are solved and have market validation.

A well-run company will add patents to their arsenal of tools for responding to business needs.

Culling the Herd

It is very hard to walk away from a pending patent application or even an issued patent.[5]  In many ways, the patent application represents the hopes and dreams of the inventor and, by extension, the company.  The patent is a business tool, however, and tough decisions have to be made when those tools are not providing value.

Culling the herd comes down to two basic forms – abandonment and sale. When is the right time to abandon a patent application?  When should an issued patent be sold off?  Both of these decisions should be relatively easy to make when the asset no longer aligns with the business objectives.

When should a patent application be abandoned?

Abandoning a patent application can happen for different reasons, but often it occurs when prior art pops up that severely changes the business value of the patent.

These decisions can be easy when there is a due diligence package done ahead of time.  The due diligence package may include assumptions for the scope and breadth of the possible patent, along with the business value of the patent if it were to issue with the assumed scope.

When an examiner finds a prior art reference that severely changes the scope of the claims, the decision to go forward or not should come from re-running the analysis done during due diligence and making a decision.

The analysis will verify that the due diligence assumptions are still valid, and if not, the due diligence analysis should be updated.

Sometimes, the changes in the patent scope during prosecution may change the business case for the patent.  Again, the due diligence analysis would be re-run.

It is easy to fall into the sunk cost fallacy.  Sunk costs have no bearing on whether or not to proceed.  Any cost analysis should be based on whether paying for the next step in prosecution is a good investment on a risk adjusted basis.[6]  Will the investment in the next step yield a useful asset with a meaningful risk-adjusted return?

As the examination process continues, the risks become more clear.  When the examiner cites a very close prior art reference, the risk that a patent will never issue with our intended scope greatly increases.  When the examiner has a hard time putting together an argument to reject our claims, the risk decreases.

When should a patent be sold off?

In the US, maintenance fees are paid at 3.5 years, 7.5 years, and 11.5 years after the patent issues.  These are the classic decision points for asking whether or not to invest more money in a patent.  If the maintenance fees are not paid, the patent will lapse and be dedicated to the public.

Startup companies typically are no longer a startup when they have enough patent assets to worry about paying maintenance fees, but larger companies with big portfolios incur giant costs at these intervals; many large companies try to sell off the patents or let them go abandoned at these points.

For a startup, a periodic review of their patent portfolio may identify patents that they are no longer using.  When that occurs, a decision may be made to sell or license the asset.

Usually, the startup’s patents come from the history of product development, and so most of the patents should remain relevant.  However, when there is a significant pivot to the company’s product and thinking, there may be some unused assets.

Patents that are not infringed will have de minimus value.  There may be value to transfer the patent assets to a new company which may seek funding and try to learn from the first company’s mistakes, but in general, liquidating a startup’s uninfringed patents are often not worth the effort.

However, the due diligence package should identify possible infringers at the time the patent is written.[7]  A list of possible buyers of the asset can be formed from the possible infringers, and a professional patent broker may be able to render a judgment as to whether they think the patents can be sold and for how much.

Selling On The Secondary Patent Market

The secondary patent market often appears to be a deep, dark mysterious place where huge deals get done.  In truth, it is not all that mysterious and not that many huge deals get done.

The secondary patent market is where patents get traded.  The players are the people liquidating assets, people acquiring assets, and the brokers and other people in between.

Several companies have been formed to try to establish a single market, but nobody has been able to generate enough momentum to consolidate the market.  These include auction houses, patent listing services, and other systems.

Part of the mystery is that most acquirers do not want other people to know that they are acquiring.  A big company that wants to move into a big market typically acquires patents in that market prior to making a move.  Once word gets out that Big Co is buying patents, the prices rocket skyward and they can no longer get assets at a reasonable price.  Further, as the word gets out, patent owners in the space start putting together war chests to enforce their patents to make it doubly difficult for Big Co.

The guide through this morass is a patent broker, and specifically, a seller’s agent.  These are people who sell portfolios for a living.  They have a deep Rolodex of names and can get an audience when they have good assets to sell.  They also have an ear to the ground and know who is buying and why they are buying.  Brokers charge 20-25% of the transaction price, and it may take six [spell out] months to three  years before they can sell an asset.

The secondary market is fueled by assets that are infringed.  It is impolite to say “infringed;” the proper term is “evidence of use.”  Basically, the only patents traded on the secondary market are those where there is solid evidence that another company is using the invention.  Patents without evidence of use are of no interest to brokers in this market.

Typically, a portfolio might have one or two killer patents that are infringed and a pile of other related patents that are not infringed.  These usually sell as a bunch and credit is only given for the good patents.  The rest are just fluff.

The secondary market is where a startup can unload assets that it is no longer using.  Some companies try to keep a non-exclusive license for itself.  This can be done, but it will devalue the assets and make them harder to sell.

Before selling patents, spend a lot of time looking for a good broker.  It is somewhat like looking for a good used car salesman, but interview two or more brokers and ask them about their own reputation as well as the reputation of the other brokers being interviewed.

The ideal thing is to ask the broker’s customers, which would be a patent buyer at a big company.  It is a pretty small world, and these deals are based on relationships.

The secondary market is ever changing, with people entering and exiting quickly and new business models popping up repeatedly.  A good broker will know about them and be a guide through the maze.

An important thing to note: the secondary market is where the good patents are separated from the bad ones.  Only a small percentage of patents actually get traded, and those that do are well written, well prosecuted, and have solid commercial value.  A broker only cares about top-drawer, investment-grade patents that have solid evidence of use.

[1] One of the fascinating aspects of patents is that they must have inherent risk.  If all of the risks are removed, the invention becomes more “obvious” and it becomes more difficult to be allowed by the examiner.  Every patent has to have an “inventive step” or “ah-ha moment”.

[2] Often, this is done by having both relatively broad and relatively narrow claims.  The narrow claims are designed to be allowed quickly, giving the startup a patent it can use for business, and the broad claims force the examiner to give a landscape of the allowable subject matter.

[3] See also where the quid pro quo between the patent applicant and the government is discussed.  Making unnecessary public disclosure of trade secrets is also damaging to the company.

[4] This is one example of why it is a good idea to keep continuation applications open on early patents.  This book does not discuss this strategy in detail.

[5] Patent abandonment is the greatest fear of any patent attorney, so be prepared for a litany of reasons why a patent should not be abandoned.

[6] It is just as easy to fall into the trap of endless nickel and dime investments into an asset that will have little commercial value.

[7] It feels like an expensive and painful process to document this knowledge, but it is infinitely helpful when making business decisions about the patents over the patent’s 20 year lifetime.

This is an excerpt from “Investing in Patents” by Russ Krajec.

Investing In Patents is available at

Leave a Comment

You must be logged in to post a comment.