The True Cost of Obtaining a Blockchain Patent

A while back I thought it might be interesting to estimate how much a large corporation ends up spending on obtaining a USPTO patent in the blockchain space. This is after I found a web page by a patent attorney in which blockchain was described as a difficult emerging technology requiring more application drafting effort [1]. You don't say!

The more I thought about it, the more I realized that there are also lots of hidden costs that run of the mill articles completely overlook, so I reasoned it would be entertaining and illuminating to include them too.

I was surprised by the final breakdown.

Summary

If you don't want to examine the whole process by which I came up with the resulting numbers, here are my estimates:

A granted blockchain patent costs a large company at least $85,000.

The USTPO costs comprise $6000 of that.

The "invention" cost, i.e. how much is spent on allowing engineers to invent, is about $3000,

Another $3000 in bonuses go into the pockets of the engineers ... if the patent is granted.

The rest is spent on lawyers, drawings, searches, management and administrators.

Startled, and feeling some disbelief? Well, then read on.

Paying the Piper

So, let's start with the simplest cost - the USPTO fees. We are looking at a "normal entity" - a traditional corporation. Small organizations, like independent inventors, qualify as small entities and pay half the cost, and universities are, bizarrely given their size, classified as micro-entities, and pay a quarter the cost.

We have to make some assumptions, so let's assume: less than 20 claims, less than 100 pages, no mistakes made during filing, no request for a time extension, only one request for continued examination, with an allowance after that.

If you're unfamiliar or not bothered with the technical details - that would be an average patent application with no administrative mistakes made.

This racks up the following fees payable to the USPTO:

So far, so good.

Just browsing

Large corporations always conduct a patent search, or more usually, outsource it, to make sure they're not wasting their time pursuing a patent for something that's already out there. There's a risk here, because searches can't find applications that were filed beforehand, but haven't been published yet (and that takes 18 whole months!), but at least it reduces the risk.

You also want the search to come with a recommendation as to how significant any possible prior art is. That's called an "opinion".

IP Watchdog estimates that a patent search of that kind is about $2,500 to $3,000 [2], so let's take the average.

I'll let you in on a secret: because I am a "small entity" (according to the patent office, and not my wife, I hasten to add), the cost of a patent search is more than the cost of simply filing an application in my case. So I take a day or two out after I think I have a valuable invention to browse Espacenet, Google Patents, and the web in general and conduct my own search.

Big companies can't do that.

A picture is worth a thousand words

The patent has to have drawings. And the more the better, up to a point. Ten seems like a reasonable number, and it appears that the service can be surprisingly cheap, at about 50$ per drawing, but can be as high as 200$ for complicated concepts. Companies don't go for the cheapest, so let's settle on a simple fee of 100$ per drawing.

When I started drafting my own patent applications I read through the patent office drawing requirements, and now use Inkscape to draw my own because it allows me to refine my sketches. It helps the inventive process, and at the end of it, I have the pictures I need. In over twenty applications and a total of about 160 drawings, I've only ever had to correct one of them.

And I could make the correction almost immediately.

The law may be an ass, but it's an expensive one

So now we come to what most people consider the "big expense" of obtaining a patent - the lawyers. I will split this into two parts: the application (writing up the invention and submitting it), and the prosecution (arguing with the examiner to finally get a grant).

Patent attorneys don't come cheap. They have both a STEM degree and a law degree, so it's fortunate that on average they aren't twice as expensive as a normal lawyer. An estimate from a legal firm is that patent attorneys charge from 300$ per hour up to 600$ or even more for the real hot-shots and specialists.

I was lucky enough to get to know one of my former employer's hired patent lawyers well enough that he gave me what we call, in the UK, "mate's rates" - that is, he reviewed a set of claims I'd written at the company rate, but put double the time in. I was more than happy with the expense, because a) I got a bargain and b) he taught me a significant amount about patent claims. But for a small entity like myself, it was still a significant expenditure.

An eye-watering one.

Drafting the patent

I am going to estimate on the upper average side (big companies don't hire the cheapest) and select an average hourly rate of 500$, with two hours billed for meetings with the inventors (more about them later), and 40 hours for the write-up.

Why 40 hours? Well, it takes me about 60 to 80 hours to get to the point that I am ready to log on to the USTPO site and submit a patent application. So I thought I'd go for half of the upper figure. A patent attorney may have the law at their fingertips, but I know my inventions inside-out.

And, after all, we are talking about high-quality blockchain patents here, so the attorney has to understand them, and probably is learning about blockchain in the process.

Again, note that this is a low-ball figure. I've seen numbers as high as $35,000 or even $40,000 quoted for the application costs of boring database patent applications.

Prosecuting the patent

Now the patent has been drafted and filed, the pretty pictures have been drawn, it's been published, and the press has mistakenly published over-excited articles that the company has a blockchain patent.

They don't - now the prosecution stage starts, in which the examiner assigned to the case repeatedly rejects the application, and the patent attorney repeatedly argues that the examiner is wrong.

I have had rejections where my rebuttal took less than half a day: five minutes to spot the flaw in the examiner's reasoning, and three hours to write a response and file all the required admin.

Most of them, however, take me three to four days of chess-like researching and pondering, and some have taken five or more days spread out over a couple of months. And I'm not billing for the work my subconscious was doing over that time.

Now, patent attorneys are going to be quicker at spotting legal flaws in the examiner's rejection, but much slower at spotting technical flaws, if they spot them at all.

Generally speaking, I think that traversing the examiner's rejections (that's legalese for explaining why the examiner is wrong in a convincing manner) is going to be much more effective if the argument points out why the examiner is wrong on an engineering matter rather than a legalistic matter. It's one of the few advantages the pro-se inventor (i.e. inventor who is pursuing his or her patent directly rather than through representation) has over professionals.

But back to bookkeeping - let's be frugal and assume a day for thinking, and half a day of paperwork in billable hours per rejection on the part of our brilliant attorney, and one phone interview with the examiner followed by a write-up of said interview. And as stated before, a total of three rejections:

I hope you'll agree that that is a very low estimate.

Miscellanea

Well, surely we are done now? All the paperwork has been completed, the patent has been written, argued about, and finally granted, and all the fees have been paid.

Oh no, no it isn't. Not when it comes to company costs, and in this case, they are mostly hidden ones.

And the award goes to...

In decent big companies a small amount of cash is handed over to the inventors when their patent application is filed, and a slightly larger amount when it is granted. In my experience, it is typically $500 per inventor (up to three) for the filing, and $1000 for the patent grant.

Companies can and sometimes do cut the latter expense by laying off their inventors before the patent is granted. But we won't account for that here. So let's go with two inventors who are actually appreciated and see their careers through to the final grant date:

... But there's more

Now unless you've been hired into a research department, it's a fact that as a software engineer in a big company not that interested in blockchain, you are going to be doing your innovating and investigating in your own time. And yet, due to your employment contract, they own your ideas.

However, there are still some hours that will be taken out of your daily routine of running flake8 over your Python code or performing Github code reviews before changes are pulled into the main code branch or something like that, to write up your invention disclosure, attend patent board meetings, and explain to the patent attorney what it is that you have actually come up with.

Here's an assumption: if you have both the knowledge of the technical area and of the processes surrounding submitting a blockchain invention, you're probably going to be at the more senior end of the scale.

I'll go with a principal software engineer, but I won't add in remuneration from stock options. That gives us $160,000 per year [3], which is an hourly rate of $77. And as if by magic, a second search [4] has the figure of $74, so that's close enough.

I then went and put in some laughably low costs for office space, medical insurance, and suchlike into an expenses calculator [5] which bumped the total cost up to 90$ per hour.

Wow, engineers may have what's considered good earnings, but they're still lagging behind patent attorneys!

So, let's allocate two hours within work to discuss the idea and write up the invention disclosure, two hours in a patent review board meeting, and two hours in meetings with the attorneys. And then another hour to review the final patent draft to make sure it's not nonsense, for two engineers:

At a total cost to the company of less than $2000 for all the actual "inventing" work, engineers are dirt cheap! It makes you think that perhaps hiring ten times as many inventors and using trade secrets rather than patents might actually be the way to go.

I think that may be what Tesla is doing [6].

Bored on the board

The patent review board, in my experience, typically consists of a Director of Research, Senior IP Counsel, a couple of junior lawyers, and perhaps a secretary to take minutes.

Big companies don't seem to take into account how much these entities cost them per hour.

I don't really want to insult them, because despite only one in four of my invention disclosures passing through their gates and being filed, I've always found patent review boards both courteous and diligent, and I like both of those characteristics. Nevertheless, I am going to make the broad assumption that the combined hourly cost of those five people to the company is only the equivalent of three of the lowest-paid patent attorneys at $300 an hour.

Oh, but hang on, for every patent filed, they spend an equivalent amount of time on three more inventions, so we have to multiply their cost by four.

And of course, they spend a couple of hours reviewing the patent application documents for inventions that they give the green light to, resulting in:

It's not over till it's done

But wait! We are not finished yet. That is just the running total for the total actual cost, as far as I can work it out (and without taking the costs of HR and the janitor into account) for obtaining a patent if it is guaranteed that the patent will be granted.

I've read a paper in the Yale Journal of Law and Technology titled "What is the Probability of Receiving a U.S. Patent?" so you don't have to [7]. In summary, about two-thirds of patents in the software space are granted.

That means we need to multiply the cost of obtaining a blockchain patent by 1.5 in order to get the true figure. The cost of winning the lottery can't just be the price of the winning ticket - it has to include all the other non-winning ones too.

And so without further ado, according to my calculations the cost to a large corporation of obtaining a blockchain patent is:

And this is a conservative lower bound. The real cost could easily be twice as high, and my gut feeling is that it probably lies closer to $120,000.

As always, I welcome responses and criticisms. If you spot something that you think is patently (sic) wrong, please let me know in the comments!

About the author: I am a blockchain researcher and inventor named on 42 patents to date. My company Chainfrog currently holds a portfolio of twelve granted blockchain patents that we are looking to sell (see www.chainfrog.com ). I am also available for invention-related consultancy. Feel free to contact me at [email protected] if you are interested in that.

[1] https://ocpatentlawyer.com/patent-application-cost-short-and-long-term/

[2] https://www.ipwatchdog.com/2015/04/04/the-cost-of-obtaining-a-patent-in-the-us/id=56485/

[3] https://www.indeed.com/career/principal-software-engineer/salaries

[4] https://www.salary.com/research/salary/posting/principal-software-engineer-hourly-wages

[5] https://www.tsheets.com/resources/determine-the-true-cost-of-an-employee

[6] https://www.tesla.com/blog/all-our-patent-are-belong-you

[7] https://digitalcommons.law.yale.edu/cgi/viewcontent.cgi?referer=&httpsredir=1&article=1113&context=yjolt