How To Save 300,000 Dollars by Gaming the USPTO through Continuations
This article is for the patent nerds out there, and involves a thought experiment for significantly reducing your patent application and prosecution fees when you have hundreds of ideas and you don’t know which ones you should attempt to get a patent for. I would love to receive any feedback as to the feasibility (or lack of it) of the proposed scheme from those who understand the US patent system better than I do.
Introduction
Imagine you have a miraculous month in which you invent a hundred different diverse things, and you want to patent them all in the United States, because they all seem like they’re novel (or at least non-obvious), and they all look like they might be commercially viable.
The Specification
Let’s make this more concrete, and assume that you are a skilled pro se inventor (that means that you have enough patent law knowledge to obtain the patents without the costly help of a patent attorney), and you have invented:
- a new kind of lawnmower
- a more efficient soft drink dispenser
- a computer database with improved search capacities
- a replacement for shoe laces
- … and so on right up to
100. a biodegradable paper bag that never tears
As you can see — they’re different and diverse.
Fees, Fees, And More Fees
What follows is an optimistic assessment of costs and patent grants — I am going to assume that every single patent application will to convert into a patent.
As an independent inventor you qualify as a “small entity” at the United States Patent and Trademark Office (the USPTO), and so the cost of filing a patent application is:
- $64 (electronic filing for small entities)
- $280 (utility search fee)
- $320 (utility examination fee)
= $664.
That means that to file your one hundred applications is going to cost you $66,400. That is a lot of money, just to start the process.
You then have to prosecute (the fancy legal term for arguing with a patent examiner that you deserve a granted patent) each one, which involves more fees, and takes a lot of time.
This study allows us to estimate the number of office actions you will have to respond to, and the number of requests for continued examinations that will be required. Office actions are what the responses from a patent examiner are called, and usually it’s the examiner explaining to you why you don’t deserve a patent, and you are required to respectfully respond to it with a counter-argument until one of you gives up first. If that’s the examiner, you get a patent.
We’ve gone with the rather bold conjecture that every patent will eventually be granted, and we can assume an average of 2.5 office actions per patent application to get a lower bound for the total cost. This is a generous assumption when it comes to cost and time.
That means half of your applications will require a RCE fee of $544, which is another $27,200. And there will be 250 responses that you need to draft and submit to the patent office.
Finally, each of those notices of allowance results in a utility issue fee of $480, so that’s another $48,000.
Time to add it up: $48,000 + $27,200 + $66,400 = $141,600.
But we are not done yet — once your patent is granted, you have to pay fees to keep it alive. I know, the patent office just keeps taking and taking!
Later on, I am going to assume a 10 year period is required to determine if a particular invention really requires patent protection, and so we should only add the first two maintenance fees.
So that is ($800 + $1504) * 100 = $230,400. Ouch.
Our final cost for obtaining patent protection of the one hundred inventions is $230,400 + $141,600 = $372,000.
Time To Respond
Furthermore, given that this is a thought experiment, you get to be a genius at assessing office actions, writing responses, and navigating the administrative requirements for all this patent work, taking a mere 4 hours per response. So you will also be spending 100 * 2.5 * 4 = 1250 hours of your time. That’s 125 working days of eight hours per day.
To put all this in context — I currently hold sixteen patents, mainly in blockchain, but with a couple of e-cig ones too. I filed about 30 applications over five years and easily spent 125 working days on those. As I said, this is an optimistic thought experiment.
A Patent Without Infringement Is Not Worth Much
The problem is, you don’t know in advance which of your inventions is actually going to be of commercial interest, and hence likely to be infringed, and which ones are going to turn out to be irrelevant in the marketplace.
Ideally, you’d like to have a crystal ball that allows you to look into the future, and find out that your new lawn mower and the computer database are viable products, and the rest are not. That way you would only have to file and then prosecute two applications. Which would comprise about three days of work and $4520 in fees.
But alas, there is no such crystal ball.
Enter The Continuation
The interesting thing about the patent system of the United States is that it includes something called a “continuation”. This allows you to file a patent application with claims (the things that you want to be protected by your eventually-granted patent), and then file another application just before your initial application is granted with new, tweaked claims.
As long as the new claims are supported in the specification of the application, the continuation is examined as a new patent application with the priority date of the original application.
And you can file continuations of continuations of continuations, ad nauseum.
Priority Dating
So what is a priority date? It is the day from which your patent is considered to be valid, and is set by the day on which you file your application.
If it later turns out that a company manufactured your invention or an academic wrote a paper about it before the priority date, then your patent is invalid. And you get to feel annoyed that neither you nor the patent examiner spotted that (and no, you don’t get a refund of the patent fees).
But if a company manufactures and sells your invention after the priority date, i.e. some time after you submitted your invention to the patent office, you can sue them for infringement and stand a chance of winning damages. Provided your patent is granted, even if that happens after they started producing and selling a product based on your idea.
This is why you can’t just put the descriptions of your inventions in a drawer and fish each one out when it appears on the market, and then file a patent application for it — the priority date will end up being after the product was being produced by someone else.
Back To Continuations
Here is an idea — what if you fold the specifications for all 100 inventions into one application? It is going to be a big application, and there is a fee for that. Let’s assume that each invention takes six pages to describe. That means the application will be about 600 pages, and the fee is $168 per fifty pages of application material beyond 100 pages.
Oh dear, that is $1680 on top of the $664 filing fee mentioned above.
And of course you have to remember to include the draft claims for each invention within the description, because any claims added later on must be supported in the specification. This is standard practice when it comes to drafting patent applications in any case.
Then you file this monster of an application, with separate claims for the first invention only. That gets you a great priority date for all the inventions described. And instead of a hundred applications, you just have the one.
Ready, Steady, Go
The clock starts ticking. For the purpose of our thought experiment, we will assume that the USPTO takes sixteen months to issue the first office action, and two months for each subsequent office action. Of course, the trick here is to select an invention that is likely to be assessed by an Art Unit that takes a long time to get around to an invention. For example, I had a patent application that took over three years to be examined.
You can delay the process of examination by taking your time to respond to the office action. Typically you have two months before fees start to accumulate.
But at a certain point, it becomes preferential to use a request for continued examination:
Finally, if none of your ideas are exploited by someone else within 10 years, then you are not the inventor you thought you were, and you should probably give up.
So how much would it cost to use continuations to keep your monster application alive for a decade?
Clearly, waiting more than a month past the first deadline sees fees mounting up pretty quickly, so you really should not be looking for extensions beyond the first month. And so I ran a couple of calculations in a spreadsheet to see whether using the one-month extension at $88 was worth it or not. And it turns out that the cost is about the same.
A ten year delay through eight RCEs, or through seven RCEs by paying for an extra month both cost about $7000. This does mean that you have to provide a convincing enough response to each office action that causes the examiner to spend time on rebutting it without actually getting an allowance.
If you do get an allowance, you have to make a judgement call. Pay the issue fee, and then subsequent maintenance fees, and make sure you file for a continuation on one of the other inventions before the patent is issued. Or perhaps not — let it lapse, and file a continuation on some other invention in the application. After all, if it turns out later that the patent you never claimed has become valuable, you can file a continuation on that too, with slightly tweaked claims, and presumably it will be allowed as well.
Now there is one problem with filing a continuation rather than an RCE, and that is that you have to pay the excess pages fee all over again. However, that works out at only $1680.
And so we have a worst-case scenario, in which you have to file a continuation after each allowance, because the damned examiner keeps agreeing that your invention under investigation is novel and non-obvious.
Except, as worst-case scenarios go, it really isn’t that bad. My spreadsheet calculations show that this results in $15,120 in processing fees, and another $10,112 in maintenance fees, for a total of $25,232. Which also gives you seven granted patents.
Finally, we have the cost of prosecuting those two patents we really needed in hindsight. That’s two continuations, and a couple of RCEs (because we are not going to delay our responses for those).
And that’s 2 * ($1,680 + $800) = $4,960
Summing this total, for the worst case we get $4,960 + $25,232 = $30,192.
Conclusion
With the assumptions I made, the cost of pursuing every invention is:
$372,000
And the cost of using continuations and then getting two patents that are actually valuable is:
$30,192
That is a saving of over a third of a million dollars.
Predicated by only 2% of your inventions being truly desirable.
Of course, the savings will be different if it turns out that none of your inventions were valuable (and the saving from that is tens of thousands of dollars larger), or much smaller if it turns out most or all of your inventions were valuable.
In fact, in the unlikely event that all 100 inventions are commercially beneficial, then you will have wasted about $200,000 because of all those excess pages fees you had to pay for each continuation. I should really take the time to work out where the break-even point is, but I suspect it’s at about half of your inventions being significant enough to warrant an infringement case.
Or, you can sell the infringed patent, because patents with an infringer are far more valuable than those without one.
An so, provided even just a few of those cases are ultimately profitable or you make a couple of patent sales, the extra costs should be small change.
And you’ve hedged your bets on which inventions to pursue.
