Before filing: how founders approach their first patent

Thrust is a startup developing bladeless propulsion technology for e-hydrofoils. Founders Tom and Pete both work in the drone industry. Avid surfers, they have more recently become interested in hydrofoiling, fitting their boards with foils that lift the board with increasing speed for a ride that is flight-like. As they spend their days engineering drone control systems for flight it seemed a matter of time their attention whether they could leverage those skills into hydrofoiling.

Bladeless ring drives is the foundation what they settled on. It's not actually used in drones, but it's an existing form of propulsion that uses all their existing control and coding expertise. So the leap is fairly modest, directly an machine through an unstable fluid steady with sensors, thrust and software.

Yes, it is attention-grabbing, but more than that they realised that engineering some directionality to the drives and using three not just one opened up an some amazing possibilities that excited them.

The result is Thrust, an electric hydrofoil board with no conventional propeller. Three bladeless ring drives sit in the foil: a large central drive and two smaller thrusters at the wing edges.

Each drive can direct its thrust within a modest but effective cone of movement. So as well as varying thrust levels across ring drives, there are differential and directional options which opens up completely different riding experiences in terms of buffering any turbulence, as well as either accentuating or moderating certain ride dynamics.

They have self-funded everything themselves to this point: prototypes, testing and tooling. The next step is production, and this needs outside funding.

They describe Thrust in one sentence: “A board that gets you up easily, corners on rails, and offers endless adjustment whether you're beginner or pro.”

Prototype testing and feedback are promising, but what are they left with if Thrust is a hit and competitors simply adopt the same technology?

“Patent pending” seems to be the answer: first to protect the product, and second to attract much-needed funding. But currently funds are limited. They have done so much work and solved so many problems that it’s not even clear exactly what they should be patenting. While they’re very fluent in the product, it’s hard to articulate which part is the critical key they should be protecting.

Two imperatives drive the push to patent. The first is the launch: anything shown publicly before a patent application is filed can count against patent validity. The second arrives sooner: investors will want to know what patent rights Tom and Pete hold before they commit. A great prototype alone means funding can be doubtful. Oonce the concept is proven and presented, the barriers to copying are relatively low for competitors well placed to exploit this new development. So Thrust faces the dilemma many founders face: they need protection to raise money, and they need money to afford protection. This means being very pragmatic about securing the most effective protection without incurring needless expense.

One product, four perspectives

The founders

The founders both engineers naturally lean technology-first, and that has been their preoccupation in developing Thrust. They have been consumed with developing ring drives that are responsive, outlets that steer the thrust, and their control strategies that coordinates them for seamless flow.

The customers

Customers, however, determine the future of Thrust. What they see and value is the novel experience of getting up on a foil with ease, carving through exciting turns, and no propeller fear near swimmers. Perhaps they value investing in a board that matches their skill as they develop. And of course a board that looks really cool.

The investors

Investors ask about exclusivity: who could build something similar, how quickly, and how much they could learn by buying a board and taking it apart. They're not only concerned about revenue but second-order effects in terms of ingress from competitors.

And there’s the product itself: what ships at launch, what stays on the roadmap, and what the launch reveals to the world.

The competitors

Competitors, for their part, are sure to take notice. Their first thought is what this means for them. What ideas or technology can they borrow? Will they tweak, or even radically adjust, their roadmap? Will their positioning or direction change? What will they learn about broader customer preferences or unmet needs?

What to do?

Most first-patent mistakes come from deciding through only one perspective. Engineers want to patent the cleverest technology. In a vacuum this looks like the winning choice as it is after all possibly the most patent-worthy and what the engineering team's proudest work.

Commercial instincts pull toward protecting the features that sell, which makes sense from a marketing perspective. But those features can be the easiest to design around or substitute. And often the hardest to patent.

Often, nobody asks about the MVC ('minimum viable copy') what a competitor would actually be motivated to copy with a demonstrated demand for the new product. This is often the lean version of the winning prodiuct: they see what customers are actually responding to, and look to serve that need at lower cost.

Lastly, investors, and the immediate hurdle. What do investors want to see in terms of protection? Investors at least are usually open to being convinced that you have a good read on the other perspectives.

The challenge is to triage the competing concerns and find a balance that meets immediate needs while mitigating the most likely mid-term contingencies. You have to live with the decision for a while, but the immediate hurdle is funding, so one factor is coming to an position that is easily presented as compelling to investors.

Realistically, it’s only pragmatic to fund a single patent at the pre-launch stage. First, the product, while promising, is unproven. Second, multiple patents risk over-investing in protection that turns out to be entirely unnecessary. Usually it is best to settle on a single patent that is centered around the features that will survive product-market fit, and being around in a few product iterations. Reaching an informed decision that integrates all four perspectives at once is what the rest of this guide sets out to do.

Taking inventory

The first step is to get everything out of Tom and Pete’s heads and onto one page. That means every new feature in Thrust, including ideas that aren’t finished yet but are likely to matter within a year or two.

Their list runs to more than a dozen items. Each is then rated against ten factors spanning all four perspectives: near-term commercial importance, long-term competitive value, the likelihood of competitors developing it independently, the ease of reverse engineering it from a purchased board, whether copying could be detected, whether it’s a genuine roadblock for competitors, whether customers actually notice it, how central it is to future development, its brand value, and its licensing potential. Each factor is scored out of ten, which gives every feature a score out of 100.

The result is a short list that looks like this.

The surprise for Tom and Pete is that the feature they’re proudest of, the bladeless drive itself, scores poorly. Ring-driven and ducted thrusters already exist in marine engineering. What’s new is what they’ve done with the drive: steering its thrust, combining three of them, and coordinating them in software. While they have limitless options to operate this arrangement in software, what they're looking to protect is probably the control architecture iteself rather than any specific operation.

The second surprise is how neatly the green and amber features cluster. The steerable drive, the take-off assist, the three-drive layout, the control method and the low-speed manoeuvring are all parts of one system. That’s good news on a tight budget, because one well-drafted application can cover this whole suite of features.

Where the perspectives disagree

The scores don’t settle everything. They show where the four perspectives pull in different directions, and those tensions are where the real decisions sit. The idea is more to use this tooling to navigate the confusion rather than let the scoring govern dictate the eventual decisions.

Take-off assist is where everything aligns. Getting up on a foil is the biggest barrier for new riders, so customers want it. It depends on the steerable drive, so it’s likely to be protectable. And it opens a beginner market that propeller-driven boards struggle to serve. This is where the first dollar should go.

Cornering is the opposite case. It’s the feature that wins demonstrations, but on its own it’s a weaker patent position, and a competitor can deliver a similar feel in other ways. Its value is better captured through brand and riding experience than through claims.

Low-speed manoeuvring barely registers with recreational riders. But holding position in a current or chop is something that may be compelling. It costs little to describe it in the application now, and it may be what makes Thrust’s technology licensable later. This is the feature that separates protecting a product from building a position.

Steering by varying power alone is the principle behind every twin-engine boat. Patenting may be possible owing to the unique factors of the foil, but this is probably lower priority. Money spent now trying to patent this feature is money not spent on what matters.

Importance of patent searching

Patenting does not occur in a vacuum. Your patent rights can only begin where the existing technology stops. Before drafting, a searching existing patents and publications tests the scores against reality. For Thrust, it confirms what the inventory suggested. Rim-driven thrusters are well documented. Steering by directing thrust is familiar from jet boats. Several patents cover ways of stabilising electric foils.

Patent searching does not give a binary result. Patent searching shapes what can be claimed. Very broad and generic claims to “a bladeless drive for a hydrofoil” or “steering a board by directing thrust” would run straight into earlier published material. A claim to a hydrofoil drive whose outlet is divided into discrete drive units, with steering achieved by varying the flow between them, is on much firmer ground and unlikely to be demonstrated within existing publications.

What “patented” really means

This is where many founders discover what a patent actually does. The claims are the fence around the invention, not a description of the product. A patent is a right to stop others using what the claims cover. It isn’t permission to sell your own product. “Patent pending” stops no one on its own, though it tells competitors something is coming. And a competitor who finds a way around the claims is free to use it.

That’s why the claims are layered. The broadest claim covers any steerable ring drive of this type on a foil, whether there’s one drive or four. Narrower claims cover the three-drive layout, the take-off assist, the banked cornering and the control method. If an examiner or a competitor finds earlier material that defeats the broad claim, the narrower ones still stand.

The 45-degree cone is a good example of the discipline this takes. Tom and Pete are proud of it, and it belongs in the application. But it belongs in the narrower claims, not the broadest one. If a competitor achieves 50 degrees, or Thrust’s own next version does, a broad claim limited to 45 degrees would catch neither.

Patent, trade secret or publish

Patenting should be selective by design. It is not costless, and some of that cost is not immediately obvious: is is publication. Not everything valuable should be patented. A patent application is published around 18 months after filing, and everything in it becomes public. So the question for each feature is whether competitors could work it out anyway. As an example, certain manufacturing know-how may be invaluable, but if you can keep it secret this is typically the better choice.

The hardware answers itself. Anyone who buys a Thrust board can see the three drives and the segmented outlets. Keeping them secret isn’t an option, so they’re patent candidates.

The control tuning is different. How much each drive responds, how quickly, and in what combination for each riding situation is invisible from outside. It’s the product of hundreds of hours of testing, and a competitor copying the hardware would still have to find it for themselves. You might think trade secrets are preferable, but patenting can be strong option as competitors could reverse engineer this work easily enough.

A trade secret only lasts while it’s treated as one. That means confidentiality agreements with contractors and testers, restricted access to the control code, and a written record of what the secret actually is. Investors will ask about these measures, because an unprotected secret isn’t an asset.

The choice is also one-way. Anything included in the patent specification can never be a secret again, so the tuning has to stay out of it. And a feature that isn’t worth patenting, but might be patented by someone else, can simply be published, putting it on the public record so no one else can claim it.

Pre-revenue dilemma: deciding under uncertainty

This is the hard part. Thrurst has limited money and limited market research. They don’t yet know which features customers will value most, which markets will buy, or what investors will make of it. Yet the decisions they make now will shape their position for the early crucial years of their business.

The way through is to separate decisions that can be changed later from those that can’t.

Decisions that can’t be undone need consideration now. What the launch reveals, once it’s public, stays public. What goes into the first application sets the limits of everything that follows, because new material can’t be added later without losing the original filing date. A feature patented can never become a secret. And ownership needs to be settled: the inventions must belong to the company, not to Tom and Pete personally, and certainly not to their drone employer. Because both work in the drone industry, their employment contracts need checking for clauses giving the employer rights in related inventions. Investors will look at exactly this in due diligence.

Decisions that can wait should wait. How narrow the claims need to be, which countries to file in, whether to file further applications, and even whether to pursue the patent at all can all be decided later, when there’s more information and, ideally, more money.

Spend on the quality of the first application, and describe everything that can be reasonably disclosed. That breadth is what gives the claims room to move later. Be frugal about everything else until the funding round and first sales provide real information.

Timeline and milestones

The patent process works in stages, and the costs follow the same rhythm. The stages also line up well with a startup’s own milestones.

Month 0: provisional application. This establishes the filing date. The main cost is the drafting, and it’s the most important money spent in the whole process. For Thrust, the filing happens before investor pitches and well before launch.

Months 1 to 12: fundraising, testing and launch. Thrust can now pitch, demonstrate and launch without losing rights to what the application describes. Improvements found in testing can be captured in further provisional applications, then combined later.

Month 12: the first major decision. Within a year, the provisional must be followed by a full application. Most startups choose an international application under the Patent Cooperation Treaty (PCT), which keeps options open worldwide. This is a moderate cost, and by now Thrust should have its funding round closed and early sales data.

Month 18: publication. The application becomes public.

Months 30: the largest decision. The international application enters the national phase in each chosen country. This is where most of the cost of a patent program sits, because it multiplies by the number of countries, with translations and local attorney fees in many of them. It’s also the point at which Thrust will know the most: which markets are buying, where competitors are appearing, and what investors want.

Examination and grant. Each country then examines the application, typically over several years, with costs spread across that period.

Spending is light at the start, when uncertainty is highest, and heaviest at 30 months, when the business has the most information and, ideally, the most funding. A good first patent strategy is designed around that curve.

What a first engagement looks like

Thrust’s path starts with a conversation. A first meeting is the opportunity to work through the product using all four perspectives: technology, customers, competitors, investors. And the product roadmap. That leaves room to identify what is new, most what’s worth protecting, and how.