Patent documentation, honestly.
Also called: IP filing · Patent drafting · Design registration · IP protection record
Writing down what is genuinely novel about your product, then choosing which protection is worth filing for and which is just expensive theatre.
Most products have less protectable invention than the founder hopes. Be honest about what is actually novel, then file the cheap, real protection (a registered design, a trademark) and skip the patent unless a true invention survives scrutiny. Brand and execution usually protect you more than a filing does.
What patent documentation is
Patent documentation is the work of recording, in writing, what you have actually invented, and then deciding which of the formal protections is worth paying for. It sits at the build stage because by now you know what the product really is, not what the pitch deck hoped it would be. That gap matters, because most of what feels proprietary turns out to be the obvious application of known engineering.
There are three instruments, and they protect very different things. A patent protects a genuinely novel and non-obvious technical invention, the underlying mechanism. A registered design protects the specific look of the product, its shape, surface and ornamentation. A trademark protects the name and the badge you sell it under. Founders reach for the patent first because it sounds the most serious. In my experience it is usually the one they least qualify for and can least afford to defend.
The honest test for novelty
- Has it been done before? If a competent engineer in the field could arrive at your solution from what already exists, it is obvious, and obvious things are not patentable. Heating a box to a steady temperature is not new.
- Could you defend it? A granted patent you cannot afford to litigate is a certificate, not a shield. A patent costs thousands to file and tens of thousands to enforce against a determined copier.
- Is the real moat somewhere else? For most consumer products the durable advantage is the brand, the supply chain and the execution, not a clause in a filing. Be suspicious of any protection plan that ignores those.
A worked IP filing
The clearest way to think about this is to fill the boxes honestly for a real product. Here is where the proofing box we ran through the framework actually landed, so you can see the shape of an honest answer rather than a wish list.
Notice the patent row is the weakest, and we walked away from it deliberately. The protection that earns its place is the design registration and the trademark, plus the unglamorous moat of supply chain and compliance that no filing captures.
- Filing on a mechanism that is obvious engineering.
- Spending thousands on a claim you could never afford to enforce.
- Telling investors “patent pending” to sound defensible.
- Treating a filing as the moat and ignoring brand and supply chain.
- Register the distinctive design; it is cheap and genuinely yours.
- Trademark the name and badge so the brand can compound.
- Document the novelty honestly, even when the honest answer is “little”.
- Treat execution, supply chain and compliance as the real moat.
How it fits the bigger picture
Patent documentation is activity 07.10.08 in the framework, near the end of Stage 07 Engineer. It builds on the earlier IP-landscape work in Discover and feeds into industry standards (07.10.09), where the compliance evidence that often protects you more than a patent gets assembled.
What it can do
It forces an honest inventory of what you really own, and points the limited IP budget at the protections that are cheap, defensible and genuinely yours. It stops you spending a patent’s worth of money on a claim you could not defend, and names the brand and execution moat for what it is.
What it can’t do
It can’t manufacture novelty that isn’t there. If the mechanism is obvious, no amount of documentation makes it patentable, and a registered design will not stop a competitor selling a similar product that looks different. The real defence is being faster, better made and better known, which no filing guarantees.
See the full 10-stage process →
Try it yourself
Write one sentence describing what is genuinely new about your product, with no marketing words. Then ask a competent engineer in the field whether they could have got there from what already exists. If they could, you have a registered design and a trademark to file, not a patent. Spend the budget where it is defensible, and write down honestly what protects you that no filing can.
Want a structured first pass at your IP position? Start the Free Sprint → and the GPT will help you separate the real protection from the theatre.
Your IP-documentation checklist
Project notes: the patent we didn’t file
▸ From the notebook · optional reading
Dan Hartley wanted a patent. Talking him out of it, in Stockport, was the most useful hour of the build stage.
3 min read · click to open
Dan and Anna had reached the build stage and Dan was set on a patent. “It heats the dough perfectly, nobody does that, we should lock it down.” I asked the awkward question first: what, exactly, is the invention? We sat in their Stockport kitchen and tried to write it down in one sentence.
What we could honestly claim
The honest sentence was uncomfortable. A low-voltage heater, a thermostat and an insulated ceramic shell holding 26°C. Every part of that is established engineering. Any competent product engineer could arrive at it from what already exists, which is the definition of obvious, and obvious things do not get a defensible patent.
I was blunt about the money too. A filing runs into thousands, and a granted patent you cannot afford to defend in court is a framed certificate. For a bootstrapped product on a £38–55 bill of materials, spending a patent’s worth of cash on a claim that thin made no sense.
Where the real protection was
So we redirected the budget. We registered the design, because the ceramic-and-wood look really is distinctive and ours, and that registration was cheap and has teeth. We trademarked the name, because the brand is what one baker tells another, and brand compounds. And we wrote down, plainly, what actually protects them: the Stoke-on-Trent ceramic relationship, the UKCA and BS EN 61010 compliance work, and the economics. A copier has to rebuild all of it.
Dan was deflated for about a day. Then a competitor’s near-copy turned up online a year later, looking different enough to dodge the design registration, and he understood. The patent would not have stopped them. Being better made, better known and first did. The hour we spent not filing saved him several thousand pounds and a false sense of safety.
— Engineer stage, project notes, 2026
— Next in Engineer → Industry standards