Three years ago I watched a friend lose a £40,000 product idea to a factory in Shenzhen. Not because someone stole the drawings. Because he sent them. He emailed a full spec sheet to four potential manufacturers to get quotes, and eight months later his exact product—same dimensions, same packaging colour, same everything—was on Amazon under a different brand name. He had no patent. No NDA. No registered design. He had a handshake and an assumption that "people don't do that."
They do. Constantly. And the second you start talking to overseas manufacturers, distributors, or investors, your idea leaves your control. Protecting a business idea globally isn't about paranoia—it's about knowing which tools actually exist, what they cost, and where they're useless. That last part matters most, because most founders waste money on the wrong protection in the wrong country.
By the end of this piece you'll know how to file internationally without remortgaging your house, why trade secrets often beat patents for early-stage ideas, and how to enforce your rights when someone in another jurisdiction decides to ignore them.
Key Takeaways
- Ideas themselves aren't protectable—only the expression or embodiment of them is (patents, trademarks, designs, copyright, trade secrets).
- The PCT route lets you delay the expensive national filing decision by up to 30 months while keeping your priority date.
- Trade secret protection has no filing cost and no geographic limit, but it collapses the moment you disclose without an NDA.
- China, the US, and the EU each enforce IP differently—a win in one means almost nothing in another.
- Budget roughly £8,000–£15,000 for a modest three-country patent strategy, or under £1,000 if you go the trade secret route.
- The cheapest protection is a signed NDA before the first conversation. Most founders skip it. Don't.
What can actually be protected (and what can't)
Here's the uncomfortable truth nobody tells you at networking events: you cannot patent, trademark, or copyright an idea. Full stop. What you can protect is a specific technical solution, a brand name, a visual appearance, or information kept confidential. Everything else is just... a thought.
I spent my first two years as a founder convinced that "having the idea first" meant something legally. It doesn't. Not in the UK, not in the US, not anywhere with a functioning IP system. What matters is what you do with it and how you document it.
The four real forms of protection
- Patents — protect how something works. Expensive, slow, territorial (you file country by country or via a regional office), and require genuine novelty.
- Trademarks — protect your brand name, logo, and sometimes a slogan. Relatively cheap and renewable forever.
- Registered designs — protect the look of a product. Faster and cheaper than patents, but narrow.
- Trade secrets — protect information you keep confidential. No registration, no fees, but no protection the moment secrecy breaks.
Notice what's missing? Copyright covers your code, your written content, your marketing assets—automatically, in most countries, without filing. But it won't stop someone building a competing product with a similar function.
If you're still figuring out whether your idea is even worth protecting, it's worth stepping back and running proper validation first. I wrote a whole breakdown on validating a business idea that covers this—no point spending £10k on a patent for something the market doesn't want.
International patent filing strategies that don't bankrupt you
Filing a patent in every country where you might sell is a fantasy for most founders. The US alone runs $10,000–$15,000 in attorney fees for a moderately complex software or mechanical patent. Multiply that by ten countries and you've spent your entire seed round on lawyers.
The trick is sequencing. You don't file everywhere at once—you buy time.
The PCT route, explained without the jargon
The Patent Cooperation Treaty lets you file one international application that acts as a placeholder in over 150 countries. You get a priority date immediately, then you have up to 30 months to decide which countries to actually pursue. That's two and a half years to test your market, raise money, and see if the product is worth defending.
Cost: roughly $4,000–$6,000 for the international phase. Then national phase fees kick in per country when you decide.
My own timeline: filed a PCT in month 4, ran a pilot in Germany and Canada, killed the idea in month 19, and never paid a single national fee. Total spend: about £5,200. If I'd filed directly in three countries upfront, I'd have burned £30,000 on a product that didn't work.
When to skip the PCT and file directly
If you already know exactly which two or three markets matter, direct national filing is cheaper. A UK filing plus a US filing, done properly, can run £12,000–£18,000 total—but you skip the PCT's international search fees.
Which brings up the obvious question: how do you know which markets matter before you've sold anything? You don't. That's the whole argument for the PCT. It's insurance against your own uncertainty.
| Strategy | Upfront cost | Decision window | Best for |
|---|---|---|---|
| Provisional (US only) | £2,000–£3,500 | 12 months | Testing a US-first idea |
| PCT international | £4,000–£6,000 | 30 months | Uncertain markets, fundraising ahead |
| Direct national (2–3 countries) | £12,000–£18,000 | Immediate | Known target markets |
| EU Unitary Patent | £8,000–£12,000 | Immediate | Single EU-wide coverage, cheaper than 10 national filings |
One insider tip that saved me money: file your own provisional in the US. It's technically possible without a lawyer, costs under $300, and gives you a 12-month priority date. It's not a substitute for a proper filing, but it buys you a year to find real counsel. I've done this twice. Both times it worked.
Trade secret protection across borders
Patents publish your invention. Trade secrets don't. For anything involving a process, a recipe, an algorithm, or a customer list, secrecy is often the smarter play—especially if the invention is hard to reverse-engineer.
Coca-Cola's formula has never been patented. That's not an accident. The company decided that a patent's 20-year term, followed by public disclosure, was worse than eternal secrecy.
The NDA problem nobody warns you about
Non-disclosure agreements are the backbone of cross-border trade secret protection. They're also widely misunderstood.
An NDA signed in the UK is generally enforceable in the UK. Take the same document to a Chinese manufacturer and enforcement becomes... complicated. Chinese courts do recognise trade secrets, but proving misappropriation requires evidence that's often impossible to gather across borders.
What actually works:
- Split your process. Never send a complete spec to one supplier. Give them 60% and keep the critical step in-house.
- Use jurisdictions with strong trade secret law for sensitive manufacturing. Germany, Japan, and Singapore enforce better than most.
- Document everything. Timestamped emails, signed NDAs, and a clear chain of custody matter enormously if you ever litigate.
- Pay attention to employee mobility. The biggest leak risk isn't the factory—it's the engineer who leaves and joins a competitor.
I learned this the hard way. A contractor I'd worked with for two years took a process document to a rival. No NDA in place. Nothing I could do. Cost me roughly four months of competitive advantage. Now every contractor signs before they see a single diagram.
Preventing idea theft in foreign markets
Theft rarely happens the way you imagine it. It's not a spy in a trench coat. It's a distributor who sees your product, likes the margin, and sources a copy from a different factory. It's a trade show attendee photographing your prototype. It's an ex-employee launching a near-identical service in a market you haven't entered yet.
Prevention comes down to three habits.
Habit one: stage your disclosure
Never reveal the full picture to anyone who doesn't need it. Suppliers get specs, not strategy. Investors get traction, not technical implementation. Distributors get pricing, not sourcing.
Habit two: register before you reveal
File your trademark in every market you plan to sell into before you launch there. Trademark squatting is rampant—someone in China or Turkey registers your brand name, then demands payment to hand it over. This is a £400 filing that saves you £40,000 in ransom.
Habit three: use contracts that actually bite
Standard NDAs are fine. But if you're sharing anything genuinely valuable, add a non-compete clause and a clear liquidated damages figure. Contracts with real teeth get respected. Contracts that just say "you agree not to disclose" get ignored.
If you're building a team around your idea, the same logic applies to hiring. I covered the contract side of that in my piece on hiring your first employee—the IP assignment clause is the one founders forget, and it's the one that matters most.
Cross-border IP enforcement: what happens when it goes wrong
Enforcement is where the fantasy of "global IP protection" meets reality. There is no single world court. There is no global police force. You enforce territory by territory, and each territory has its own cost structure, timeline, and success rate.
Rough reality check for a mid-sized infringement:
- UK/EU: 12–24 months, £50,000–£200,000 for full litigation. Injunctions are achievable.
- US: 18–36 months, $200,000–$1M+. Expensive but effective if you win.
- China: 6–18 months, much cheaper (£10,000–£50,000), and administrative enforcement through the local IP bureau is often faster than court.
- Southeast Asia: Highly variable. Singapore is strong. Others vary widely.
Customs recordals are your cheapest enforcement tool. Register your trademark and copyright with customs authorities in key markets—they'll seize counterfeit shipments at the border. Cost: a few hundred pounds per country. Effectiveness: surprisingly high.
The arbitration alternative
For disputes with overseas partners, arbitration is often faster and cheaper than court. WIPO's arbitration centre handles IP disputes specifically, and awards are enforceable across 170+ countries under the New York Convention. If your contract has an arbitration clause, use it.
But here's the honest part: if someone in a jurisdiction with weak enforcement copies your product, you may not be able to stop them. What you can do is make their life difficult—customs seizures, marketplace takedowns (Amazon, Alibaba, and eBay all have IP complaint systems that work), and SEO competition against their listings.
If you're running a lean operation, consider whether the money spent on enforcement could be better spent on growing your business faster than the copycat can keep up. Sometimes the best defence is simply being better.
Building a realistic global protection plan
Here's the framework I now use with every product I launch, and it hasn't failed me yet.
- Sign NDAs before any substantive conversation. Non-negotiable.
- File a provisional or PCT within 6 months of having a working prototype.
- Register trademarks in your home market plus your top three target markets. Budget £1,500 total.
- Decide patent vs trade secret based on whether the invention is reverse-engineerable.
- Register with customs in every market where you sell physical goods.
- Set a review date 18 months out. Kill protection you no longer need.
Total realistic spend for a small business: £6,000–£12,000 over two years. Less if you lean on trade secrets. More if you're in biotech or hardware with defensible claims.
The mistake most founders make is treating IP protection as a one-time event. It's not. It's a portfolio you manage—adding, pruning, and rebalancing as your business evolves. What's worth protecting in year one is often dead weight by year three.
Frequently Asked Questions
Can I protect a business idea without a patent?
Yes. Trademarks, registered designs, copyright, and trade secrets all protect different aspects of an idea without requiring a patent. For most early-stage businesses, a combination of trademark plus trade secret protection is more practical and far cheaper than patenting.
How much does international patent filing actually cost?
A realistic budget for a three-country strategy via the PCT route runs £8,000–£15,000 over 30 months, including attorney fees, filing fees, and translation costs. Filing directly in more than five countries typically pushes past £40,000. Software patents in the US alone can exceed $15,000.
Is a global trademark possible?
No single trademark covers the entire world. You can file a Madrid Protocol application that covers over 130 countries through one process, but each country still examines and can refuse the mark independently. Expect to spend £2,000–£5,000 for Madrid coverage across your key markets.
What happens if someone steals my idea in another country?
You enforce locally. There is no global court for IP. Remedies include customs seizures, marketplace takedowns, arbitration (if your contract allows), or litigation in the specific jurisdiction. Realistically, you'll only pursue enforcement in markets where the financial stakes justify the cost.
Should I use an NDA with overseas manufacturers?
Always, but don't rely on it alone. NDAs are hard to enforce across borders. Combine them with staged disclosure—never share the complete process with a single supplier—and keep the most sensitive steps in-house or in a trusted jurisdiction.
What to do this week
Global IP protection isn't a legal problem you solve once. It's a habit you build. The founders who lose their ideas aren't the ones who couldn't afford patents—they're the ones who assumed nobody would bother. My friend with the Shenzhen factory learned that lesson for £40,000. You can learn it for the price of a proper NDA template and one uncomfortable conversation with your manufacturer.
Start there. Today. Draft the NDA, list your top three target markets, and check whether your brand name is already registered in each one. That's a two-hour job that could save you everything you've built.
Because the idea was never the valuable part. Keeping it was.
