Startup patent strategy: the mistakes you don't get a second chance to fix

Key facts

• Europe has no general grace period for novelty: anything made available to the public before the filing date counts against the application. Source: European Patent Office, Official Journal 1999, p155
• The US allows a 12-month grace period for the inventor's own disclosures under 35 U.S.C. 102(b)(1). Source: Mondaq, The impact of public disclosures on patentability
• At FlatFrog, none of the top five filed patents were the ones competitors infringed; they worked around what the team thought was core. Source: Sifted, 17 September 2026
• Ola Wassvik's rule of thumb is that a regular startup spends no more than about 2% of its total budget on patents and IP, while businesses where the IP is the product spend as much as they can afford. Source: Sifted, 17 September 2026

A startup patent strategy usually fails in six ways: relying on one patent, protecting one corner of the product, trusting an NDA instead of filing, having no multi-year budget, filing after public disclosure, and stopping after the first filing. Fix those six and your portfolio can survive pivots, copycats and due diligence.

I've filed over 40 patents myself and helped manage more than 300 patent families, most of them at FlatFrog. I've also got IP strategy wrong more times than I like to admit. These six cost me the most.

One patent isn't a patent strategy

Our first patent at FlatFrog took months. Clear claims, clean drawings, exactly how I wanted it. Then we got scooped by someone who filed something broader.

It took me about ten years to really see it. We brought out a technology, a competitor copied it, and none of our top five patents were in their way. They worked around what we thought was core.

One patent is a lottery ticket. File fast, file broad, narrow down later. The full framework is in our guide to patent strategy for growing tech companies.

All our IP sat in one corner

We put strong patents on one piece of our touch-sensing technology. Then a key piece of the technology leaked. A competitor rebuilt our system from scratch, and we didn't have the protection we thought we did since we had focused too much on one aspect. Luckily we had filed some patents in all areas and could double down there.

Cover the concept, the system and the component. Then ask what the worst patent a competitor could file against you would be, and file that one first. That is what a patent portfolio strategy looks like in practice.

We trusted an NDA instead of filing before the meeting

We pitched to a massive global brand under NDA. We showed the technology, the approach, the diagrams. Almost immediately, they started filing patents on our ideas.

We could have sued for breach of NDA and tried to get our IP back, but that would have been a very expensive exercise.

An NDA is a contract. A patent is a right. File before you disclose, especially under NDA. Whether to patent or keep something secret is a separate call, and our guide to trade secret vs patent walks through it.

We had no budget for keeping patents alive

Dozens of applications. It looked impressive. Then the cost of keeping them alive arrived. Some went public and never got granted. Others we abandoned when budgets got tight, and valuable ideas went straight into the public domain.

My rule of thumb for a regular startup is around 0.5-2% of total spend, rising as investment and revenue arrive. Businesses where the patents are the product spend much more. Budget five years ahead before you file the first application. One solid patent, followed through, beats ten you abandon in year three.

We waited too long to file a patent

Build first, patent later. That was the plan. By the time later came, demos were out, slides were shared and features were visible. Parts of the invention were no longer patentable in several countries.

Public disclosure kills novelty. Europe has no general grace period. The US gives you 12 months. Don't plan around either. File when you can describe the idea in one sentence. Our guide on when to file a patent shows where that line sits.

We never planned the second move

We got decent at filing. We never planned what came after. No continuations. No divisionals. We protected the first version of the product and nothing after it.

A patent strategy isn't a checklist. It's chess. You file for today and set up tomorrow's moves. The trade-offs between licensing and building a portfolio are in Patent licensing vs. portfolio: the 10:1+ rule.

How Lightbringer handles this

All six mistakes came from the same place. Strategy lived in a document, and cost lived in somebody else's invoice. When an investor told us to file 100 patents, the old way stopped working completely.

So we built Lightbringer to fix that. Lightbringer is the AI-native patent service for tech companies: patents drafted with purpose-built AI, reviewed and filed by Lightbringer's own patent attorneys, for one flat fee per application.

A flat fee means you can afford to file broad, because you know what it costs before you start. Official office fees are separate, and we say so up front. Attorney review means the AI never files alone. Your strategy sits in a live view of your portfolio, your competitive landscape and your invention pipeline, so the second move is on screen instead of in your head.

We work with more than 200 deep tech companies across the US and EU. Get started for free.

Most founders don't get a second chance with IP. I got several. You probably won't.

FAQ

When should a startup file its first patent?

Before any public disclosure, including demos, decks, pitches and launches. Europe has no general grace period for novelty, and the US allows 12 months. File as soon as you can describe the idea in one sentence.

Does an NDA stop someone else from patenting my idea?

No. An NDA is a contract, not a patent right, so if the other party files first you are left enforcing a contract. A narrow exception exists in Europe for disclosures made in breach of confidence, but it is not a plan. File before you disclose.

How much should a startup spend on patents?

Ola Wassvik's rule of thumb is around 2% of total spend for a regular startup, rising as investment and revenue grow. Businesses where the patents are the product spend much more. Budget five years of costs before filing the first application.

How does Lightbringer's pricing work?

Lightbringer charges one flat fee per application. The Starter plan is priced per application per year until a decision on grant, with official fees charged separately. The Growth plan, for five or more applications a year, is available on request.

DISCLAIMER: THIS IS NOT LEGAL ADVICE. YOU SHOULD CONTACT A PATENT ATTORNEY IF YOU NEED A FORMAL ASSESSMENT.