Nobody needs to copy you to beat you.

Deep tech companies spend an extraordinary amount of time protecting what they've built. Patents filed. IP strategies developed. Freedom to operate analyses commissioned. Trade secrets guarded. Lawyers engaged. All for good reason. If you've spent five years and millions of dollars building something genuinely novel, the last thing you want is somebody copying it.

But there's a question that gets asked far less often. What happens if nobody needs to copy you to beat you?

A patent can stop a competitor replicating a particular invention. It cannot stop them solving the same customer problem another way. It cannot make a procurement team choose you. And it cannot guarantee that when your market thinks about the problem you solve, they think about you.

To be clear, there are markets where the patent is the business. In pharmaceuticals the molecule is the asset and the patent is the entire commercial position. In licensing models, the portfolio is the revenue. At acquisition, the IP is often what gets valued.

If you're in one of those, file everything and defend it hard.

But most deep tech companies are not selling a molecule or licensing a portfolio. They're selling a capability to a buyer who has options. And in that market, the patent protects the invention while leaving the position completely undefended.

When I talk about brand in deep tech, I don't mean logos, colours or advertising. I mean the position a company occupies in the mind of its market. The association between a problem and a name.

Those associations are valuable precisely because competitors cannot engineer around them. You can build a technically superior product and still lose to the company that has become the default answer to the question your buyer is asking.

Which raises an uncomfortable question. When someone in your market describes the problem you solve, do they say your name?

If the answer is no, your technical moat may be considerably stronger than your commercial one.

I've sat through industry pitch days at a few conferences this year. The pattern is consistent enough to be a bit uncomfortable. Almost every founder opens with the technology. The architecture, the performance, the novelty, the IP. It's usually genuinely impressive. Almost none of them open with the problem.

And more often than not, the company has a name that tells you nothing. A generic compound of two technical words, or a Greek god, or something that could just as easily belong to a fund, a fintech or a fusion startup.

So the room hears a brilliant answer to a question nobody has stated, from a company whose name gives them nothing to hang it on. Then they move on to the next pitch and do it again.

That's not a technology problem. Every one of those companies is defensible on paper. It's a translation problem. And it's the reason the association ends up forming around somebody else.

I call the distance between those two things the Commercial Translation Gap. It's the gap between the value embedded in the technology and the value understood by the market.

Deep tech companies assume technical defensibility will become commercial defensibility on its own. It doesn't.

A company can have world class engineering, genuinely differentiated IP and an impressive portfolio, and still struggle to explain why any of it matters to a customer, investor, prime or government buyer.

The wider that gap gets, the more room you leave for somebody else to occupy the commercial position your technology should have earned you.

That competitor may have weaker technology and fewer patents. But if the market understands what they do and where they fit before it understands you, the association starts forming around them. Once it sets, it's very hard to dislodge.

Your competitor doesn't need to copy your technology to beat you. They need to become the answer before you do. You can't patent your way out of that.

There's an assumption that brand matters less as products get more technical. I think the opposite is true.

In defence, aerospace, advanced manufacturing and critical infrastructure, the cost of a wrong purchasing decision is enormous. So the buyer isn't only evaluating technical performance. They're evaluating whether you can deliver, whether you'll integrate, whether your supply chain is secure, and whether you'll still exist in ten years. And underneath all of that sits something more human.

Somebody inside that organisation has to put your name in front of their boss. That person is staking their own judgement on you. Nobody gets promoted for backing an unknown supplier that fails.

Brand is what makes that internal advocacy survivable. It lowers the personal cost of choosing you.

It's also why brand is most valuable in exactly the markets where technical founders are most likely to dismiss it.

So, a patent or a brand?

If you're licensing IP or selling a molecule, take the patent. Everywhere else, the patent alone is not enough, and the strongest companies stop choosing.

Technical IP creates barriers to replication. Brand creates barriers to substitution.

The Commercial Translation Gap determines whether the first ever produces the second. Because technical superiority that nobody understands is not a commercial advantage yet. It's potential.

Ask your last three buyers to describe the problem you solve, in their words.

Listen for whose name comes out of their mouth.

If they need your deck in front of them to answer, you have one moat, not two. And while you're busy protecting the invention, somebody else is quietly getting on with owning the problem.