Studio
“So you changed niche again?” is a fair thing to ask a studio that described its market one way at the start of the summer and another way by the end of it. Yes, we did. And we did it while it was still cheap to change. But the same decision also made the market harder to get into. The fields went from one to five, and the test a company has to pass got stricter, because it now asks about the problem a company has, which an industry label never told us.
The products looked alike before anyone explained why
When I started looking for companies to write to, Mistaek was pointed at nature tech, so I spent a lot of evenings looking for companies, opening their websites and trying to find their product pages. After a while they started to blur together (which probably says as much about how many I opened in one sitting as it does about the pages). Similar maps, similar layers, and similar claims about turning environmental data into better decisions.
A founder on one of our research calls explained why. In his part of the field, a large share of the data comes from the same public sources, and the technical users already know those datasets well. Getting a project started is expensive, the people at the earliest stage of a project rarely have budget to spend, and once a company raises and has to step up, there isn’t much left to stand apart on. Most of the teams he competed with, he said, had moved toward consultancy.
That says nothing bad about the people building those products. It’s hard work in a market where the money arrives late, and I’d expect anyone working from the same inputs under the same pressure to end up close to their neighbours.
The sameness started in the data, where design can’t reach
For a design studio, that was an awkward thing to learn. We had been reading the sameness as a design problem, something a sharper brand or a clearer interface could solve. It started earlier, in the data and in the money, and if we had walked in promising to make one of those products stand out, we’d have been promising something the work couldn’t deliver.
Knowing a company’s industry told us what it worked on. Whether it had the problem Mistaek is built for was a separate question, and the label couldn’t answer it.
The problem we chose has three conditions
So we stopped filtering by field and wrote the problem down instead. A company fits when the technology is hard to understand from outside, when the people who buy it aren’t the people who use it, and when the decisions that shape the company’s future are made by people without direct access to the science.
Those three conditions cut across fields. A quantum hardware team, a diagnostics platform and a satellite data company can all meet them, and in each of them, we believe, the product ends up doing some of the explaining: the investor without the domain background, the procurement committee and the operator who has to trust an output before they understand how it was made all decide partly on what they can see. Nature tech stays in, wherever a company meets the three conditions.
This is a targeting decision, and we treat it like one. It tells us where to look and what to ask on a first call, and it doesn’t prove anything on its own.
What we still have to prove, and what would change our minds
We have no frontier tech cases yet. The cases on our site come from banking and mobility: a design system for a regulated data that teams in other countries chose to adopt without being told to, and a direction worked out with a company while its brief was still open. What carries across is the complex data, the high stakes and the work of getting people who didn’t build something to rely on it.
Before we sent the first email in the new market, we also wrote down when we’d look at the decision again: after three months of steady outreach, or sooner if founders tell us, in real conversations, that the problem isn’t worth paying for. If replies are slow in the meantime, we take it as a sign that we’re asking badly, and we fix the asking before we touch the market. The time between projects goes into the parts a founder can check before talking to us: the first Foundations memo is live, and the Method page lays out Root Key and Standing Decisions, the two tools both programs run on.
A description of a market that nobody can check is one more thing asking to be believe, so we published the condition that could prove ours wrong. If your company meets the three conditions, book an intro call and tell us what the people deciding on it see first.
