What succeeds in the lab moves out. Why the model is rigged that way, when the step is taken, and what it does to both the lab and the company that leaves.
By Simon Torngren · July 29, 2026
The usual way to handle a successful idea is to keep it. It was built here, it is starting to pay for itself, so why let go? Formise does the opposite. When something shows that it holds, it moves out: its own name, its own site, its own company.
The thesis fits in four words: assets up, risk down. What holds is lifted out of the lab and becomes an asset with its own economics. The risk, everything still being tested and allowed to break, stays down there.
A lab that keeps everything eventually becomes a warehouse of half-finished things nobody maintains. Every retained build draws a little attention, a little maintenance and a little guilt. In the end the time goes to keeping old builds alive instead of testing new ones, and then the lab has stopped being a lab.
So the spinout is built into the model, not decided after the fact. From the moment a build starts, the plan is for it to leave the lab if it holds. That can sound cold, but it is the opposite: a build meant to move out is built from the start to manage on its own. Dependencies are kept few, the technology is owned the whole way, and nothing important lives in one person's head.
The model also makes the lab more honest. When success moves out anyway, there is no reason to keep something alive just because it has already cost money. What does not hold gets to die cheaply and early. Why that difference matters is under Studio.
The move happens when an experiment goes from play to serious. Three signals decide: customers using it for real, revenue you can point at, or outside interest in the build itself. One of them is enough to ask the question. Without any of them the build is still an experiment, and then it stays in the lab.
Before that, the idea has already travelled a path that looks the same every time: explore, define, build, launch. The whole path is on the process page. The spinout is not a fifth step but what happens after the fourth, once the launch has held up over time.
Waiting for the signals is half the protection in the model. A company formed too early carries all the costs of being a company without the proof that justifies them. An experiment allowed to be an experiment costs almost nothing, and can therefore be allowed to be wrong.
The spinout is a separation. What holds is separated from what is still being tested. The company gets its own name, its own site and its own economics, and from that point the goal is that it manages without the lab. The risk stays with the experiments, where it is cheap. The asset moves out, where it can grow without sharing daily life with things that break.
Distribution is the other half of the model: build in the lab, distribute through Memorise's client network, and let what holds become its own company. Formise and Memorise are siblings sharing craft and people, not an org chart, and how that structure fits together is under About.
The separation tidies in both directions. The lab is spared the upkeep and can keep testing. The company is spared the lab's mess and can meet clients without one of the neighbours exploding mid-meeting.
The model is not an idea on paper. It has already taken an observation all the way to its own company.
The observation came out of client work: technical SEO is stuck in the code, and an improvement that takes five minutes to write can take weeks to ship. The definition became a claim that could be wrong. The build became a layer that answers at the edge and shapes what search engines read without the site being rebuilt. The launch came when it held: its own name, its own site, its own company, in private beta today. The whole sequence is in the case.
A proof that still lives at home is no proof. It is the spinout that makes Dynamic SEO an argument. The company stands on its own, and the lab can point at it without holding it.
The model is above all a question of stability. Experiments never weigh on a delivery: what is still being tested lives in the lab, and what delivers has left it. Nobody pays to keep half-finished ideas alive, and no delivery shares the fate of an experiment allowed to fail.
Anyone meeting one of the companies also meets something built to be left. That is a higher bar than it sounds. A build that needs its builder to survive has failed, however good it is, and that rule applies from the first sketch.
Simon is the one who builds. He takes an idea from observation to something that runs: prototypes, products and the technical ground they stand on.
Why doesn't Formise keep the companies in the lab?
A lab that keeps everything becomes a warehouse of half-finished things nobody wants to maintain. The spinout keeps the lab lean enough to absorb failures, and what holds no longer has to carry the experiments' risk.
When does an idea spin out?
When the experiment graduates from play to something serious: customers, revenue or outside interest. Then it gets its own name, its own site and its own economics.
What happens to ideas that don't hold?
They are shut down, cheaply and early. When whatever succeeds is going to move out anyway, there is no reason to keep something alive just because it has already cost money.
Has the model been tested for real?
Yes. Dynamic SEO started as an observation that technical SEO is stuck in the code, was built in the lab, and runs on today as its own company in private beta. The whole journey is in the case.
Want to see where your idea sits in the model?