The spinout model: assets up, risk down

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 hold its own, 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. This page explains why the model looks like that, when the step is taken, and what the spinout does for both the lab and the companies that leave it.

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 still allowed to break, stays down below.

A lab that keeps everything eventually becomes a layer of half-finished things nobody wants to maintain. Every kept build draws a little attention, a little maintenance and a little guilt. In the end the time goes into keeping old builds alive instead of testing new ones, and at that point the lab has in practice stopped being a lab.

That is why the spinout is built into the model, not decided after the fact. From the day a build starts, the plan is for it to leave the lab if it holds. It can sound cold, but it is the opposite: a build that is meant to move out is built from day one to manage on its own. Dependencies are kept few, the technology is owned all the way down, and nothing important lives in one person's head.

The model also keeps the lab honest. When whatever succeeds is going to move 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. That is part of the point of having a lab instead of an agency, and why that difference matters is under Studio.

The step is taken when an experiment graduates from play to something serious. Three signals tend to decide it: customers actually using it, revenue you can point at, or outside interest in the build itself. One of them is enough to raise the question. Without any of them the build is still an experiment, and then it should stay in the lab, where experiments belong.

Before that, the idea has already walked a road that looks the same every time: explore, define, build, launch. The whole road is described step by step on the process page. The spinout is not a fifth step but what happens after the fourth, once the launch has shown it holds 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 that gets to stay an experiment costs almost nothing, and can therefore afford to be wrong.

The spinout itself is a separation. What holds is parted 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 for it to manage without the lab. The risk stays with the experiments, where it is cheap. The asset moves out, where it can grow without sharing a daily life with things that break.

Distribution is the other half of the model. The pipeline looks the same every time: 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. How that structure fits together, and why it is not a group, is under About.

The separation tidies in both directions. The lab is spared managing and can keep testing. The company is spared the lab's mess and can meet customers without one of the neighbours exploding in the middle of a client meeting.

The model is not an idea on paper. It has already taken an observation all the way to a company of its own.

The observation came out of client work: technical SEO is stuck in the code, and an improvement that takes five minutes to phrase can take weeks to ship. The definition became a claim you could be wrong about. 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, today in private beta. The whole journey 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 feet, and the lab can point at it without holding it up.

For a client or partner the model is above all a question of stability. Experiments never burden 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 its fate with an experiment that is allowed to fail.

Whoever meets 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.