Why we measure the off-switch before growth.
Growth without an off-switch is a hostage situation. Every project we ship has one number we check before we celebrate anything else: can it stop, and still pay rent?
The off-switch is the single test we run on every project before it is allowed to graduate from MVP. It is not a phrase we use for decoration. It is a literal calculation, and projects fail it more often than they pass.
The test
Imagine, on the day in question, we stop spending on growth. Marketing budget goes to zero. Sales effort goes to zero. We stop the experiments that cost money but don’t earn money. Everything that remains — existing customers, organic, the product as it stands — runs forward unchanged.
The question: does the project still cover its own costs? Hosting, support, the slice of the operating team it consumes, payment processing, the rest of it. If yes, the off-switch works. If no, the project is dependent on growth spend in a way we are not willing to underwrite.
Why it’s the first test, not the last
Most studios run the off-switch test, if they run it at all, after the project has scaled. By that point the team is large, the contracts are long, the customer expectations are set, and the unit economics are entangled with growth spend in a way that is difficult to undo. Reversing course is expensive. Sometimes impossible without restructuring.
We run it at MVP, before scale, because that is the only point at which the answer is cheap to act on. A failed off-switch test in month three means we either redesign the unit economics or close the project. A failed off-switch test in month thirty means we go back to investors for a bridge.
A company that cannot stop growing does not belong to its operators. It belongs to whoever funds the next round.
What it changes about the work
Knowing the off-switch test is coming reshapes the upstream decisions. We don’t build features that only make sense at scale. We don’t pick categories where the only path to profitability is dominance. We don’t run pricing experiments that are subsidies in disguise.
The constraint is uncomfortable. Some interesting projects fail it. We close them anyway. The category of project that survives this filter is narrower than the category we could otherwise attempt — but it is also the category we can hold for a decade without flinching.
What we report
Every project in the studio carries a public off-switch number: the months of profitable runway it has if we cut growth spend tomorrow. Investors see it. The team sees it. We see it on the weekly review. When the number drops below a threshold, growth spend goes on hold until it recovers.
It is the boring metric. It is also the one that has saved us from at least two projects we would have otherwise scaled into a wall.
← all essays and notes