Backed, not bought.

This house was built around a refusal: no client work, no briefs, no sign-offs from outside the building. So it is fair to ask why a company like that accepts a grant from ElevenLabs, or a place in Google's startup programme — and the answer is worth writing down, because the line between backing and dependence is exactly where young product companies quietly lose themselves.

What backing is for

Good backing buys speed on things that were already decided. Credits pay for compute the roadmap already demanded; a partner programme opens doors the product was already walking toward. In every case the test is the same: would we be building this exact thing, in this exact order, if the support vanished tomorrow? If yes, take the help gratefully. If no, the support is not backing — it is steering, and steering is the thing we built this house to avoid.

Support may buy speed. It may never buy a vote.

What it is not for

Recognition is pleasant, and pleasant things are dangerous to strategy. A grant logo is not a business model; an acceptance letter is not revenue. The moment a programme becomes part of a company's identity rather than its toolkit, the company has started optimising for selectors instead of customers — pitching upward instead of shipping outward. We put recognition on the record because it is checkable fact, and facts about a young company are scarce. We do not put it on a pedestal, for the same reason.

The ledger stays ours

Owned is still the word that governs. Every programme we join is weighed like a supplier: what it costs in attention, what it returns in speed, and whether we can leave it without limping. That last clause is the whole doctrine. Backing a house can walk away from keeps it honest; backing it cannot walk away from has already bought it. We intend to remain, permanently, the kind of house that can walk away.