Why we built the wedge before the platform
Platforms are a reward for being useful, not a strategy for becoming useful. Y0 exists because we refused to skip that step.
Every few weeks someone asks me why Mynd Labs — a company that openly wants to build an eight-lane ecosystem — shipped exactly one product. The honest answer is that we tried to write the platform pitch first, and it kept failing a very simple test: who, specifically, wakes up tomorrow and needs this?
A platform answers that question with an audience. A wedge answers it with a name. When we wrote the spec for Y0, the cognitive runtime, we wrote it against one person — a freelance designer in Indiranagar named Priya, composited from about forty user interviews. She does not want an ecosystem. She wants her Thursday deadline flagged before it becomes a Friday apology, and she wants the invoice drafted with the client, scope, and amount already known. That is a wedge: a problem narrow enough that you can be the best in the world at it within a year.
A platform is a promise about everyone. A wedge is a promise to someone.
The seduction of building wide
The platform-first temptation is real because platforms look like leverage. Build the substrate once, the thinking goes, and every product after that is cheap. But the substrate only knows what to be after products have stressed it. We learned this in our own codebase: the context graph that powers Y0 today is the third design. The first two were built in the abstract, for hypothetical products, and both collapsed the moment a real workload arrived. The version that survived was extracted from a working product, not designed ahead of one.
There is also a colder reason. Platforms have brutal cold-start economics. Every horizontal feature you ship before retention is proven is capital spent buying optionality you cannot exercise. We are a small company in Bengaluru, not a memo at a hyperscaler. Our optionality budget is zero.
What the wedge must prove
Internally, Y0 carries three burdens of proof, and the platform stays frozen until all three clear. We review them monthly, in writing, with the actual numbers on the page:
- ↳Retention — week-eight retained usage above our floor, cohort over cohort, without paid re-acquisition.
- ↳Revenue — people pay, at a price that survives contact with their finance team or their own wallet.
- ↳Reach — a distribution channel whose cost per retained user we can fund from margin, not from hope.
Only behind those gates does the platform make sense — because by then it is not speculation, it is extraction. The runtime, the context graph, the trust kernel: each becomes a platform surface the day a second product needs it, and not a day before.
The order is the strategy. Wedge before platform, platform before ecosystem. Everything else we publish — the lane map, the principles, the trust architecture — is downstream of this one refusal to skip ahead.