Build the platform. Let others build on top. Capture the ecosystem. Become the foundation.
This is the pitch. The reality is more expensive.
⸻
The Abstraction Burden
A product serves a user. A platform serves developers who serve users. This additional layer of indirection multiplies every cost.
Your API must be stable, because breaking changes destroy your ecosystem's trust. Your documentation must be comprehensive, because your users are building production systems on your specifications. Your error messages must be precise, because the developer debugging at 2 AM cannot call your support line.
Building a platform means building a product where the tolerance for ambiguity is zero and the feedback loop is measured in months, not sprints.
⸻
The Ecosystem Dependency
The platform is only valuable if the ecosystem builds on it. The ecosystem only builds on it if the platform is stable, well-documented, and unlikely to be deprecated.
This creates a chicken-and-egg problem. You need adoption to justify investment. You need investment to earn adoption. Most platform plays solve this by subsidizing early adopters with engineering resources, dedicated support, and custom integrations that do not scale.
By the time the platform reaches general availability, the cost of supporting the early adopters has consumed the team's capacity. The platform launches, but nobody is left to maintain it.
⸻
The Lock-In Calculus
The darker version of the platform play is deliberate. Build the platform not to serve the ecosystem, but to trap it.
Make the migration cost high enough that switching becomes impractical. Embed proprietary APIs deep enough that extraction requires a rewrite. Offer a generous free tier until the customer's architecture depends on your infrastructure, then adjust the pricing.
This is not strategy. It is a hostage negotiation with a long fuse. The customers who build on your platform know this, and the sophisticated ones will maintain an abstraction layer specifically to protect their exit. You have not earned loyalty. You have earned wariness.
⸻
The Test
Before building the platform, ask one question: would developers choose your platform if switching costs were zero?
If the answer depends on lock-in, pricing tricks, or the absence of alternatives, you are not building a platform. You are building a trap. Traps work until someone finds the exit.
End.