Going from being on top of public APIs to feeling the way with MCPs

If you’re running an API product business—one where APIs are your primary revenue stream—you’re likely grappling with a critical question right now. What should our strategy be for MCP (Model Context Protocol) in the age of agentic AI? This has moved on from being a theoretical concern. The signals are everywhere. Your established API business is humming along nicely—you’ve got sophisticated infrastructure, mature governance practices, strong customer onboarding, and the revenue numbers to prove it all works. But then the landscape shifts. Agentic AI arrives. Suddenly, you’re fielding questions about how agents will discover and use your APIs. You’re watching competitors experiment with MCP servers. You’re wondering if your current approach is future-proof.