This deep dive expands on the Loop Asia conversation with Erik Wilde, ambassador for the OpenAPI Initiative.

Every platform investment case gets asked the same question: what does this get us? It's fair, and the wrong shape for what Erik Wilde described in Paris — the honest answer is nobody knows yet. That's not a weakness. It's the mechanism by which platforms pay off.

Erik's example was Uber Eats. Uber's founding plan was personal transportation, replacing taxis — food delivery almost certainly wasn't the original vision. But once routing, mapping, and invoicing existed as reusable services, launching Uber Eats meant adding one missing piece, the restaurant side of the marketplace. His read: it probably took Uber a fraction of the effort a standalone competitor would need, because the platform was already there.

This is the Ansoff matrix, running on system components instead of products

Marketers already have a name for this: a matrix for growth — new product into an existing market, or existing product into a new market. Erik's combinatorics example runs the same matrix one level down, on technical capabilities instead of go-to-market moves: a team with customer purchase history as a reusable service and a team with a recommendation engine as one can combine them into an offer neither was asked to build.

That reframing matters for whoever owns the P&L: it changes what you're funding. You're not buying this quarter's feature — you're buying next year's option to combine things cheaply, and options don't show a return until someone exercises one.

The caveat Erik wouldn't let go of

Erik was careful not to oversell this: "APIs aren't magical." Whether a company like Uber goes looking for adjacent opportunities is a cultural question, not a technology outcome. The platform makes cheap experimentation possible — but that only pays off if someone is running experiments. Most organizations fund the technical side of that without funding the cultural side: room to combine services into something nobody asked for, and tolerance for the ones that go nowhere.