While I like that @Havn wishes for a world of APIs and protocols for services, it’s mostly a dead end 1.

We tried the world of services and APIs and called it Web 2.0. There are all kinds of places on the web that are talking about APIs and protocols again, very, very slowly gaining traction. The problem with the idea that some businesses are services and that they don’t need to be clients is that those are almost always terrible businesses compared to the alternative. Maintaining an API or protocol makes client applications your customer. But these clients generally do not want to pay for access to services or don’t have financial models that support passing those costs on to customers.

And for all the costs of maintaining a service or API, consumer preferences have longed shows that third-party or alternative clients are minuscule portions of usage compared to first party offerings, even when they’re superior. Apollo never impacted overall Reddit usage. Twitter apps barely registered on monthly active users.

“Services”, as though that boundary was clear, make less money with a more difficult to serve customer that represents a vanishingly small part of their business when they build to allow clients. So they eventually stop.

This just won’t happen unless users actually adopt clients, but they mostly don’t. This won’t happen unless clients and their customers are willing to pay for services access.


  1. Yes, I know he said he’s not predicting. It’s still worth examining what prevents this world. ↩︎