Here’s some free advice to developers of new web apps that want their application to be adopted, spread, enhanced and embraced by third-party developers: do the API first.
Standard operating procedure with web apps these days appears to be “get the app released, write a blog post about how an API is coming Real Soon Now, wait 6 months, release API.” Sometimes the last step is never reached.
By following this timeline you’re losing 6 months of developer interest, and handing a 6 month leadtime to any competing app to become the standard platform for whatever world-changing thing your app does.
As you might imagine, I’m a keen follower of the YALBS space and at this point any app that’s missing an “API” link in the footer doesn’t interest me, as it means that the service is likely focused on building a walled garden, and doesn’t deserve my attention.
Comments
Well said!
Well said!
I totally agree (now). Some
I totally agree (now). Some time before I would have made the mistake of thinking: let’s not do an API while we experiment with what functionality our app provides, wait until it stabilizes and then come out with a complete API offering that is rock solid and won’t change until the end of all times. Only to avoid possibly having to introduce incompatible changes, which may not be as bad as initially feared at all. At least not if you make it clear from the beginning to offer an experimental web service and that the API may need to follow core changes. As you said, delaying an API is just a way of excluding external developers from having an influence in shaping the application, which makes it harder to impossible to ever suit them at all.
Add new comment