by Chris Lundberg
At the previous company we founded we used a dozen different APIs — for payment processing, spam checking, you name it.
And we HATED them. (Well, mostly I did because often I had to figure out how to fix them when they’d break) There are very few things more annoying than spending months or years building and deploying a product only to have it collapse when the critical payment system takes a tumble, RIGHT BEFORE AN ELECTION. Of course it’s the point provider’s responsibility, so thus ensues the delicate balance of notifying thousands of clients that “Our apologies that we went down, the real reason was our payment provider, but passing the buck is not what you need, so… well… just know that we’re really sorry.”
Here’s the reality of working with APIs: They break. They change. All the about time.
Now, things have been getting a bit better as we’ve all collectively recognized the importance of a stable, consistent API, but that is offset by the massive number of APIs we all now use. When Google Auth goes down for a few minutes it blocks access to thousands of products. The average time until API breakage is not really going up.
If you’re trying to do any integrations with other databases, you have the additional challenge of figuring out how to correctly map data — and everyone has a different opinion about the *best* way to connect two systems together. And those systems are continually changing and upgrading as well. It’s a Sisyphean task.
So why, oh why, would we choose to build an API-First company? One where APIs and integrations are the main point? It may be masochistic, but I prefer to think of it as grabbing a tiger by the tail. The “API Economy” is here. Integrations WILL happen. The teeth of the tiger are not going away.
So rather than running away, we’ve got our Bots grabbing that tiger by the tail, and we’re in a dead-on sprint to keep programming them against all the APIs and integrations we can manage. And if we can keep that tiger chasing us, maybe it will stop bothering YOU just long enough to actually run an effective marketing campaign.
Now, if only we could make a Bot with wheels.