API Half-lives


At DuckDuckGo, we use a lot of external APIs and services. Sure, we use more than most, but the trend is clear: startups are building more and more services and other companies are relying on them more and more.

I wholeheartedly agree that it's annoying that these APIs aren't as stable as things that came before, but the bigger problem from my perspective is that their half-lives are tiny. Already we've done several integrations where subsequently the company has either gone out of business, "pivoted" and ditched or abandoned their API in the process, or got acquired and the new company shut down the API.

And its not just startups. I've lost count on how many services Google has sun-setted this year -- not just ones they acquired, but also their own.

Don't get me wrong, I don't fault the individual decision makers. I don't have full background information, and for all I know I'd make the same decisions if I did. But what I do know is that my process for looking at APIs and online services is changing.

At my first company I got asked this question a bunch: how do we know you're going to be around in five years? (We weren't.) 

Now five years seems laughable! Almost all entrepreneurs I meet don't think they're going to be doing their startup in five years. Side note: it's a good negative investment filter.

I don't really have good answers yet for what to do about it or how to better evaluate a particular case. I'm annoyed at myself for doing this, but I find myself gravitating to startups that have taken in a lot of funding. I love working with startups because they are hungrier and more flexible, and so this type of startup seems like a sweet spot. They have resources to spare, both to help you out and stay in business.

Some other things I've found myself thinking about in particular cases wrt to this question about API half-lives:
  • How active is their API discussion forum?
  • How responsive are they to API bugs/issues?
  • How core is the API to their business?
  • How long have they been around already?
  • What is their runway?
  • Do they seem like they're passionate about what they're doing?
  • Are they likely acqui-hire targets?
Another approach is to apply Chaos Monkey philosophy and expect anything to die at any time. We pretty much do that. However, we still need to place bets. With so many startups, there are often several choices in any given category. 

Update: good comments on HN, especially this one.


If you have comments, hit me up on Twitter.
I'm the Founder & CEO of DuckDuckGo, the search engine that doesn't track you. I'm also the co-author of Traction, the book that helps you get customer growth. More about me.