Facebook And Twitter Only Now Beginning To Police Their API Applications
30 Apr 2018
I’ve been reading about all the work Facebook and Twitter have been doing over the last couple of weeks to begin asserting more control over their API applications. I’m not talking about the deprecation of APIs, that is a separate post. I’m focusing on them reviewing applications that have access to their API, and shutting off access to the ones who are’t adding value to the platform and violating the terms of service. Doing the hard work to maintain a level of quality on the platform, which is something they should have been doing all along.
I don’t want to diminish the importance of the work they are doing, but it really is something that should have been done along the way–not just when something goes wrong. This kind of behavior really sets the wrong tone across the API sector, and people tend to focus on the thing that went wrong, rather than the best practices of what you should be doing to maintain quality across API operations. Other API providers will hesitate launching public APIs because they’ll not want to experience the same repercussions as Facebook and Twitter have, completely overlooking the fact that you can have public APIs, and maintain control along the way. Setting the wrong precedent for API providers to emulate, and damaging the overall reputation of operating public APIs.
Facebook and Twitter have both had the tools all along to police the applications using their APIs. The problem is the incentives to do so, and to prioritize these efforts isn’t there, due to an imbalance with their business model, and a lack of diversity in their leadership. When you have a bunch of white dudes with a libertarian ethos pushing a company towards profitability with a advertising driven business model, investing in quality control at the API management layer just isn’t a priority. You want as may applications, users, and activity as you possibly can, and when you don’t see the abuse, harassment, and other illnesses, there really is no problem from your vantage point. That is, until you get called out in the press, or are forced to testify in front of congress. The reasons us white dudes get away with this is that there are no repercussions, we just get to ignore until it becomes a problem, apologize, perform a little bit to show we care, and wait until the next problem occurs.
This is the wrong API model to put out there. API providers need to see the benefits of properly reviewing applications that want access to their APIs, and the value of setting a higher bar for how applications use the API. There should be regular reviews of active APIs, and audits of how they are accessing, storing, and putting resources to work. This isn’t easy work, or inexpensive to do properly. It isn’t something you can put off until you get in trouble. It is something that should be done from the beginning, and conducted regularly, as part of the operations of a well funded team. You can have public APIs for a platform, and avoid privacy, security, and other shit-shows. If you need an example of doing it well, look at Slack, who has a public API that is successful, even with a high level of bot automation, but somehow manages to stay out of the spotlight for doing dumb things. It is because their API management practices are in better alignment with their business model–the incentives are there.
For the next 3-5 years I’m going to have to hear from companies who aren’t doing public APIs, because they don’t want to make the same mistake as Facebook and Twitter. All because Facebook and Twitter have been able to get away with such bad behavior for so long, avoid doing the hard work of managing their API platforms, and receive so much bad press. All in the name of growth and profits at all cost. Now, I’m going to have to write a post every six months showcasing Facebook and Twitter as pioneers for how NOT to run your platforms, explaining the importance of healthy API management practices, and investing in your API teams so they have the resources to do it properly. I’d rather have positive role models to showcase rather than poorly behaved role models who I have to work overtime to change perception and alter API provider’s behavior. As an API community let’s learn from what has happened and invest properly in our API management layers, properly screen and get to know who is building application on our resources, and regularly tune into and audit their behavior. Yes, it takes more investment, time, and resources, but in the end we’ll all be better off for it.