Last Thursday, Twitter did something that the Brexiteers have been trying to do since the 23rd June 2016: It took back control.
At the flick of a switch, several of Twitter’s APIs — the tools that enable outside developers to access Twitter’s internal systems — fell silent. After a decade of growth, powered in part by outsiders building their own Twitter apps, the company was instead telling apps that had happily co-existed and enriched Twitter society for years that “we don’t want you here”.
Reminder‼️ This week, on August 16th, User Streams, Site Streams, and the legacy Direct Message endpoints will be deprecated.
Today's forum post shares what to expect as well as instructions for those who have not yet migrated off of these endpoints. https://t.co/hedzTXO6as
— Twitter API (@TwitterAPI) August 13, 2018
Okay, so the Brexit analogy is perhaps a little overwrought, but the decision by the company to make it harder for external developers to use Twitter data is just as poorly thought through, and will have similarly disastrous consequences on the future of the platform as Brexit is having on our country.
The damage is immediate. Just as the value of Sterling tanked in the hours following the referendum, apps like Tweetbot — my favourite Twitter client — were rendered unable to use all of Twitter’s features. It could still load tweets, but it could no longer stream them in real time. Information about who had retweeted or liked a given tweet were no longer accessible. Rather than support a mutually beneficial relationship with third parties, Twitter was now trying to demonstrate how it doesn’t need them — and force users to stay stuck in the company’s own first party ecosystem instead.
So why has Twitter made the decision to cut off certain APIs? The company would probably argue that it hasn’t. To get into the weeds for a moment, the company has cut off the “user streams” and “site streams” API, which enable apps like Tweetbot to function. In its place, is a new “Account Activity” API which offers developers similar levels of access but there’s a catch: the price.
As Paul Haddad, one of the team at Tapbots, the company that makes Tweetbot points out, the cost for developers to access this new API is way too high — and would require his company to charge at least $11 per month, per account in order to use it. He says it would probably cost upwards of $20/month once extras like taxes and giving Apple a cut of the revenue was taken into account. So functionally, shuttering one API in place of a vastly more expensive is having the same effect — it is cutting developers off.
For those wondering this is the pricing, each subscription is an account. So a minimum of > $11/account/month. pic.twitter.com/ZdmIjVonof
— Paul Haddad (@tapbot_paul) August 16, 2018
So… why? Why would Twitter do this?
Writing on the official Twitter blog, Rob Johnson, the company’s senior data director said:
“Our goal is to deliver the best Twitter for you. This year, we’re moving faster towards this goal by focusing on improving Twitter for iOS, Android, and twitter.com. As part of this, we’ve chosen to stop supporting some other experiences.”
In other words, a load of corporate waffle. But what is actually motivating the changes appears to be as a reaction to politics: the fact that everyone seems to hate Twitter.
Over the past couple of years, the company has persistently come under attack for being a major carrier of “fake news” thanks to legions of Russian bots and fake accounts. Founder Jack Dorsey has been criticised for his seeming lack of interest in the problem that Nazis and white supremacists use Twitter — and thanks to the way Twitter perpetuates outrage, are able to spread their hate around with relative ease.
And on the other side of the political spectrum, many right-wing commentators in the US have become obsessed with the idea that Twitter is acting to censor through “shadowbanning” — something that has even attracted the attention of the manbaby in the Oval Office. Essentially then, everyone seems to believe that Twitter has its knives out for them.
There’s also the cold-hard corporate reality that Twitter isn’t growing as quickly as investors would like, and is still a social networking minnow compared to Facebook.
Obviously shutting down API access isn’t the magic bullet that solves these problems, but it's conceivable how Twitter arrived at the decision: by exercising greater control over what happens inside its borders, Twitter is theoretically more able to control the experience its users have. By restricting APIs, it could be harder for bots and nefarious actors to use the system to spread false information, and by forcing everyone to use Twitter Dot Com, Twitter can ensure that the content people see is less likely to be abuse from Nazis. Hell, it probably means they can promise advertisers more eyeballs too if users have no choice but to use the official Twitter app.
Twitter Means Twitter
Unfortunately, I can’t help be pessimistic about the changes. “Taking back control” might seem desirable in the abstract, but as with Brexit, this isn’t a decision made in isolation: though Twitter talks about a focus on “improving” Twitter, the reality is that it is making a trade off. Just as Britain is choosing more flexibility to regulate the number of scary foreigners coming to the country, over having a functioning economy, Twitter too may face damaging consequences of the decision.
What sort of consequences? The big loss might be in terms of the Twitter’s ability to innovate — harming its long term prospect. Developers are being scared away already — and shuttering important APIs creates the impression of a hostile environment for any outsiders hoping to play with Twitter data.
And we can see the potential cost of this by looking at Twitter’s history: it is a platform built on innovations made by outsiders (insert another tortured Brexit/immigration comparison here).
For example, hashtags were not dreamt up in the Twitter boardroom — they emerged organically as early Twitter users wanted a means of organising information. Similarly, the convention of referring to other users with an “@” came about after users wanted to refer specifically to other users, or to reply to each other. In time, these conventions were formalised by the company, and built into its core products.
Retweets and quote tweets are directly attributable to third party Twitter apps which were only able to exist because of the readily accessible Twitter API. Long before Twitter.com contained a retweet button, third party apps like Twhirl and Tweetdeck made it possible in one click to essentially copy and paste someone’s tweet and add a credit — enabling information to spread incredibly rapidly.
And you know how you can upload photos to Twitter? Long before we used Twitter for it, we used a third party service called Twitpic. You could sign in with your Twitter account using the API, and other third party Twitter clients would be able to display the images without the user needing to click a link.
Hell, even Twitter’s search function was invented by an outside company called Summize, and was one of Twitter’s first acquisitions in 2008.
And speaking of Tweetdeck, Twitter’s power user interface, this wasn’t invented by a coder at Twitter HQ wanting a more effective way to deal with tweets en masse — it was an independent British company that Twitter itself acquired in 2011 for $40m. I’m still grumpy about how Twitter changed the original, brilliant app in the process all of these years later.
Today, many Twitter users — especially power users like myself — still vastly prefer third party experiences for viewing Twitter. I’m an enormous fan of Tweetbot which is available on Mac OS, iPad and iPhone — my three primary devices. It offers similar functionality to web-based Tweetdeck, such as support for multiple accounts and chronological view, but in my view also offers an extra layer of gloss, which makes it a much more pleasurable user experience than the “official” channels. Sadly, the app, like many others, has been crippled by Twitter’s latest changes.
Essentially my point should be obvious: the Twitter of today, which plays such a fundamental role in sharing the world’s information doesn’t exist solely because of the work of Twitter. The company has built its global position by building a platform on which others can innovate.
And rather than close itself off from the outside, Twitter would be far better served if it continues to embrace this fact.
So what if people choose to use third party Twitter clients? Surely that should be celebrated as a good thing? Users are still on Twitter — they are just choosing to experience it in a way that better suits their needs. If Twitter’s one-size-fits-all (well, two-sizes if you count Tweetdeck too) approach doesn’t meet their requirements, then no need to worry.
To be clear, while Twitter definitely has issues it needs to address (see the little Nazi problem described above), screwing with third party developers like this really seems like the wrong intervention here.
If Twitter was truly confident, it wouldn’t put up barriers to keep people inside, it would build its own products, like Twitter.com and Tweetdeck, and enable them to compete on a level playing field with the same access to the same tools. If Twitter wants us to use its “official” products then it shouldn’t coerce us to use them, it should persuade us that they are better.
And by forcing everyone — Twitter itself, and third party developers — to keep innovating, this will result in a better platform for everyone overall.
So Twitter, I beg you — like Brexit, there is still time to reconsider and step away from disaster. Yes, you’ll have to make a massive screeching U-turn and there might be some political fallout from it, but it might be better for everyone in the long run.
James O’Malley is Interim Editor of Gizmodo UK and tweets as @Psythor.