The BBC as a platform? Get ready for APIs from news providers
Trushar Barot
Mobile editor, BBC World Service. Twitter: @Trushar
If you don’t use APIs in 2017, your media business will die.
An API, application programming interface, is essentially a set of building blocks made available by a platform - for example, a messaging app or social media company - to enable you to build your own functionality on top of what is offered by the platform as standard. Depending on what those building blocks are, you can create your own services or user experiences on the platform.
APIs have been around a long time. What will be significant about 2017 is just how widespread they will become, meaning news organisations will have the potential to distribute content and engage with audiences at a greater scale than ever before.
Want to develop an app for a Google, Tesla, or Apple self-driving car? That could soon be possible because of APIs. Want to pull insights data from YouTube, Facebook, and Twitter into your own in-house analytics tool? APIs mean you now can.
Want to connect your TV production systems into Facebook Live? An API will sort it for you. How about developing bespoke automated accounts or content that are more likely to be picked up with AI assistants like Amazon’s Alexa, Microsoft’s Cortana, Apple’s Siri, or Google’s Assistant? All of these AI assistants are likely to offer sophisticated APIs in 2017.
A lot of the tools currently used in the industry - Dataminr, Chartbeat, SocialFlow, and Crowdtangle, among others - take advantage of APIs offered by Facebook, YouTube, Twitter, and Instagram to deliver deeper data, analytics, alerts, or publishing mechanisms.
Among the other big platforms likely to release some sort of API in 2017 are Snapchat and WhatsApp, opening up potential access to hundreds of millions of millennial users and more than a billion users in emerging markets.
Now, more than ever, media companies will need to have access to developers who can focus on experimenting with APIs and develop uses that will work regardless of the size of your company, reaching the audiences you want on the platforms they’re already on. If you don’t have developers already embedded in your newsroom, make sure you at least have processes to enable your editorial teams to work closely with them.
So far, all the API-momentum is coming from the platforms, as they try and make it easier for third parties to engage more deeply with them.
But two can play at that game. 2017 will also the see the rise in APIs being offered by news organisations.
The BBC News Labs team (among a number of other news organisations who also do this) already works with coders during hack days where they get access to an API that pulls in BBC News content, and then allows it to be adapted for a range of different digital tools and platforms.
Last November, for instance, we facilitated a hack challenge (pictured here) for The Trust Project, which is exploring ways to make high quality journalism stand out in the age of fake news. Here, participants were given access to, and made imaginative use of the BBC Juicer, a news aggregation and content extraction API.
By offering greater API access, news organisations can also experiment with becoming capable of being platforms themselves.
The idea of the BBC as a “platform”, outlined by director-general Tony Hall last autumn, is that we would become a space for creative content - created by others, hosted by the BBC. Equally, our content could increasingly be hosted on other people’s spaces as long as our brand travels with it.
That already happens in some areas of specialist interest, such as business and technology news videos. In future, it could become standard practice to be able to embed a wider range of BBC content across third party sites and platforms, with the video player still carrying BBC branding within those videos.
If developed and deployed effectively, APIs have the potential to release content and make it more discoverable, distributable, and recognised as your brand than ever before.
This is a slightly edited version of an article originally published by NiemanLab, posted here with the author’s kind permission.