Newspapers innovating with API’s the way you’d expect newspapers to innovate with API’s


Metered-digital paywalls, restrictive Terms of Use, and data-limited API’s present problems for newspapers trying to enter the 21st century

If you had a time machine and went back to the early 1990’s, I bet you could sit in a conference room with newspaper executives and hear them talk about the Internet as though it were merely a series of tubes — something that a dump truck could quite literally get stuck in — and nothing they should concern themselves with.

Fast forward to 2013 and most newspaper executives have come around to the fact that their industry is hemorrhaging readers and burning through money faster than people can use newspapers to start campfires (I believe kindling for camping trips has become one of their big selling points). In short, the Internet killed the newspaper star.

But there has been a glimmer of hope on the horizon for some of the papers that are trying to offset the losses from the dramatic decrease in print distribution and advertising — the metered digital paywall — which has successfully increased digital profits by requiring someone to purchase a subscription after viewing X articles per month. This type of system helps to prevent huge drop-offs in digital advertising revenue by ensuring that ads will still be shown to new visitors who organically found a store, while also encouraging new digital subscriptions.

So that’s the end of the story, right? Metered digital paywalls + digital advertising + physical delivery + print advertising = 100 more years of the newspaper golden age!

Not quite … newspapers are still trying to figure out how to actually build a 21st century product. Last year, there were four big publishers who had API’s for their newspaper content: The Guardian, The New York Times, USA Today, and NPR. Today, there are a couple more papers with API’s, including the Washington Post’s nascent efforts, and the sophisticated Zeit Online API archive, which is unfortunately only available for non-commercial use.

And why do newspapers need to build an API? The main reason is that API’s spur innovation, experimentation and they empower 3rd party developers to build apps on top of existing data. For a newspaper, an API means that someone could actually build something out of newspaper articles, or test new designs, or new ways to read and manage newspaper articles, or explore the big data world of a newspaper archive. API’s present newspapers with a hope of being able to cross the bridge into the 21st century, but they also lead to a series of problems due to the current metered digital paywall strategies, and restrictive Terms of Use.

A newspaper API is useless if it’s hampered by Terms of Use and paywall restrictions

Less than 30 people have watched this YouTube video from August 2013 featuring Erik Bursch, the Director of IT Operations and Content Systems for USA Today, and it explains everything that is wrong with the newspaper industry. First, probably the most important news from the video comes around the 17-minute mark — Gannett papers are developing an API for ALL of their newspapers using Mashery, which could be as many as 100 papers opening up data. Erik stated:

“We’ve really come full circle since 2010, you know with the Terms of Use change speaks loudly to what the perception changes inside of Gannett and USA Today. And then at that point, Gannett has looked at what the USA Today API has done for them, done for us, excuse me, and is now replicating that app to all the Gannett properties, which is in development right now. So Gannett and USA today will have that same API layer across all properties, which will be a huge win all around for us.”

In the video, Erik also goes into great detail about how USA Today has worked to be “the first and the best” and how they have iteratively developed their API based on developer and licensee feedback. He also spoke about how USA Today has three or four meetings a week debating how to provide the API data, and they came up with two really important conclusions:

1.) USA Today made a rather revolutionary change to their API Terms of Use to allow commercial use of their data, which brings them up to par with The Guardian. This means that a 3rd party developer can actually build an app that generates a source of revenue and not be in violation of the API Terms of Use. This is huge — revolutionary, especially since Gannett is using USA Today as their model for their big newspaper API expansion. That being said, one big difference currently between USA Today Terms of Use and the Guardian Terms of Use is that the Guardian has a clause that is very smart—and is part of the framework for a Newspaper Article API + Advertising Network that would require someone to, “display on Your Website any advertisement that we supply to you with the OP Content. The position, form and size of any such advertisement must be retained as embedded in the OP Content.”

Unfortunately, the New York Times and NPR still have non-commercial clauses on their newspaper API Terms of Use, which dramatically hurts their ability to offer a developer-friendly API.

NPR is doing fantastic work with their API archive going back to 1995, which includes a cutting edge transcript API for their radio shows, butunfortunately their non-commercial Terms of Use restrictions and a couple dozen other restrictions make it much less likely that people would want to develop innovative apps on top of their API. The New York Times has 14 API’s currently available, but their app gallery only lists 32 apps developed from the data, and if you actually click through the list of apps, it’s a bunch of half-baked apps, literally dead links and websites that are available to be purchased — all in all a pathetic excuse for an API gallery, and a resounding rejection of their API strategy.

2.) USA Today and Gannett can’t figure out how to build an open API within their existing metered paywall structure — This is the saddest news to date within the newspaper API debate — due to the state of the industry and their new reliance on metered paywalls, it’s nearly impossible to find a model where they can actually fully open up the Article API data. Developers can get headlines, excerpts and link backs to the original article, but they can’t get the full text, which dramatically limits how someone could use the data. It’s like opening up the largest and nicest golf driving range in the world and then telling everyone they can only use putters.

So what does this all mean? Essentially, Gannett is currently developing the largest newspaper API the world has ever seen. They are breaking down barriers and have relented to let developers use their data for commercial use. But they also appear to not have a solution for how to ACTUALLY open up their data within their existing metered paywall structure. And it also appears they aren’t taking the lead from The Guardian by building out an advertising network tied to their Newspaper Article API, even though both companies use the fantastic API company Mashery.

A good analogy would be that developers are now like that kid who comes down for Christmas to a room full of presents, only to find out that their terrible parents put bike locks around all the presents and swallowed the keys and are forcing the kids to just stare at the presents with big locks around them. Okay that’s a bad analogy, but you get the picture — an API needs to be completely open and available for commercial use in order to spur innovation, otherwise it’s just newspapers using API’s exactly the way you’d expect newspapers to use API’s.

So what are newspapers to do? What are some ways they could innovate? Good question and it’s something I wish that Gannett and other papers would start discussing more openly. There should be hundreds of people engaging in this dialog about what could be valuable within a newspaper archive and article API, along with other data they could hold, and how the entire ecosystem be opened up while still generating profits for the paper.

Newspaper API ideas should come from more than one person, but here are a few

1.) Combine the Article API with an ad network in order to serve up ads on 3rd party websites and in 3rd party apps — The Guardian is already planning/doing this and it seems like the logical way forward. In fact, newspapers building out their API’s would be smart to model a lot of their work off of The Guardian’s API, especially the breakdown of their Content API Terms and Conditions. This model creates a distributed advertising network, provides new opportunities for niche advertising, and puts newspapers back on track to generate more of their money from digital ads.

2.) Breakdown the API access to categories and don’t allow a single domain/app to access more than one API feed. So essentially you could facilitate sports apps, or politics apps, or local news apps — you wouldn’t be giving away the whole enchilada to one organization, but you also would be opening up the full articles/archives so developers could actually do something with them.

3.) Build out a Newspaper Archive API with full-text articles and commercial access and just limit it to stories older than X months/years. This would limit problems with licensing clients while also opening up possibilities for really innovative apps. And with this type of system, merely integrate an Ad Network into the API feed in order to monetize it.

4.) Take niche events like the 2016 Presidential Election and open up articles just around that topic. So essentially create an Election 2016 API that would provide 3rd party developers with a platform and time to think and build out a wide range of innovative apps. Someone could do that for a wide range of niche categories, which could work for a large network like Gannett.

5.) Provide a platform within a paper to compile local content through a “Community Collaboration API.” Newspapers used to be the central place a community would look for news, but with the advent of bloggers, craigslist, independent news organizations, and a host of websites, it’s becoming harder and harder to find one central place for community news. Newspapers could develop a “Community Collaboration API” and build out services/plugins for the major blogging platforms like WordPress, Drupal, Wix, etc, so that bloggers could push their content to a larger audience housed on the newspaper website. The content could be categorized and organized, and newspapers could become the “Drudge Report” of local link farming, but focused on niche local blogs and issues.

6.) It needs to be easier to create API’s and people need to be better educated about how a data mashup can reveal trends and facilitate data-driven decision making. A great company working on this is, and one of their funders is the new Washington Post owner Jeff Bezos. Perhaps we’ll see something from WaPoLabs that will make API management and data-mashups between papers, bloggers and archives easier, but we’ll have to wait and see.

Finally, I’m just one guy ranting about what would be nice to see in a Newspaper API and developer environment– but there are hundreds if not thousands of people more qualified to talk about this subject. News organizations like Gannett need to open up discussions when they are developing their API, not AFTER they have already developed the API. The developer community shouldn’t be finding out about the largest newspaper API in the world through an errant comment on a YouTube video. It’s time to start talking honestly about how newspapers can prosper in the 21st century, how they can encourage innovation, and how developers and news organizations can work together to better inform, educate and entertain the public.