VIENNA — More than 750 senior news executives, journalists, media innovators and startups hailing from over 65 countries gathered in Vienna last week for the seventh Global Editors Network Summit to discuss trends and technologies shaping the media industry now and in the future.
The conference host, the Global Editors Network (GEN), is a six-year-old non-profit journalism organization headquartered in Paris with the mission to define a vision for the future of journalism and further freedom of information and independence of the news media.
The following is a recount of the most salient topics debated at the summit.
Technology is Quickly Becoming Ubiquitous – and Paramount – for News Orgs
Automation and artificial intelligence dominated keynotes, panel topics and conversations in Vienna. The resounding message was that these technologies are no longer nice-to-have tools — they are now widely regarded as an indispensable element of journalism no news operation can afford to ignore.
Amy Webb, founder and CEO of the Future Today Institute, said during her keynote that the future of news is inextricably linked to the future of artificial intelligence. “AI will become so sophisticated,” she said, that those who build and control AI systems will get to make the decisions about “how people get informed” and about “what constitutes news itself.”
If news organizations wanted to be in a position to shape the “information landscape,” Webb continued, they’d need an “equal seat and an equal say at the table” with the big tech companies. Merely producing content and refining existing processes and workflows no longer suffices, Webb noted. Instead, news organizations need to become aggregators themselves and develop their own infrastructure, such as proprietary “algorithms to parse data,” platforms as well as a “visual object recognition strategy for revenue, advertising or for content.”
During his highly-anticipated keynote, The Washington Post’s executive editor Martin Baron took this idea one step further, arguing that news orgs should try and become “vendors of news technology to other organizations.” In other words: monetize in-house technology.
Baron said the Washington Post “is evolving into a technology company, even as it remains a journalism company.”
“At the most basic level,” the news veteran said, “it means that technologists are first class citizens in our news organization. They are not there merely to provide support to journalists or the ad department. Our engineers serve as a creative force that drives growth.”
Francesco Marconi, strategy manager and automation co-lead at the Associated Press, said the AP’s artificial intelligence strategy consists of automation, or automatically creating stories, which the AP started in 2014, and augmentation, which means A.I. “enhancing journalists to do more work and find new [story] angles,” Marconi said.
Marconi cited the “Seafood from Slaves” project as an example of how the AP leveraged A.I. to advance its storytelling.
“One component we used to capture this information is … computervision, a sub-domain of artificial intelligence, to track how the fishing ships that were carrying slaves were moving,” Marconi said during his panel. “The story is not what technology we used, it’s the impact that story has.”
Robert Unsworth of news app News Republic, echoed co-panelist Marconi’s sentiment, arguing that the business model ought to drive the technology, not the other way around. “It’s very much about looking at the carpenter bottlenecks that you have, areas you … could perhaps automate in an intelligent way,” said Unsworth, who is currently building mobile operations in North- and Latin America for News Republic.
— Benjamin Bathke (@BenjaminBathke) June 21, 2017
Unsworth also stressed that “even the cleverest of systems” cannot replicate journalists. “The legwork is done by machine-based platforms, but the last mile is the human intervention,” he said.
Although nobody at GEN Summit denied A.I. and automation can be a boon for journalism, some attendees had a few words of caution. The Financial Times’ Andrew Jack, for instance, said his experience with A.I. has been “quite mixed” in terms of how far it can really deliver targeted content that’s valuable.
“Clearly, there’s a lot of potential for that over time, but it’s not in the leading position. It requires all sorts of information, including integration with third-party data and platforms alongside the information on readers that any media outlet has,” Jack told me in an interview.
Amy Webb had a few additional recommendations for publishers and newsrooms:
- Develop a strategic plan and find true industry collaborators
- “Listen for signals and look for patterns from sources you wouldn’t normally look at,” such as the MIT Technology Review
- Track real trends and technologies, and spend ten minutes a day reading up on new ones
- “Move trends into action,” but map scenarios for the implications of a given technology first
Webb also advised news orgs to pay close attention to Chinese news aggregator Toutiao, an AI platform that currently has 78 million daily active users. Toutiao, Webb said, offers a glimpse at the future where publishers are “responsive wire services” that use “machine learning algorithms to surface and deliver information that’s relevant to what I’m doing …, is empathetic to my background and to my worldviews and offers radical transparency.”
Monetization: No One-Size-Fits-All Solution in Sight
Speaking of China: Toutiao, along with the two other dominant platforms WeChat and Weibo, was was a focus of a panel titled “How to monetize content: Lessons from China.” Overall, publishers’ revenue models in China are similar to those of their Western counterparts, said China Business News executive editor-in-chief Yudong Yang. Advertising is the main source of revenue, accounting for 90% of income, according to Yang.
Chartbeat’s senior data scientist Sonya Song said Chinese platforms avoid a “news media” label and produce very little political news as China’s media remain heavily regulated by the government. Since lifestyle and utility news dominate, Song said, micro-celebrities and influencers often have a better ROI than traditional Chinese media. A case in point, according to Song, is the fact that of the 35,000 Toutiao accounts approved for content contribution, 73% belong to micro-celebrities. This isn’t, however, dissimilar in the West, she added.
In China and the West alike, monetization models differ by content type — the production of ‘hard news’ in China, for instance, mainly relies on subsidies and subscriptions; In the West, it is mainly financed through subscriptions as well as memberships and donations. Song also said native ads on social and mobile are a strong revenue engine in China, a point Condé Nast chief digital officer Wolfgang Blau underscored with his statement that one third of Vogue magazine’s digital revenue in China comes from WeChat.
Globally speaking, the Digital News Report, once again a highlight at the summit, had some noteworthy findings when it comes to the monetization of news. In the United States, the number of people prepared to pay for online news grew from 9% in 2016 to 16% this year, and donations to major news organizations, often referred to as the ‘Trump Bump,’ tripled since last year.
New payments mostly came from young people under 35 as well as those on the political left, with almost a third saying they want to ‘help fund journalism.’ The report also noted, however, that “it is too early to know whether these increases constitute a groundswell, or simply a knee-jerk reaction to a political shock.” Plus, it is worth pointing out that the surge in subscriptions during the Trump administration was by and large limited to major publications like the Washington Post, the New York Times, non-profits like ProPublica, as well as higher ratings among major cable networks.
Whether or not the ‘Trump Bump’ will last, at least for the moment news organizations are able to do something thought almost impossible until recently: monetize investigative reporting. “For a long time, people said this is work that couldn’t be done anymore,” Baron told me in an interview in Vienna, “simply because it took too much time and required too many resources. What we are discovering now is that the public will not only pay for investigative reporting, they also want us to do that kind of work.”
— Benjamin Bathke (@BenjaminBathke) June 22, 2017
Andrew Jack, the Financial Times’ head of curated content, told me in an interview the FT currently has 860,000 paying subscribers. “A large number of those are premium subscribers, particularly businesses and other organizations,” for which the FT has a meter model in place — the more content they consume, the more they pay. “This is also a powerful metric in showing how much they value our content,” said Jack, who has worked as a journalist for the Financial Times since 1990. At the same time, he acknowledged the ROI of higher engagement through newsletters is “difficult to quantify.”
Personalization of News and Engagement are Closely Intertwined
The fundamental idea behind audience engagement, in the words of Andrew Jack, who oversees the FT’s more than 15 newsletters, is both to “encourage discovery of content” — help busy readers cope with the “information overload” — as well as “distill a smaller number of things worth reading, and build a narrative and analysis around them.”
According to the Digital News Report, email has made a comeback in many markets on the back of more personalized technology and mobile delivery. In Belgium, for instance, more people (34%) access news via email than through social media; And in the US, where a quarter use email for news (23%), the charge to “drive both habit and subscription” has been led by Quartz, the Washington Post and other players.
The Financial Times’ broader content personalization strategy, Jack told me after his panel, is to get readers with “particular interests and focuses” come back more often and turn them into loyal readers.
But in the light of the Digital News Report finding that “brand attribution in distributed environments is a problem,” engaging readers who already have an affinity to a given brand is likely enough. According to the Reuters Institute’s annual study, fewer than half of people polled could remember the news brand behind a story when it came from social media or search.
A second area of focus at London-based FT, according to Jack, is user-led personalization — encouraging readers to flag up content they want to follow. “This led to a 35% increase in engagement,” Jack said, “precisely because they were seeing things that were targeted, which tempted them to read a bit more.”
Jack also pointed out that while many readers lean on “editorial judgments” to select content for them, personalization has its limits. “There’s still a lot of appetite by readers to make their own preferences,” he said.
In another presentation about the personalization of news, Vesselin Popov discussed how news is increasingly delivered to us based on our digital footprint — demographics, location, age, etc. — as well as our psychological profiles or interest. To Popov, who works at the University of Cambridge Psychometrics Centre, ‘psychometrics,’ or the psychological prediction of our personality based on our data trail, especially on social media, is a double-edged sword.
One the one hand, psychometrics is a technology that can improve experiences for readers — from real-time personalisation to business model innovation.
“Personalizing content and explaining how it’s personalized can help people break out of their filter bubble,” Popov told me in an interview. These “reverse recommendations algorithms” show users articles they otherwise wouldn’t read. At the same time, this maximizes time with your audience. “People are more likely to engage with content if it is personalised,” he said during his speech, adding that personality-matched ad content is also twice as profitable. In the “battle for attention,” Popov said, personalization is “essential.”
For newsrooms, psychometrics offer an opportunity for “topic modelling based on psychological footprint data,” Popov said. And, in case they only speak to people with a particular personality, it can help them break out of their own echo chamber by diversifying their audience.
Yet the technology also poses serious risks, according to Popov. One example is the potential to manipulate voters by delivering false news, as seen in the case of British firm Cambridge Analytica, which claims to have used psychological prediction and micro-targeting techniques to sway people in the UK and the US to vote for a certain candidate. To Popov, this poses a “fundamental threat to freedom of speech and democracy.”
“It’s undeniable that these techniques have an individual impact,Popov told me. “One of the best ways of controlling people is limiting their access to information … We need to be having a conversation about how they should or should not be used, and how they can be regulated to ensure democratic principles are protected.”
Trust is on everyone’s mind
One of the most widely discussed findings of this year’s Digital News Report is that (dis)trust in the media is heavily driven by your political viewpoint, particularly in countries with high levels of political polarization. So the most important question in journalism right now might be how the media can (re-)establish trust with their audiences, especially after Trump’s election victory exposed shortcomings when it comes to reporting on certain demographics and taking different viewpoints into account.
Here are a few selected GEN Summit attendees weighing in on this vital question.
“There needs to be much stronger pressure on explaining how personalization has come about. Is it based on public data, is it age and gender, or is it psychology and intelligence? If the goal is to increase understanding of the salient issues in our society across the spectrum, then we can’t just be elitist about the people who are tech-savvy, well-educated and high critical thinking to distinguish fake from real news. We need to speak to the audience and take on a journalistic voice we’re less comfortable using maybe.”
“It’s time for a nutritional label for news, where you’d explain the sources of personal information that’s being included: every single person who worked on the project, the context and how data was used. Not everybody is going to pay attention to this, but this helps solve some of the other problems we have, and it makes your information more useful.”
Rob Wijnberg, founder and editor-in-chief, “De Correspondent”:
“As a journalist, you have to think about how to establish a trusting relationship with your audience. A key part of this is letting readers know about you, your personality and your character, and also about why you care about the topics you write about — and why you think other people should care about those topics. That’s something journalists aren’t used to because they’re supposed to be objective and report facts.”
Dan Gillmor, ASU professor:
“Transparency is incredibly valuable. It’s the first step journalism organizations can take to help their audience be news literate and to do things that in the end will come back around to help people in journalism itself … Consumer Reports completely screwed up a story about baby seats … But they retracted it and reprinted the correct one, and they sent out a note to everyone saying ‘Here’s what happened, and here’s how we’re not going to screw it up next time.’ … It’s easy to notify people who want to be notified that they’ve seen a mistake, and the correction. But as far as I can find, nobody is doing that at the moment, although it’s trivially easy to do technologically.”
“We have to think long-term and keep doing our jobs. It’s quite possible that over time if people see that our work is validated, that we have been steadfast in our effort to fulfill our mission, and that the work that we did proved to be true and had impact, then people might change their view of the American press. I don’t think that we should be responsive to just short-term swings and approval ratings. And I certainly don’t think that our job is to pander and to placate and to appease. I don’t think the public would forgive us if we abandoned our mission.”
Ben Smith, editor-in-chief, BuzzFeed:
“[Political reporting] is not really a place where you win trust, because the waters are so poisoned. People use political stories to club each other with, not to build community around. Another path of how we do build trust is … investigations that don’t have to do with politics. We published this year a huge exposé of the largest psychiatric hospitals chain in the United States that was essentially tricking people into being institutionalized and making enormous amounts off their insurance … We found a huge appetite for that … At a moment of really historic doubt in institutions, we shouldn’t expect people to trust us, because we are keeping secrets from them. They should trust us for the opposite reason — because we’re sharing what we know and not confusing professionalism, which is really important to keep.”
Throughout the conference, speakers recommended resources newsrooms and journalists can tap into. Here’s a selection:
- “The future of augmented journalism: A guide for newsrooms in the age of smart machines,” provided by AP’s Francesco Marconi
- DropBox folder with readings, research, a decision matrix, open-source tools and tech trends, made available by Amy Webb
- VR tipsheet with tips, links and resources to “better navigate through the emerging world of virtual reality and journalism,” courtesy of USC’s Robert Hernandez
- The eight finalists of the Startups for News competition; the 16 prototypes of the Editors Lab; and the dozen projects that took home a Data Journalism Award
- News Integrity Initiative, with the aim to advance news literacy and improve trust in journalism; encompasses the Trust Project, funded by Craigslist founder Craig Newmark
- Journalism grants by the European Journalism Centre
- Arc Publishing, The Washington Post’s proprietary suite of tools, including their custom CMS
GEN Summit Competitions
As a GEN Summit tradition, three competitions crowned their winners during the conference: the Data Journalism Awards, Startups for News and the Editors Lab. The Data Journalism Awards, which took place at Vienna’s opulent Hofburg Palace, awarded a $1801 cash prizes — a nod to the estimated date of the creation of the pie chart — to a dozen projects including the Electionland Coalition in the “News data app of the year” category for its coverage of the 2016 U.S. elections.
In the Startups for News competition, a global pitching contest with eight finalists, Flourish took the top spot. The London-based news startup lets journalist turn spreadsheets into interactive content, including data visualisations, maps and explainers. It caters to non-coders and newsroom developers alike. The runner-up, Urbs Media, is an editorial agency that combines human-authored data journalism with automation to deliver data-driven stories.
Finally, the Editors Lab, a series of global hackdays, culminated during the summit after a three-day hackathon. Team BBC Visual Journalism fenced off the competition with its prototype Appy Helper, a “conversational interface that will sit within long-running complex stories that can help users catch up with recent developments or learn all about the background to this story.”
Benjamin Bathke is an entrepreneurial freelance journalist covering media innovation, startups and intractable global issues for Germany’s international broadcaster Deutsche Welle, as well as several other international publications. In 2015-16, Ben was a Global Journalism Fellow at the Munk School of Global Affairs and a multimedia storyteller for Washington University in St. Louis. You can follow the 2017 Reynolds Fellow on Twitter and see more of his work here.