Digitisation has changed almost everything — and very quickly.
Annie Lidesjö, head of operations and people at NTM in Sweden, said many of the most successful tech giants use automation and data to create better services for the consumer. Lidesjö mentioned Facebook, Google, Instagram, Netflix, and TikTok.
“How can the news media do this in a good way, while still keeping our mission to keep a free and independent press?” she asked during INMA’s Leadership and Transformation Opportunities Master Class in January (future INMA Master Classes can be found here)
Lidesjö pointed out a few things:
The industry already use a lot of automation in news media.
Programmatic is one, that has changed the entire advertising landscape.
Consumer data analytics have changed strategy.
New devices and tech have changed consumer behaviour.
The industry will continue to automate more to find a sustainable business model.
COVID-19 has changed almost everything, and users and customers will have different perspectives moving forward.
She presented a graph that showed the value relative to the difficulty of bringing automated analytics into news media publishing, stressing that creating data-driven organisations is a good start to taking full advantage of automation.
“We need to believe in the data and have one true version for everyone in the company, and clear goals on what to achieve,” she said. “If you do this right from the beginning, you will have so much from that later on.”
Piotr Turek, head of data products at Ringier Axel Springer, said building an organisational foundation around data that can support automation and AI initiatives requires addressing fears around automation, which are usually around fears of singularity and making humans obsolete.
“Even though some of these fears are actually true and relevant, in my opinion, the biggest problem we see today with AI is an education gap,” he said.
Everyone in an institution is not equipped with the tools to use or even understand the underlying tech of AI, Turek added. This lack of understanding compounds with hyperbole in the media and fear of the unknown to create denial and hostility toward new technology.
The true risks of automation and AI, Turek said, lie in failure to embrace them. Social media giants using data and machine learning to a clear advantage are evidence of this.
The challenge of incorporating automation is worth it. Ringier Axel Springer’s flagship project was to create an automated, personalised home page for Polish news portal Onet. The brand now distributes 4x more articles daily, increased engagement on widgets by 50-100%, and increased loyal users by 24%. Also as a result, Onet was deemed the most cited and trusted media brand in Poland. The whole transformation, the elevation of quality content and audience trust, happened simultaneously.
“There is a common misconception that automation is about removing the human element,” Turek said. “We proved this was wrong.”
Automation is also amplifying journalism and making life easier for the newsroom at the BBC. Nathalie Malinarich, the company’s digital development director, said automated content production is her favourite way the company is using the technology.
“It enables us to do more with the journalism we have, to make it travel further — but also it does allow us to provide a better service, just being able to reach people with stories in the way they like them told.”
The team relies heavily on bots for this facet of automated production. For example, around elections the bots are used to automatically produce graphic cards related to the election.
“Robojournalism I’m sure is familiar to most people,” Malinarich said. “We have a programme called SALCO, which is semi-automated local content.”
Basically, she explained, SALCO takes databases (such as crime or health care) and pipes the data into story templates. The results include a short text story and a graphic to accommodate it. These graphics can also be used in other ways such as social media and blogs.
“However, we’re not getting rid of journalists to do that,” Malinarich assured. Journalists still must verify every story that is auto-generated, but the overall process is much less resource-intensive than the old, completely human-powered way.
Empowering staff with data and tech capabilities is the future for news media. Dagens Nyheter in Sweden is working to build healthy data foundations into the company’s organisational structure. Lovisa Bergström, data scientist and analytics team lead, said choosing the right analytics tools is a key foundational decision. Becoming data-informed creates the risk of data-overload.
“We want full control, but it’s very hard to keep several targets and KPIs in mind at the same time,” Bergström said.
Old editorial tools and dashboards no longer served current teams and goals, so Bergström and a team of four other data scientists spent two days a week for six months researching, surveying, designing, and developing new dashboards.
Interviews with staff told Bergström and her team that the new dashboards had to be in Swedish, update in real-time, make it easy to change KPIs, be easy-to-use, and scale to work in the company’s multiple newsrooms.
The dashboard is now rolled out across 49 Bonnier news brands with about 1,000 users each week. By building an internal tool that allows newsrooms to cancel their expensive analytics subscriptions, Bergström and her team save the company about €1 million per year.
Bergström’s team onboards users so they can take full advantage of the tool, and they also receive feedback for improvements from users. One key lesson over the whole project has been that audiences change over time, and systems become outdated if they do not evolve with the needs of users. Onboarding really is critical, she said: “Tools and KPIs are useless if no one knows how to use them.”