There are a wide variety of different ways AI can be implemented in a newsroom to help reporting, audience growth, engagement and more. Here we explore just some of the benefits local media organisations have seen from their AI projects in this extract from our new report, Practical AI for Local Media.
Doing more, faster
From data analysis to content creation, AI can do far more, far faster, than human reporters.
A 2022 audit conducted by LION Publishers, a US member organisation supporting independent news publishers, found that one of the group’s biggest challenges was ‘adding more editorial content without the operational infrastructure to support it.’
That’s exactly the problem McClatchy solved with its property bot. The publisher was creating content around property sales across several markets, but not on a regular basis. “If one of our reporters had time, we would have that information,” said McClatchy’s VP of Audience Growth & Content Monetization Cynthia DuBose. “But if it was busy, we might go weeks without. The robot has allowed us to have regularity and scale in markets where we didn’t have anybody pulling the information.”
In other contexts, speed of delivery is crucial and robots are simply quicker. Jens Pettersson, Head of Editorial Development at NTM used traffic reports to make the point: “We have one robot that works on traffic and it is so much faster than our own reporters. They will see an email in their inbox from the authorities telling them there’s been an accident. They need to put that text into the CMS and get it on our website. The robot is instant.”
The point with these use cases and the others we looked at is getting more content that matters to local audiences quickly. “It’s about getting more journalism in those areas that we know make a difference. That’s where the robots come in, they help us out with that,” explained Pettersson.
Practical AI is all about ROI. Stavanger Aftenblad was facing challenges in growing its subscriber base, with high rates of churn hampering growth. The introduction of junior football coverage allowed the team to deliver on the promise to deliver the best football coverage in Stavanger. Online Editor Elin Stueland said, “It’s an endless promise so they just want to stick with us. Churn was lower from the start.”
McClatchy has seen an increase in first time visitors, playing into ambitions for audience growth. DuBose estimates that two thirds of AI content readers are first time visitors. “Now we have an opportunity to drive them from that coverage deeper into our site and our reporter created journalism,” she said. “It definitely helps with the growth piece, because it allows us to reach people that typically might not come and interact with us.”
Looking at standard email KPIs, audience engagement is higher on NRC’s automated newsletters than for those that are manually produced. But the most important metric for NRC’s Data & Innovation Manager Luuk Willekens is digital activation. He said, “This newsletter was more of a retention use case than to get more subscribers.” His team monitored specific subscriber segments over 12 months and saw that people receiving the personalised email were more active on a weekly basis than on NRC’s website or app.
“We might go weeks without property sales coverage. The robot has allowed us to have regularity and scale in markets where we didn’t have anybody pulling the information.”
Cynthia DuBose, Vice President, Audience Growth & Content Monetization, McClatchy
McClatchy’s property project allowed the publisher to extend its coverage in more newsrooms. It also supported new product development. DuBose, tasked with growth and monetisation, said the property bot gave her enough content to think about new offerings. “I could say, ‘Oh, can we build a newsletter now?’ And that’s a product that advertising can go out and say, ‘Hey, we also have a real estate newsletter’.
Stueland at Stavanger Aftenblad talks about the promise made to readers to cover junior football like it’s the Champions League. Without AI, it would never have been able to report on 7,000 junior matches every season needed to keep that promise.
In a similar vein, Stavanger Aftenblad uses AI to produce business reports at a local level. Stueland said, “It might be really small businesses, but they have a great impact on their neighbourhood or city area. We’re closer to people with automatic journalism. That’s kind of thing I couldn’t think of that much before.”
Freeing up time
Everyone we spoke to for this report said that AI saves time for journalists, enabling them to focus on adding real value. NTM’s Pettersson said, “AI frees up time for human reporters to bring along their analytical mind. The robot can do the boring work.”
RADAR Editor Joseph Hook saw a huge amount of value to be brought through automation, simplifying processes, and doing a lot of the things in a local newsroom that ties people to their desks. “It’s freeing them up to go out and meet people… something that’s being lost a bit in journalism.”
Ahead of the arrival of 2022’s Hurricane Ian, the deadliest hurricane to strike Florida since 1935, McClatchy built a hurricane bot. It used AI to scrape data from the National Hurricane Center and quickly created update articles.
“That’s not really journalism,” said DuBose, “Our journalists reported evacuations, city preparedness, other efforts. That’s journalism that cannot be replaced by a robot.”
NRC’s newsletter automation has saved editorial time in newsletter production. Prior to automation, manual newsletter creation could take two to three hours including story selection, choosing images, headline tests.
“This time is not needed now,” said Willekens. “And with the personalised layer, readers are reading more articles than in newsletters picked by the editorial staff.”
“With the personalised layer, readers are reading more articles than in newsletters picked by the editorial staff.”
Luuk Willekens, Data & Innovation Manager, NRC Media
You don’t know what you don’t know and one of the things AI is really good at is surfacing stories that otherwise would have been missed.
Pettersson explained how NTM’s property bot acts as a story alert for journalists and Stueland agreed. “I never thought of AI making us more capable of doing the things that we’re really good at, but sometimes we don’t know they’re happening. We don’t read all those lists of real estate sales. It gives us the ability to get closer to the details of people’s lives.”
United Robot’s Chief Marketing Officer Cecilia Campbell spotlights the case of a robot reporting junior hockey for some sites in the US. “The robot picked up that one team, it was the first time they had won in 40 games. The editor didn’t know that, but the robot found it and turned it into a headline.”
At RADAR that results in national stories being written with local relevance. Hook said the problem with traditional newswires is that it can be difficult to make stories relevant to specific publications, especially on a local level.
“We can pick up, not just a data set of the top level numbers, but we can dig into those numbers. We produce a specific version for every area that data contains. So if that’s local authorities in the UK, we’ll be writing 350 versions of the story in England – 400 if it’s across the UK. We’re making stories which are very tailored and personalised to papers’ local readership.”
This article is an extract from our new report, Practical AI for Publishers, sponsored by United Robots. Download it for free below, and listen to our special podcast documentary episode featuring the voices and experiences of the report’s interviewees here.
Your details will be used to send you the Practical AI report, as well as information about future Media Voices reports and United Robots communications. Please note United Robots and Media Voices are joint data controllers for this report.