When robots drive inclusive journalism

Update: The Q&A function won a third place in the Best New Technology or Digital Product category of the 2020 Global Media Awards.

Talking to a BBC journalist the other day, I was laying out the drivers behind newsroom automation at Swedish local media groups. We agreed it sounds somewhat contradictory that robots now play an active role in creating journalism that involves hundreds – thousands – of people that were rarely seen in local news stories before. I was talking specifically about our Sports robot’s Q&A function, which – we’re very happy to say – has been nominated in INMA’s 2020 Global Media Awards.

In short, this is what happens: Almost all Swedish local news groups publish automatically generated post-match reports of all games in all divisions in all sports, using United Robots NLG based Sports robot. The new Q&A function adds a human element. After writing the match report, the robot sends relevant questions via text message to team coaches. The coaches' quotes are automatically inserted into the articles published on the news publishers' sites.

“There had been earlier attempts at using tech, like bespoke apps, to get teams involved in the stories,” Henning Johannesson told me. Henning was Head of Sports at Mittmedia when Q&A was launched (he’s now my colleague at United Robots). “I think the processes were too cumbersome though, and nothing ever really took off. I confess was very sceptical it would work this time either.” But it seems the simplicity of using automatic text messages has won through. Since launch in 2018 and as of January this year, the Q&A function has included 602 coaches providing a total of 6272 published match comments. 536 teams have taken part, 28% of them women’s teams. The articles have been published across some 30 Swedish news sites. 

There’s plenty of anecdotal evidence, including posts on social media, that players and coaches appreciate the coverage they get in the local press through the Q&A quotes. There’s some more scientific proof too. In a survey among the coaches, of the 180 who responded, 90.5% said that the function enabling them to comment via text message after matches is either “good” or “very good”.

News publishers are positive too, not least because Q&A enables more and better coverage in small local communities, which are sometimes underserved due to restricted newsroom resources. Says Bärgslagsbladet’s (part of the Mittmedia group) Editor-in-chief Helena Tell: “For a small local title like ours, with no dedicated sports reporters and limited staffing during evenings and weekends, the coaches comment function is a huge asset. We get match reports – immediately after the final whistle – which are comparable in quality to a story over phone by a reporter. The clubs and fans get the attention and the information they hunger for. The newsroom can instead spend our time chasing the ‘real’ stories, in other words content which we know a broader set of readers interested in sports want to read.”

The Sports robot Q&A function is not primarily about technology, it’s about people. To quote Nick Diakopoulos, author of Automating the News: How Algorithms Are Rewriting the Media: “Human values are at the heart of algorithms and AI and journalists need to think about how to design journalistic values into technology.” With Q&A, that’s exactly what we’ve done. The automatic interviews are a win-win for local sports teams, fans and publishers, as they help create local sports journalism which is both inclusive and engaging.