Reaching young readers using AI
With national and international news now so readily available, a challenge many local publishers face is competing for this engagement, particularly when it comes to younger audiences.
One local newspaper in regional Norway decided to address this issue through an unlikely candidate; AI.
Stavanger Aftenblad is a daily newspaper and online publication, concerned with their social mission, critical journalism, breaking news, and a focus on freedom of expression. The established news publisher has 63,000 subscribers, but with the average age of subscribers being 70, Stavanger Aftenblad decided they needed to find an innovative way to brand the newspaper for the future subscribers; young people.
A team designed the MÅÅÅL soccer/football portal, covering all matches across an entire region for players aged 13 and up. While this may seem an impossible task for local newspapers with just a handful of staff (or less!), the strategy is not entirely out of reach. Because Stavanger Aftenblad’s MÅÅÅL articles are generated by a robot…
Innovate Local, a global program for local news to share concrete and actionable ideas, hosted a webinar with the AI Project Lead Elin Stueland to unpack the project.
For busy folks who don’t have time to watch, here are some key takeaways about MÅÅÅL from the webinar:
How does it work:
Coaches or team managers add data to an app following the completion of a match, including penalties, goals scored, players and substitutes. This data is then converted into a news story for each match; 80 to 90 football matches a day throughout the season, with 10,000 players mentioned and featured on the MÅÅÅL (‘GOAAL’ in English) football portal.
How did they do it:
Stavanger Aftenblad teamed up with the Norwegian News Agency (NTB) to build an AI in natural language generation that could take data submitted by coaches or team matches after a junior football match, and convert this into news stories. The project was developed by a team of 11, and included building codes that convert data such as the timing of the first goals scored, into phrases like “take an early lead”.
While a lot of manual work was required initially to systemise the data, the program now operates using the one robot and two journalists, who contribute with critical journalism, breaking news and human interest stories, engage with clubs and teams, and oversee the articles produced by the robot.
What was the outcome:
To work effectively, a project such as this requires significant uptake from the community to enter the data required for the articles.
Elin points out that coaches and team managers were quick to engage, seeing the morale boost players would get from having their game reported on in the paper. With young players, parents, friends and grandparents checking in to read about their favourite players, the MÅÅÅL football portal’s average subscriber age is 30 years, far below the newspaper average. The team’s goal is to foster this relationship with younger people, branding the newspaper for the future.
Tips from the Project Lead:
While a project to this scale may seem out of reach to many small local publishers, a great idea that seems likely to generate widespread community engagement could very well attract the funding or partnership opportunities to make it happen.
Here are some tips from Stavanger Aftenblad’s AI Project Lead Elin Stueland:
- Find your journalistic area that is too comprehensive to cover for the reporters (so as not to take away a job, but rather add an area previously uncovered).
- Find a structured dataset that a robot can be taught to interpret.
- Combine robot journalism with a human interest journalist.
Find out some other ways AI is being harnessed to help busy newsrooms here, or watch the full webinar from Innovate Local: