top of page
elinstueland.jpg

Elin Stueland,
Project Leader AI,
Stavanger Aftenblad

February 7 2024

Reaching young readers: How Stavanger Aftenblad covers junior football like it’s the Champions League

Looking for new ways to attract younger readers, Norwegian regional title Stavanger Aftenblad (part of Schibsted) decided to create a project writing about them rather than just for them. Combining the latest automation technology with more traditional reporting on the ground, the football vertical Mååål! (goal), has a reader promise to cover junior football like it’s the Premier League. A text robot, co-created with Norwegian news agency NTB, takes care of turning all match results into articles, while two dedicated sports journalists report from some of the matches, doing interviews and showing up for the junior football community. The result has been higher engagement – every match has a small group of really involved readers. And thanks to the reporting on the ground, Stavanger Aftenblad is building its brand with a new generation of readers.
 

By Cecilia Campbell

Background and context of the job to be done

Stavanger Aftenblad is the biggest newspaper in its region, on the south west coast of Norway. The paper has 65,000 subscribers and get about 1 million daily pageviews, 70% of which is from mobile. The average age of print subscribers is 70 years old. “We know they get exactly what they look for from the print product,” said Elin Stueland. “But what about all the young people in the region – who are never the center of attention, except with their family and friends?” This question was the starting point for the project to launch a local football vertical, to cover all the clubs so many of the young people train with every week. “We decided this was an opportunity to do branding of our newspaper for the future.”

What did Stavanger Aftenblad do?

The product, called Mååål! (Goal!), is a football portal, which exists as a section on the Stavanger Aftenblad website and also as its own URL. It includes all recent matches – 80-90 a day – and readers can search on their team/club or on the series or tournament they are interested in. 

Mååål! covers all local football played by players from 13 years of age: “This is how we reach straight into the lives of our young readers – 10,000 players get mentioned in the largest newspaper in the region,” said Stueland. The promise to readers is to cover local football as if it were the Champions League.

To do this, the content comes from two different sources.
1. Robots, built in collaboration with Norwegian news agency NTB, which write automated match reports based on structured data from the Norwegian Football Association. Every match is covered this way.

2. Two full time reporters, one of whom is the project manager for Mååål! The journalists cover breaking news like new players, interview players and coaches, cover some matches live, and do live streaming of 80 matches a month.

How did they do it?

A significant part of the project was to get the robots in place, and in order to do that, correct match data was needed. The data is gathered through a process of coaches putting it into an app run by the Norweigan Football Association. Match data includes team lineups, goal scorers, expulsions and substitutions. From the app the data is transferred to a website which the robots access through an API.

The texts are then generated using NLG (Natural Language Generation), a type of AI that is based on conditions to build the story. In other words, the texts are put together from prewritten text segments, which are activated depending on the data. The texts are constructed based on a certain logic. The main event, like a hat trick, is included in the headline. Under that is a visualisation of the match result (scores for the teams under the respective team logo), followed by one paragraph about the first half of the match and one about the second. The final paragraph includes information about the teams’ standing in the series and who they will play next time.

 

“This technology creates narratives from structured data,” said Stueland. “This significantly predates generative AI, and is a tech that gives us 100% control of the outcomes. There is no risk of hallucinations, in other words.”

Who did it?

Elin Stueland presented what she called the Stavanger Aftenbladet “Starting 11”, who got the project off the ground. The team, which included some external people from the Schibsted group, was made up of people with expertise in tech, UX design, commercial opportunities, editorial strategy and football. The Starting 11 also included two people from NTB working on developing the robots.

NTB’s role was key – Stavanger Aftenblad would not have been able to develop the robots in-house. “NTB had had this idea for a long time, but as a news agency they lacked the forum to reach readers. So it was a perfect collaboration – they built the tech, we did the product and project. We actually started it the day before the pandemic hit, so we had a lot of time to sort it out. And then there was a ketchup effect, once people were able to play football again – it was a great success from the first day."

What were the project KPIs?

Mååål! had a couple of key KPIs from the outset. One was retention of paying subscribers, which was particularly significant in the first couple of years – readers who converted on the football content stayed subscribers a lot longer than the average local subscriber.

“The second KPI was around attracting young readers,” said Elin Stueland. “We had done a lot of projects around what was going on in the city, young people’s nightlife, events and so on – the extraordinary things in their lives. But the football is about their everyday life, which we hope is even more powerful.”

In terms of revenue, in addition to subscriptions, Mååål! has attracted a sponsorship by Coop, one of Norway's biggest food retailers.

What have been the results?

“The reception of Mååål! has been exceptional. Readers feel like local football is being taken seriously by their local paper,” said Stueland. While the average age of the printed newspaper is 70 years and the average age of the digital subscribers is 52, the average age of people who have converted on the football content is 30. And of course, many of the readers are much younger than that. “Since most young readers have their Stavanger Aftenblad subscription through their parents, our main goal is to build our brand for the future. These youngsters get a relationship with the newspaper from a young age, and will hopefully become subscribers when they are older.”

Key take-aways and tips

Elin Stueland gave three key tips for local publishers thinking about launching a similar type of project.

  1. Find a content area which has data, but that is too comprehensive to cover through reporters. “This is not about replacing journalists, it’s about adding coverage that you didn’t provide before and which can add value both in terms of engagement and actual revenue.”

  2. Find a structured data set, that’s key for reliable and predictable results. “For the football, we needed to do some work to get it structured. We worked with the coaches to get them to put in all the data points from the matches. In the early days, some thought it would be fun to write that they’d won 30–0, but once you see that written on the site, it’s not so fun anymore. As time has gone on they see that it’s important to respect the teams and players, and to do their best. If the coaches don’t put a lot in, the text ends up being short, and in those instances we write that it’s because we don’t have enough data. That also motivates everyone to do a good job.”

  3. Combine robots and reporters. “A key success factor for Mååål! has been the comprehensive match coverage provided by the robots, combined with the colour, flavour and spice produced by our journalists. Our reporters care more about 5th division local football than they do about Manchester United or Real Madrid and that is key,” said Elin Stueland.

What is next?

Using a similar technology to the one used for the match reports, Stavanger Aftenblad also publishes automated content about real estate sales and companies’ annual results. “Next up is an automated weather product – in Norway, we’re obsessed by the weather, in particular any chance of sunshine,” said Elin Stueland.

Mååål! has proved itself and will live on as an important part of Stavanger Aftenblad’s local journalism. “It’s a great success and such a good way to engage with our young people. Local young football players now count down until they reach 13 and will get mentioned in the paper,” said Elin Stueland.

Useful links and contact information

Here are some useful links:

  • The case is about a service at Stavanger Aftenblad. aftenbladet.no

  • A direct link to the service itself at Stavanger Aftenblad.: Mååål!

  • Stavanger Aftenblad is part of the Norwegian publishing group Schibs´ted

  • The sister newspaper Bergens Tidende, also in the Schibsted group, has a similar service called "Lokalfotball"​​

E-mail address to Elin Stueland who presented the case and can reply to questions

We also want to mention a similar service reaching out to passionate local fotball fans which can found at the french newspaper La Voix du Nord in northern France. A full description of this case can be found in the WAN-IFRA Report from 2022 "Building and engaging specific audiences: Lessons from the third round of Table Stakes Europe" which can be downloaded here. The specific case is described on page 25-27. Table Stakes is a WAN-IFRA led programme in partnership with the Google News Initiative Digital Growth Programme created to help establish and grow the online business of news publishers who have more recently started developing their digital platforms. The program has had four rounds up til today (in February 2024). Here are two links to the case:

You can also contact Cecilia at the WAN-IFRA Innovate Local team if you have questions: c.campbell@wan-ifra.org

 

If you, dear reader, know of other relevant cases that are similar, please let us know, We will add information here. Thanks in advance!

bottom of page