JUST RELEASED: New report details practical ways media companies are using GenAI, click here to download

A Personalized News Experience
2024 Finalist

A Personalized News Experience

Bergens Tidende

Bergen, Norway

Category Readership and Engagement

Media associated with this campaign

Overview of this campaign

Bergens Tidende automated the frontpage in 2019, with over 80% of teasers automatically selected. In 2021 we introduced segmentation, with subscribers getting a front page focusing on engagement, and non-subscribers focusing on conversions. Since then BT has increased the digital subscriber base by 50%, with the front page as the main driver for growth.

However, the algorithms can be better, and we identified these user needs:

  1. User frequency greatly differs. How can we be as relevant as possible both for the frequent and infrequent user?

  2. User interests greatly vary. Some users love sport, some don't. How can we accommodate all without sacrificing our societal mission?

  3. Our target group is younger than our main audience. With algorithms tailored towards engagement, how can we avoid the bias that the majority bring along?


To address this, we believe in a product that
adapts more to your needs than everyone else.

In 2023 we introduced:

The topic ranker, that looks at each user's reading pattern on a section level, and recommended similar stories. We combined this signal with our other rankers, and increased the CTR on stories on the frontpage with 9%. This is the highest lift ever on an A/B on the frontpage.

It is now released for all subscribers. The ranker was developed by Schibsteds Curate team, and iterated by Bergens Tidende.

Our ML initiative using sBERT, co-developed by MediaFutures and BT/Schibsted, to provide similar articles based on the sentiment and content of what you already read. We are using a Norwegian language model developed by the national library. 

This model is trained by the AI lab of the national library on a wide variety of Norwegian text – both Bokmål and Nynorsk – from the last 200 years and further finetuned for the purpose of calculating textual similarity. In A/B tests comparing it to our old recommender system, we saw a consistent lift of 20%. We are currently moving this from proof of concept, to a system that can scale.


Results for this campaign

2023 has been a tremendous year for Bergens Tidende, in part because of our investment in better recommendations and personalization.

We demonstrated significant results by serving content more tailored to each user, and overall numbers are also record strong.

9% lift in CTR on the frontpage in A/B test with the topic ranker
20% lift in CTR on articles with sBERT vs tag based recommendations
30% article page views increase YoY
9% increase in daily active subscribers YoY
7% digital subscriber growth


Contact

To contact a company representative about this campaign, click here for the INMA Member Directory

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.
x

I ACCEPT