Front Page Reinvented: The algorithm that supports our journalistic mission - and creates a smarter, more engaging and informative product
Media associated with this campaign
Overview of this campaign
One year ago we were facing a basic challenge well known to many publishers around the world: Our users differed greatly in visitor frequency and preferences, but we were still serving the same “one size fits-all product” to everyone of them. The result? Our “news junkies” were exposed to the same stories over and over and again, while the more infrequent users missed our best and most relevant stories because they could no longer be found on our most important user interface; our digital front page. In other words, we were not catering well enough to meet our user’s needs. We therefore set out to do the following: build a more relevant front page that takes into account user frequency and preferences.
This could have been a straight forward algorithm optimized for engagement, but we wanted to do something more: we wanted to create a product run on sophisticated algorithms that would steer clear of so-called “filter-bubbles” and “echo-chambers” so often associated with Facebook and other algorithm-based news products. Instead we wanted to build a news-algorithm based on our journalistic mission, and one that actually closes the gap between what people know, and what they should know. We had four main objectives when we started on the journey:
We wanted to create a more relevant and engaging product for our users.
We wanted to build a product that better supports our business-model of digital subscribers.
We wanted to make sure more of our users are exposed to a higher humber of our best stories.
We wanted to release a product that enables us to speed up product development and build completely new and enriching features.
Results for this campaign
Most importantly, we have demonstrated that it is possible to make use of the great advantages of data science, algorithms and machine learning in news, without creating “filter-bubbles” or “echo chambers”. It is our responsibility as a trusted news brand to enlighten public debate and create the foundations for a better working democracy. We are doing that by combining sophisticated data and algorithms with editorial judgement and ethics.
One year in, the results of the project have been extraordinary:
We have created a much more relevant front-page, illustrated by a 10 % lift in stories read from front page. We have built a product that better supports our business objectives, illustrated by a 2% lift in conversion-rate. We are also starting to test different front-pages for subscribers and non-subscribers in order to drive more subscription sign-ups and engagement. We are now also able to show the most important stories to users who have missed them, by using meta-data from each story and our algorithm. We can also remove stories the user has already read. Finally, the combination of relevant meta-data and a sophisticated algorithm enables us to build a wide set of different products to better cater for user needs. This could be the most important stories from a favourite section or writer, stories they should not miss from a subject the ought to know about, stories we think will challenge their world-view and so on. The opportunities are almost endless.
In addition to all of the above, a data-driven front page such as this enables us to spend less time on “pixel-perfect editing”, instead focusing on improving quality of our stories and distributing our content. Eliminating legacy workflows is therefore a very positive secondary effect.