How News Corp is using UX design to create meaningful change at scale

How News Corp is using UX design to create meaningful change at scale

News Corp started web publishing in 1988, and at the time, the online news division didn’t really have a sustainable business model. It was simply a value-add to print. More recently, we’ve seen three key business models emerge: paywall - which commercialises the customer; targeted advertising; and exchange of data.

Nigel Macquet, Head of User Experience and Design at News Corp opened his talk at the ADMA/AIMIA Techmix conference  on meaningful change at scale, by saying that there wasn’t really a commercial plan for online when things first kicked off.

He said that it was through the development of these more commercial models that online publications have started to track and analyse user data. More recently News Corp started developing experimental frameworks to test machine learning and automation to help with customer experience and user experience design across products. 

Macquet explained that massive media disruption and diversification that has brought us all the way from print to desktop, then to mobile and now to smart tech e.g. voice activation, have made it harder to track the user. There are ever more variables in customer experience and user experience, and with every new entry point comes a new user journey. The more this diversifies, the harder it is to understand behaviours, motivations and psychological concepts and respond with an appropriate experience.  

With Macquet’s team responsible for 126 content sites altogether, his department serves just under six million readers nationwide. Their primary goal is to enhance engagement.

“Everyone is fighting for engagement. This will become tomorrow’s currency - it will be quantifiable and it will be tradable.”

To that end, Macquet explained that “Machine learning can help us to track customer behaviour and serve up the right experience at scale.”

News Corps has started experimenting with such machine learning to dynamically respond to some key variables that impact their users’ experience of the site. Relevance, time, context and value are four such variables, because users want content they’re interested in, at the time they demand, in relevant form, and they don’t want to pay for it.

Macquet went on to explain how he uses these four constructs to create better user experiences.


If you know when your user wants to engage with your content, and for example, that they don’t want to see the same content at night as they do in the morning, you can use smart logic from machine learning to move content off the site and serve up fresh, relevant articles for the user. This prevents the site appearing static like it would be in print.

Ranking content assists with this. News Corp’s media outlets can have personalised blocks of content on the page, and display only seven of 15 articles that are actually there, at any one time. The rest is used as personalised content for a particular cohort. This means they different stories are served up in the evening.


The media industry is challenged with the ability to continue to produce high value content to increase propensity to engage. While print media is 100% curated, the online environment gives News Corp a chance to combine curation and automation to show users what the media outlet wants them to see - and more of what they want to see.

Machine learning can collect implicit data by monitoring user interaction with content and use it to generate predictive modelling. This helps to continually optimise editorial content for different cohorts.


User behaviour varies hugely depending on the user’s pathway. For example, if a user clicks on a newsletter from their inbox, they will probably know the site, the brand and be engaged with multiple articles as well as the editorial agenda of the brand. However, if the user comes to an article via search, News Corp can assume they will be more interested in the article itself, and not necessarily the brand. Then, if the user comes from social, they have a 90% chance of bouncing - they simply find something they want to see, and then they’re immediately off to post it to friends and family.

Macquet’s point here, is that with three different contexts of the very same user, the site needs to be smart enough to respond with the appropriate format of the article, e.g. giving a summarised version of the article when the user comes from social.


"Nobody wants to pay for news," explained Macquet. He went on to explain that the perceived value of editorial content is low, because there is plenty of supply. However, exclusive content including videos and interactive experiences that keep high engagement stories alive can lead to a higher propensity to pay. 

Additional to this, different types of content might be tiered in their value to the user.

Journey to a better customer experience

While Macquet admits there is still a way to go, he is optimistic about the value of machine learning in creating a more engaging experience for customers, at scale.

“If we can track customer behaviour, we can group customers into cohorts that share certain patterns of behaviour, then match marketing, sales and pricing to these cohorts.” 

Looking to get more sophisticated with your marketing automation, but the thought of integration is making you break into cold sweat? Never fear! Download our No-nonsense guide to streamlining and integrating marketing technologies today!

The no-nonsense guide to streamlining and integrating marketing technologies