NewsUK - Leveraging Data and AI at News UK

Joanne Kuai 22, February 2019 | London

NewsUK – Leveraging Data and AI at News UK

//
Posted By
/
Comment0
/

Joanne Kuai

22, February 2019 | London

The Times and the Sunday Times now as 508,000 subscribers, 263,000 of whom have digital-only subscriptions. The News UK-owned publisher has constantly leveraging Data and AI in driving subscriptions, delivering personalised content to subscribers and reduce churn rates. Michael Migliore, Head of Customer Value, News UK says in the past 24 months, they have achieved a 10-percentage point reduction in digital churn rates.

“Here, we believe in data-informed but not data-led. Data should help us to under our readers, who they are and what resonate with them. It helps with augment editorial judgement but not replaying it. The journalism, the content is still the key,” says Mr Migliore.

Dave Dale, Head of Data Science, News UK, introduces there are 15 data scientists in News UK based in Bangalore and London. They have highly achieved academics specialising in computational statistics, algorithm design, computational linguistics, statistical genetics having work backgrounds in self-driving cars, investment banking, online retail, search engines, media and also entrepreneurship.

They’ve spilt the Data Science team into central DS team and project data scientists. While the central DS team monitors and launch improvements to existing models and provide tools to standardise the model building process, the project-based data scientists are embedded in cross-disciplinary teams to “find the product fit”. Now, they can launch a model within three weeks of the start of a project, but they are still constantly experimenting with the process.

Within the newsroom, The Times’ house analytics platform serves the outputs of machine learning models to journalists and editors and is used in news conferences to discuss which stories have performed.

Called INCA (Intelligent Newsroom Contextual Analytics) the platform is designed to support the exposing of our data in the most appropriate ways given a journalist’s context.

“The first way we have chosen to do this is via a Chrome extension that displays data about an edition or an article directly next to it in the browser. This helps answer some of the most common questions quickly and most importantly in a way that makes editorial staff self-sufficient,” says Nick Petrie, Deputy Head of Digital, News UK.

The experts were speaking to a room of media practitioners across the globe at a study tour to News UK’s headquarters in London on 22 February 2019. As part of the 9th Data & AI for Media Week, the tour aims to inspire participants to imagine how to earn revenue with AI-driven products for their own media companies.

The Data & AI for Media Week was created by World Newsmedia Network in 2013. Since then, the conference has been held in London five times, Hong Kong once and twice in New York. The 10th Data & AI for Media Week is planned for San Francisco and Silicon Valley for 29 April to 3 May 2019.  https://sanfrancisco.dataaimedia.org

Leave a Reply