Best Ever Growth for The Times & The Sunday Times Thanks to Usable Data Science
Media associated with this campaign
Overview of this campaign
Our overall objective was to help safeguard the newsroom and investment in quality journalism, offsetting challenges in advertising, print circulation and distribution. Essential to this objective was an understanding of individual subscribers, with a data-driven lifecycle to revolutionise retention rates. We wanted to reduce churn rates by 25% to best ever rates, and in turn generate an incremental £1.2m.
Our data modelling required a fair way to score reader engagement based on relative impact each key variable had on churn using 6 years of data. From here we wanted to ensure this information was engaging and usable for marketing, editorial and in the call centre.
From understanding engagement and churn, we wanted to utilise our data scientist and produce a churn propensity model, enabling us to assess the 'vitality' of our base and gauge which aspects of our offering had more potential.
Following this, we wanted to create the ability to serve our customer the right message for the biggest return. For example Times+ our loyalty programme releases over 200 offers a year but only used by 20% of new subscribers, whose cancellation rate is four times that of an in-life customer. Another focus area was surprise and delight, to see how we could us to both retain high-risk customers, and simultaneously 'judo-flip' them into annual contracts.
An additional objective was to use 'non-traditional' channels across social and display to engage new audiences not active on our digital products or email channels.
Furthermore, our final ambition was to go one step further and bring automation into execution, serving the right message, in the right format, at the right time. We sought and secured investment via the Google DNI fund to develop a Journey Automated Machine for Higher Engagement through Self Learning or JAMES for short.
Results for this campaign
Our data-driven, score-led approach has increased our digital subscription revenue by 10%. Churn rates are at best ever rates, now just 60% of the rates of 2015.
We created a churn-propensity model, correlated with an engagement score, identifying frequency as the best indicator for vitality. Volume of articles read per visit was the second biggest influencer, followed by recency between visits and finally interactions including commenting, sharing and engaging with Times+. These four variables spell the acronym FAIR and this has been trained out across both editorial floors and beyond. Our data modelling enables us to assess the FAIR score for segments of subscribers, resulting in more scientific churn forecasting which in turn has given our board heightened confidence in our outlook. Our FAIR score is up 8 points to 73 and our gamified lifecycle is focused on nudging readers up to their optimal level based on their 'look-a-like' audience.
Our dynamic on-boarding journey consisting of 14 personalised touch points has reduced early life churn by 9%, resulting in £0.8m incremental revenue.
Activity across social and display has resulted in a 16% churn versus control aided by continual testing and segmentation.
Times+ has become a key tool for converting trialists, with users converting at 93% and now reaching 28%, up 8 percentage points safeguarding 45,000 subscribers. Data analytics has informed what offers and events we launch, and enabled us to track relative impact.
Surprise and delight has enabled us to move 12,000 high churn risk subscribers into 12 month full price contracts by generating an incremental £0.6m net, but has created a on-going challenge now expectations have been set with this group.
We now have a data-centric marketing division, which has become extremely effective and an embedded data function.
JAMES has revolutionised churn further, and advisors informed by readers interests underpin an award winning contact centre.