£6m in additional revenue.
Retention and marketing ROI, both up.

Marketing spend and editorial calls were made without the data to back them.

Marketing, commercial, and editorial teams were each operating without a data-driven backbone. Campaign effectiveness was hard to measure, so spend wasn’t reliably going where it worked best, and the business had no predictive view of which readers were about to churn or why.

Nothing was structurally broken, but every function was running well below what its own data could support, and revenue that better targeting and attribution could unlock was going unrealised.

Three constraints were non-negotiable:

• No reliable marketing attribution
• No predictive view of reader churn
• Editorial decisions made without data

One data science team, four models, three functions.

A cross-location data science team, split between London and Bangalore, was directed to build the analytics the business was missing. Churn analysis and customer preference models gave the retention side a predictive view of which readers were at risk and why.

Article popularity prediction models gave editorial teams a data-backed steer on what to run and promote. Attribution models mapped marketing spend to actual customer behaviour so budget could move to what was working, while recommendation engines layered personalisation onto the digital platforms themselves.

The output wasn’t just models — it was strategic insight fed directly into the marketing lifecycle, giving commercial, editorial, and marketing teams a shared, evidence-based view of the reader.

Six million pounds, and a data-driven culture to match.

The business moved from siloed, gut-feel decisions to a shared, data-driven approach across marketing, editorial, and commercial teams, with the models and culture in place to keep compounding that advantage.