Bonnier Publications
53% Subscription Growth Through ML-Powered First-Party Data
The situation
The challenge
Bonnier Publications was operating in a media market under pressure — declining readership and rising competition from digital platforms. Subscription sales via digital channels were underperforming, and marketing budgets were being spent broadly without precise targeting of the most purchase-ready users.
What we did
We consolidated historical customer data from CRM, Google Analytics 360 and the newsletter platform into Google BigQuery. On top of this data infrastructure, we built a Machine Learning model predicting the likelihood that a given user would purchase a subscription. The model segmented 33.3 million users into 5 actionable segments and activated them directly in search and display campaigns.
Results
After just a few months, the project delivered measurable results: 53% growth in digital subscription sales, an 18% increase in conversion rate, and 327 million predictions generated. A critical side effect was the ability to shield existing subscribers from new acquisition campaigns — reducing wasted spend and improving customer experience.
This project was delivered by the team behind Dear Future, formerly operating as IIH Nordic. Winner of the FDIH E-Commerce Award 2020.
The approach
Data consolidation in Google BigQuery · ML model: subscription propensity scoring · Audience segmentation & activation
Data consolidation in Google BigQuery
Linked CRM, Google Analytics 360 and newsletter data into a single Google BigQuery data warehouse, creating a 360-degree view of every customer.
ML model: subscription propensity scoring
Built a Machine Learning model to predict the likelihood that a given user would purchase a subscription — trained on historical purchase behaviour.
Audience segmentation & activation
Divided users into 5 segments and pushed them to search and display campaigns. Existing subscribers were shielded from new-customer acquisition messages, reducing cannibalism.
The results
Numbers that held up after the project closed.
In a market where most publishers were losing subscribers, Bonnier grew digital subscription sales by 53% — entirely driven by smarter audience activation, not increased ad spend.
Delivering the right message to the right user at the right moment lifted conversion rates by 18 percentage points across digital campaigns.
The ML model scored every user across 327 million prediction events, enabling Bonnier to act on real-time intent signals at scale.
33.3 million users were classified into 5 actionable segments — including a do-not-disturb segment that protected existing subscribers from being targeted by new-customer ads.
Bonnier has created one of the strongest data frameworks in the Nordic media industry.
The pattern
Bonnier's pattern applies
to every organisation.
It might be email routing, invoice processing, support ticket triage, form classification, or customer enquiry handling. Work a person does because the system hasn't been taught to do it yet.
This case is relevant if your team...
- First-party data is the moat — Publishers sitting on years of CRM and newsletter data have a significant advantage — but only if that data is connected and activated.
- Protecting existing customers is as important as acquiring new ones — The ability to exclude current subscribers from acquisition campaigns reduced wasted spend and improved brand experience simultaneously.
Is your team doing work like this?
Tell us the problem. We'll tell you if we can help.
We're direct. If we've solved something like this before, we'll say so. If we haven't, we'll say that too. Either way — the first conversation costs nothing.