AI Technology for Subscribers Growth

JAMES is an innovative technology for subscriber retention that pioneers the use of machine learning for distribution of news content. Powered by a combination of algorithms,  JAMES leverages the rich content data produced daily by newsrooms, as well as the vast amounts of reader behaviour data collected by publishers, to produce personalised recommendations for each individual user with the aim of increasing their engagement.

JAMES’s first prototype, developed together with News UK and Google DNI, showed a 49% cancellation rate decrease by sending 14.1 million personalised emails to a cohort of 117.000 subscribers of The Times and The Sunday Times (read more about the project below).

How Does it Work

The aim of JAMES is to help increase retention of subscribers by tailoring distribution for each user. To do so, JAMES exposes readers to different content and measures their reaction to understand what triggers them. After the algorithms are trained with this data, JAMES generates lists of recommended articles per reader that publishers can use through different distribution channels providing indications of best suitable trigger time and format.

Twipe opened up the Launch Partner Program in Q4 2019 that will release JAMES algorithms to more publishers with the aim of further development.

Book an Introductory Call

We are currently selecting two partners for the second roll-out of JAMES starting in April. If interested in knowing more fill out the form and our team will be in contact shortly.

[wpforms id=”46225″ title=”false” description=”false”]

JAMES – The Project

JAMES was born as a collaborative project between Twipe and The Times of London. After receiving Google DNI funding, a group of 22 people worked for a year in fine-tuning a product that could help accelerate digital subscriptions growth. Over 100.000 subscribers of The Times were served with individualised newsletters compiled from the content of a daily edition. Several optimisation algorithms were developed and tested to achieve the best results.

Google DNI
The Times

The Results

Throughout this one year project we learned that there is a high potential for the use of AI to create better customer experiences, by creating reader engagement with automatically  generated emails. JAMES continues to be used and developed at The Times and The Sunday Times and is now being made available through the Twipe digital publishing SaaS platform.

14,1 Million Emails Sent

70% of the Exposed Readers interacted with JAMES

49% Less Churn on Exposed Cohort

Highest Impact on Readers with Low to Medium Engagement


Sign up to be notified of the latest developments from JAMES.

Further reading