Case Study: Reuters achieves higher engagement and monetizes video assets with Microsoft Azure Machine Learning

International news service uses machine learning to increase website engagement

Reuters, the 160‑year‑old global news organization, faced a content‑matching challenge: thousands of video assets were underused because editors could only manually link a few videos to articles each day, missing the short 12–48 hour news window to engage readers. Reuters needed a fast, scalable way to surface contextually relevant videos alongside articles to improve reader experience and monetize its growing video library.

Using Microsoft Azure Machine Learning (with Azure SQL Database, App Service, and Python tools), Reuters built an automated recommendation engine that extracts titles, descriptions, and transcripts, scores candidate article/video pairs, and updates matches every 15 minutes. The system places the top matched video in a related‑coverage pane on articles, driving a 4% increase in video views, average completion rates above 80%, growing editorial trust in AI, and new engagement and revenue opportunities.

Reuters

Isaac Showman

Managing Director