India Today, one of India’s most influential media organizations, has launched an artificial intelligence initiative aimed at forecasting audience engagement patterns, marking a significant shift in how newsrooms leverage data to inform editorial decisions. The project, called Audipulse, was developed in collaboration with the 2025 edition of the Newsroom AI Catalyst program, a WAN-IFRA accelerator supported by OpenAI. The tool is designed to analyze historical engagement metrics and generate predictive insights about story performance, optimal publishing times, and content formats, addressing a long-standing challenge in digital journalism: the delay between data collection and actionable decision-making.
The initiative emerged from a critical evaluation of existing analytics tools, which, while effective at measuring past performance, struggled to provide forward-looking guidance. Editors at India Today reported that manual analysis of engagement data was both time-consuming and prone to subjectivity. “In today’s digital space, where people do not actively choose the source of news, and instead become passive consumers of whatever the algorithm pushes, it is important to know what your audience expects from you to keep them loyal,” said Bal Krishna, head of the Fact Check team at India Today. The organization sought to address this gap by integrating AI into its workflow, leveraging local GPU infrastructure to avoid reliance on external cloud services for sensitive data.
Audipulse functions by combining real-time engagement data from sources like Chartbeat and Google Analytics with draft headlines prepared for the following day. The system evaluates variables such as click-through rates, time spent on content, and topic-specific trends to predict how stories might perform. During a 15-day pilot, the tool achieved a 64% prediction accuracy rate, outperforming the 52% baseline of human editors. “AI is very efficient at analysing data and identifying trends,” Krishna noted. “Even if ample data on audience behaviour is collected by an organisation, it is extremely difficult to reach a definitive conclusion without the help of AI.”
However, the project also revealed the limitations of purely data-driven approaches. Early results showed that incorporating contextual taxonomies—such as categories like “cricket,” “elections,” or “Bollywood”—increased prediction accuracy by 11 percentage points. This highlighted the importance of aligning algorithmic insights with cultural and social nuances. “The biggest concern was that while the data-driven approach can be good for predicting trends, it struggles to capture the deeper context associated with the stories and the topics,” Krishna said. The team emphasized that refining the model required ongoing monitoring, additional data inputs, and manual adjustments to ensure relevance.
Editorial skepticism initially hindered adoption, but confidence
