It’s not uncommon for food manufacturers to use social listening platforms to react to a PR-crises, such as an ingredient thought to cause health or environmental concerns: think of instances such as talc in baby powder, 'yoga mat' materials in bread, or the more recent titanium dioxide concerns in candy. These crises are often a surprise and can lead to unexpected product reformulations, sourcing new ingredients, and modifying supply chains. In the worst case, these events can lead to billions of dollars in losses due to lawsuits and declines in shareholder value. There is a better way. Social listening to monitor news headlines alone is often a lagging indicator of a concern – not proactive but reactive. A system that synthesizes not just social data streams but also leading indicators such as academic research, peer-reviewed journals, research studies, and regulatory sources allows manufacturers to respond proactively. Simply pulling billions of documents isn’t enough either. Machine learning that is proven to work in other fields, provides an efficient solution to analyze massive datasets and titrate down to the most crucial documents that signal a potential concern or opportunity. This analysis uncovers patterns in the data that allow for holistic understanding of what the next potential villain (or hero!) ingredient might be. Speaker: Richard Hughes, PhD