Public Discourse & Trust Modeling
MerckModeled how themes, events, reach, and timing shaped public sentiment and Trust in a complex media environment.
Challenge
A global healthcare company needed a data-driven view of how large-scale media-scraper, news, social, event, and public-discourse data influenced vaccine sentiment. The goal was to identify communication watchouts, Trust-building opportunities, and moments where proactive storytelling could shape the narrative.
Approach
Used predictive AI and statistical modeling on high-volume news and social discourse to evaluate how selected topics and events influenced sentiment over time. The model separated baseline sentiment from topic-driven swings and measured Trust-driving and distrust-driving mentions separately before calculating net discourse impact.
Impact
The work showed that media volume alone did not explain impact. Reach, timing, topic alignment, and sentiment dynamics mattered. Confidence-oriented conversations reached four times as many people, while negative sentiment tended to peak up to five days after positive sentiment. That timing created an opportunity for more proactive narrative planning.
Selected slides
A closer look at the work
Want the deeper story?