We took a two-phased approach, starting with an examination of how factual content is talked about implicitly on social media, using Natural Language Processing to quantitatively measure the extent different perceptions people have of factual media content from text data.

Following this, we moved onto a comprehensive quantitative phase using MaxDiff to uncover the relative importance of different content features when it comes to factual and exploring occasions, drivers, topics, brand associations, preferences and audiences in detail.