They address the engagement issue in the article, but not in a particularly convincing way : "Internal studies conducted by the agency showed that KPIs like bra

They address the engagement issue in the article, but not in a particularly convincing way : "Internal studies conducted by the agency showed that KPIs like brand lift and intent to purchase were not affected by whether the viewer watched the video with the sound on or off" I wonder if they are actually measuring engagement over a complete feed stream, or just some subset of videos that were allowed to play longer than 10s (guaranteeing a baseline level of engagement).