A big example is Facebook looking at how long people spent on their feed, how much stuff they liked, how much they responded, if those responses were negative o

A big example is Facebook looking at how long people spent on their feed, how much stuff they liked, how much they responded, if those responses were negative or positive, if they spent more time on negative content, etc; Once they had a decent understanding of that they literally A/B tested if people stayed on Facebook longer if they made them sad or if them made them happy. So yeah, if 'data' allows you to do clandestine psychology experiments on a massive scale, that is a big deal.