Rescue time, tag at the application level (obviously not fantastically accurate), and sub tag in chrome based on website.
It’s directionally correct at best, but still helpful.
Would also do a secondary form of manual evaluation by paying someone to text or call with a daily list of questions, but only have about 1.5 of years of data there. I ran the regressions and they are highly correlated in a time lagged way. In particular, sleep, for me, drove a lot of everything else. But from there it’s hard to say what the causality is (were you tired because of work to begin with? Or were you unproductive because of sleep?)
