Only problem is that averages are borked. For example-- 20 people in a room. 1 is 10' tall. 2 are 9' tall. 2 are 8' tall. 5 are 3' tall and 10 are only 2' tall.
The average height is 4' tall. Except, no one is 4' tall. Averages are broken data.
Incidentally, I used those weird heights because they loosely follow long tail theory and web visits and buying behavior tends to fall under long tail theory.
On Medium, there's tons of stories that get no claps while a select few get 5K and 10K. Which, incidentally, doesn't translate to dollars. I had one go viral with tends of thousands of views and made me pennies. I wrote about that. :)
The info in here is amazing. Might write about it, and if I do, I'll tag you for credit. Bookmarked.
Awesome piece -- I sure do love the way you sling data. My bookmarks are full of your posts. :)