It's just bologna

May start using the more common paid search phrase: "qualified clicks"

In that case, maybe topical groups might be interesting to see if there's still a bias within?

i.e.: definitional, how to, vs, etc.

add auto-play video and disable the back button /snark

Split by page type - may have an intrinsic bias where more volume is more likely to correlate with informational pages which is more likely to have users reading longer information than a product page.

Sadly that's not what the data is saying. It's saying that links barely matter (<~30% correlation) and now matter less.

Looks like it has been suspended now...

Now what do I try to understand with this data...

Hrm...maybe I should do more prominent ones...guru99.com?

comparitech.com