Last week SEMRush put out a very interesting study on ranking factors they were able to observe from their data set.

The results highlighted a few metrics that aren't traditionally considered primary ranking factors - a huge correlation between direct traffic visits and high ranking (favoring big brands), and user behavior metrics like bounce rate, time on site, and pages per session were more highly correlated than even linking domains.

Study is here (warning, email gated content): https://www.semrush.com/ranking-factors/

SEMRush claims that their algorithm looked past just correlation, but do you think that these factors actually have this much influence? At the very least, this should make it easier to sell A/B testing and other UX services in to clients. Thoughts?