I spent time looking at actual studies on content performance rather than just repeating what everyone says works. The gap between common advice and research findings is pretty wide.
Reading level doesn't correlate with rankings like people claim
You'll hear "write at an 8th-grade level" constantly. But analysis of top-ranking content shows reading levels all over the map. Medical content ranks fine at college level. Simple topics sometimes rank better with more sophisticated writing. Context matters more than formulas.
Bounce rate is a terrible optimization metric
High bounce rates on informational content often mean you answered the question quickly. That's good, not bad. People obsess over keeping visitors on page longer when sometimes the best user experience is getting them their answer fast.
Image optimization has measurable impact, but not how you think
Alt text and file names have minimal ranking influence. What actually matters is load time and whether images help users understand your point. Pretty basic stuff that gets buried under SEO mythology.
Internal linking shows inconsistent results
Adding more internal links sometimes helps, sometimes does nothing. The pattern seems to be whether the links actually help users find related information they'd want, not just link volume.
The frustrating reality is that content optimization is more contextual and less formulaic than most advice suggests. What works depends heavily on your specific situation, topic, and audience expectations.