Experiment 4 min read
Impressions Up, Clicks Down? Three SEO Reports Worth Reading Carefully
Three practitioners describe a widening impression-to-click gap. Their reports also show why falling CTR is not proof of AI-caused traffic loss.
SEOVision library
SEO experiments
Transparent hypotheses, methods, evidence, and limitations.
Explore the library
10 resources Latest first
Experiment 4 min read
One AI Answer Is a Snapshot: Three Experiences with Citation Variability
People tracking AI answers report changing sources and unstable opportunity labels. Repeated, consistent sampling makes visibility claims more useful.
Experiment 4 min read
AI Discovery Beyond Your Blog: Three Experiences with Off-Site Content
Reports involving Reddit answers, LinkedIn posts and local business mentions show why a site-only visibility check can miss part of the picture.
Experiment 4 min read
AI Referrals Can Convert Well, but the Denominator Changes the Story
Three reports show stronger AI-referral conversion, while a fourth finds the reverse. Intent, sample size and conversion definitions explain why comparisons need care.
Experiment 4 min read
Does llms.txt Help AI Visibility? Four Reports Separate the Questions
Four practitioner reports distinguish file access, page discovery and AI citations. A positive example also explains why causation remains uncertain.
Experiment 4 min read
AI Internal Linking: Three Workflows That Keep Suggestions Reviewable
A prompt, a custom URL registry and an editor plugin show how people use AI for internal links. Their evidence is about workflow usefulness, not ranking lift.
Experiment 4 min read
AI Content Refreshes Work Better as a Data-Grounded Editorial Process
Three teams describe refresh workflows built around existing URLs, search data and editorial decisions. None supplies a comparable ranking-lift benchmark.
Experiment 4 min read
AI SEO Audits: Three Practitioners Found Reasons to Verify the Findings
Practitioners report both useful audit assistance and confident mistakes. An evidence requirement makes those outputs easier to evaluate.
Experiment 4 min read
AI Content Gets Published Faster. Does the SEO Value Keep Up?
Three public workflows show faster production or early visibility. Their evidence suggests a better metric: cost per useful, durable page.
Experiment 4 min read
Human Editing and AI Content: Three Reports on Ranking Durability
Three teams reported more durable performance after human editorial input. Their comparisons also reveal why authorship alone cannot explain rankings.