SEO CONTRADICTIONS: Volume II — Experiments
About this ebook
Observations tell us what is happening. Experiments reveal why.
If Volume I was about noticing the anomalies hiding in plain sight, Volume II is about pushing the system until it reveals its underlying rules.
Marketing advice tells you to follow guidelines. Engineering demands that you test boundaries. In this volume, I move past passive observation and into active hypothesis testing. I apply a software engineering mindset to the most uncomfortable questions in modern search.
I don’t run tests to produce polished marketing collateral or validate official documentation. I run them to see where the machine breaks, where the incentives misalign, and where automated infrastructure creates asymmetric advantages. Volume II is a forensic log of what happens when you stop believing "best practices" and start believing raw data, server logs, and retrieval outcomes.
Inside Volume II:
The Geo-Restriction Myth — A forgotten for years experiment testing whether blocking international traffic actually impacts search visibility in a target market.
Precision vs. Breadth — Documented proof of why "Direct-Match" content can beat LLM query fan-out in Google AI Overviews.
The Multimodal Vision Audit — A forensic look at how LLMs "see" images. I expose the fundamental tokenization difference between SVG and JPG, and show how HTML semantics can completely override visual reality.
The Identity Gap — A technical investigation into why llms.txt adoption is stalling and a proposal for the "Semantic Anchor"—the missing identity layer for AI trust. The first-ever proposition to relate semantically the entities behind a website using JSON-LD.
The Triangular Authority Chain — Defining the machine-verifiable relationship between the Person, the Service, and the Website.
Industrialized Tactical Outreach — How I forced commercial crawlers to act as a zero-cost PR department for 26,000+ businesses simultaneously.
The Provenance Paradox — Can AI transform expert YouTube knowledge into ranking articles without the creator ever consuming the source material?
The Framework:
Every chapter follows a strict, unvarnished methodology: The Question, The Hypothesis, The Experiment, and The Interpretation.
This is not a book of static truths. It is proof of work from the front lines of semantic engineering.
Don’t accept official documentation at face value.
Don’t optimize for a version of the web that no longer exists.
Run the setups. Break the hypotheses. Build your own infrastructure.
Discover more
About the author
Marin Popov is a Technical SEO consultant and software engineer, known for his forensic approach to search systems and his AI‑driven semantic infrastructure projects, including 1 Euro SEO.
He holds a Master of Science in Telecommunications from the Technical University of Sofia, a background that shaped his focus on machine readability, structured data, and the intersection between search engines and AI.
Before transitioning fully into software engineering and technical SEO, Marin taught secondary school computer science in Bulgaria. That combination—engineering, teaching, and software development—formed a defining principle in his work: listen, but never blindly trust.
Marin treats the web as a live laboratory. He tests assumptions, reads raw server logs, audits source code, and documents the anomalies that contradict industry narratives. For him, a failed experiment is not a failure; it is diagnostic evidence.
This book is a collection of those experiments.
