Episode Details

E1191: One SEO spent the summer going through more than 30,000 documents and narrowing them down to 9,249 SEO case studies, tests, and papers with measurable results.

The final dataset covered 552 cited sources and 182 SEO tactics.

I go through some of the weirdest findings, the tests I agree with, the ones I question, and the tactics worth experimenting with on your own sites.

Some of the findings: - Writing title tags entirely in capital letters increased organic traffic 17.5% across two independent tests - Removing synonym keywords from title tags reduced organic traffic 27% - Removing the word "compare" from title tags increased organic traffic 24% - Putting product prices in title tags reduced organic traffic 15% - Number emojis in meta descriptions reduced organic traffic 5% - Using a 410 instead of a 404 got removed pages dropped from Google's index much faster - Editorial backlinks earned through digital PR increased organic sessions 71% - Disavowing links without a manual penalty showed no measurable direction - Adding internal links to orphaned blog posts produced major ranking improvements - Homepage footer links to category and subcategory pages increased organic sessions 5-10% in one controlled test - Adding prominent brand CTA blocks to product pages increased conversions 80% across two tests

I also dig into the AEO and GEO findings.

The evidence there is much thinner, but some of the reported results are worth testing: - Adding quantitative statistics increased AI citations 37% - Adding quotes from credible named sources increased citations 22-41% - Keyword stuffing reduced AI citations 10% - Putting a direct answer in the opening paragraph increased citations 280% - Writing content as discrete factual statements produced a 5X increase - Moving source references directly next to claims increased citations 40% - Adding author bylines increased AI citations 113% - Human editing of AI-written content increased citations 67% - Targeted AI-written additions to specific pages increased citations 73% - Publishing AI-generated text across an entire site reduced organic sessions 35%

I also cover one of the most interesting findings around LLMs.txt.

Across the underlying measurements, results ranged from -19.7% to +81%. More importantly, there wasn't a single controlled LLMs.txt test in all 9,249 documents.

Anyone claiming LLMs.txt definitively improves SEO or AI visibility is working with extremely weak evidence.

We also get into Google Business Profile categories, keywords in URLs, emojis in title tags, internal linking, disavow files, title tag testing, mass AI content, and why so many GEO studies are difficult to trust.

The biggest takeaway: traditional SEO testing still produces some of the strongest measurable results, especially around title tags, internal links, backlinks, and page structure.

⭐️ Reddit: I Read 9,249 SEO Case Studies to Build the Biggest SEO Meta-Study - https://www.reddit.com/r/TechSEO/comments/1wmb181/i_read_9249_seo_case_studies_to_build_the_biggest/ ⭐️ Ovisy: The Biggest SEO Meta-Study (Full Study) - https://ovisy.hr/blog/seo-meta-study ⭐️ Ep 763 - The SEO Trick That Breaks All the Rules: How Ultra-Long Page Titles Boost Rankings - https://youtu.be/ufjwb51hJuc

💎 Compact Keywords - My SEO Course - Get paying customers through SEO - Clear step-by-step video breakdowns - SEO templates to be copied and adapted for your products and services: https://compactkeywords.com/

00:00 Meta Study Setup 00:52 How The Data Was Built 01:38 Title Tag Oddities 03:15 Breadcrumbs And Status Codes 04:01 Backlinks And Disavow 05:27 Internal Linking Wins 07:31 AEO And GE