Show notes
E1127: Google's biggest SEO leak gave marketers more information about how search works than almost any event in the industry's history. But according to Mike King, most SEO strategies, tools, and reporting systems barely changed after the leak. Mike explains what the leaked Google API documentation confirmed, what it revealed about links and ranking systems, and why the SEO industry has failed to act on some of its most important findings. We break down Google's different index tiers, why a backlink's value may depend on where the linking page sits in the index, and why common third-party metrics such as Domain Authority and Domain Rating may not tell you what Google actually values. Mike also explains how Google uses vector embeddings to understand pages, passages, websites, authors, brands, and entities. This changes how marketers should think about relevance, authority, mentions, links, and content creation. We also go deep into AI search and how platforms such as ChatGPT retrieve and evaluate information. Topics covered include: - What Google's API leak confirmed about ranking systems - Why Mike believes the SEO industry failed to adapt - How Google separates content into different index tiers - Why links from pages with rankings and traffic may carry more value - Why Domain Authority and Domain Rating can be misleading - How vector embeddings help Google understand relevance - Why brand mentions can sometimes matter more than backlinks - The difference between optimizing for prompts and query fan-outs - How ChatGPT breaks a prompt into multiple search queries - Why comprehensive content can create more chances to appear in AI answers - How to inspect the searches ChatGPT performs behind the scenes - Why AI search should not be treated as another version of traditional SEO - How brands should measure AI visibility, citations, accuracy, and performance - Why Reddit, YouTube, LinkedIn, earned media, and your website all matter - How content format affects whether AI systems select your page or video - Why starting your own subreddit may help protect your brand's position - The biggest mistake companies make with AI-generated content - Why publishing an article from a single prompt often produces weak results - How human review, original sources, structured prompts, and subject experts improve AI content - Why Google may rely more on user behavior than unreliable AI-content detection - Which older SEO tactics are returning because of AI search - What "relevance engineering" means - Why future marketing teams may need SEO, PR, UX, content, data, and engineering skills - Why SEOs should learn to build tools and test AI platforms directly Mike also shares a specific AI search case study involving 499 response codes. His team found that ChatGPT was abandoning page requests because the client's website took too long to respond. After identifying the problem in the site's log files and improving response speed, the client's AI search visibility increased by roughly 300% over three months. Mike's practical recommendation is simple: Check your server logs for 499 errors. Identify the affected URLs. Improve page speed, especially time to first byte. According to Mike, fixing this problem can improve AI search performance in 30 days or less. This conversation is for SEOs, founders, content teams, PR professionals, and marketing leaders trying to understand how discovery is changing across Google, ChatGPT, Claude, Bing, YouTube, Reddit, and other search surfaces. ⭐️ Mike King's agency, iPullRank - https://ipullrank.com ⭐️ Mike King's music - https://verywellversed.com ⭐️ Mike King on LinkedIn - https://www.linkedin.com/in/michaelkingphilly/ ⭐️ Mike King on 𝕏 - https://x.com/iPullRankAgency 💎 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 Welcome and Setup 00:17 Google Leak Takeaways 05:44 Index Buckets Explained 07:50 Embeddings and Entities 09:19 Mentions vs Links 11:51 Earning Mentions Tactics 14:27 Black Mirror Launch 16:05 Seeding Clips and Ads 19:15 Leak Frenzy and Twiddlers 22:14 Optimizing for AI Search 23:50 Inspecting Query Fanout 26:18 Fanouts vs Keywords Strategy 30:01 C-Suite Measurement Framework 31:05 Input Metrics That Matter 31:43 Why It Is Not Just SEO 33:01 Tooling For Measurement 33:38 Query Fan Out Research 34:54 RAG Pipeline Visibility Gaps 37:06 Who Should Lead GEO 38:46 Pilot Win Fixing 499s 42:09 Getting Into Training Data 44:22 Relevance Engineering Explained 48:04 Content Formats For AI Search 49:29 Using Reddit Without Astroturfing 51:58 AI Content Mistakes To Avoid 54:10 How Google Judges AI Content 57:09 Future Of Niche Publishing 59:01 Unpopular SEO Take And Skills 01:02:44 One Action For 90 Days 01:03:04 Wrap Up And Where To Follow The Edward Show. Your daily search engine optimization podcast: https://edwardsturm.com/the-edward-show/ #seo #searchengineoptimization #generativeengineoptimization #answerengineoptimization