Metropolis to Metadata

Superman’s Guide to Internal Links and LLMs for Gaining Authority

Transcript

1. Metropolis to Metadata Superman’s Guide to Internal Links and LLMs for Gaining Authority Grant Simmons LinkedIn.com/in/simmonet
2. Grant Simmons • SEO for a long time • Work with InLinks & Waikay • Run my own agency • Advisor to startups • Live in Virginia • Not that “Super” J LinkedIn/in/simmonet x.com/simmonet
3. Grant Simmons Metropolis to Meta Data • Let’s get ready to learn • SEO ≠ GEO ≠ AEO • SEO is evolving (still SEO) • We just need to do it “better” • What we should be doing J • Superman Movie anyone? LinkedIn/in/simmonet x.com/simmonet
5. We’re going to use Superman’s universe as a metaphor for building or improving your website’s authority & visibility in the age of AI
6. It’s All About Connections
7. Authority is from Trusted Connections
8. Bad Connections Weaken Authority
9. Today We’re Talking About Internal Links
10. And Entities
11. And the Power of Knowledge
12. “Fortress of Knowledge” = Entities + Links
13. What We’ll Cover • Entities • Internal Links • Knowledge Graphs • LLMs Love Connected Entities • Spotting Gaps & “Kryptonite” • Building Better Graphs
15. Internal Links Build Knowledge Graphs (from Entities)
16. Who is Clark Kent?
18. Wikipedia Knows
19. How Do You Know?
20. You Have Context & Knowledge
21. Context Helps Us Understand Meaning We know Superman is an alien from Krypton, found and named by the Kent family from Kansas.
22. Context Helps Us Understand Meaning We know Superman is an alien from Krypton, found and named by the Kent family from Kansas.
23. Superman is an Entity
24. Clark Kent is Connected to Superman Entity
25. So Let’s Define an (Named) Entity
26. Retrieval is a Key Part of Extraction Retrieval > Extraction > SERP/AI Results
27. Named Entities Used in NLP
28. Retrieval -> Extraction Search Engines & LLMs Do This!
29. Quick break. We’re going to talk a lot about search engines and LLMs BUT ultimately we’re creating content for human satisfaction. What I’m showing today is how to get users to your site. When they get there the content must satisfy their query. No shortcuts.
30. Better Graph. Better Understood. Better Retrieval.
31. Super Knowledge Graph Superman Super Dog Krypton Fortress of Solitude Robot 4 Lex Luthor Eve Teschmacher Jimmy Olsen Lois Lane Steve Lombard Metamorpho Mr. Terrific Green Lantern Hawk Girl
32. Graph Relationships are Defined Too Superman Super Dog Krypton Fortress of Solitude Robot 4 Lex Luthor Eve Teschmacher Jimmy Olsen Lois Lane Steve Lombard Metamorpho Mr. Terrific Green Lantern Hawk Girl Justice Gang Man<>Dog Boy/Girl Friend Co Workers Boy/Girl Friend Man<>Robot Nemesis Boy/Girl Friend
33. Internal Links + Relationship Context = Graph Superman Super Dog Krypton Fortress of Solitude Robot 4 Lex Luthor Eve Teschmacher Jimmy Olsen Lois Lane Steve Lombard Metamorpho Mr. Terrific Green Lantern Hawk Girl Justice Gang Man<>Dog Boy/Girl Friend Co Workers Boy/Girl Friend Man<>Robot Nemesis
34. Building Topical Authority • Entities connected with clear contextual links • Relevant page to relevant page • Relationships defined (girl friend <> boyfriend) • Clear anchor text for context • In copy blocks for context (surrounding copy) • HELP search engines and LLMs
35. “Bad” Connections + “Bad” Context TRUST/ACCURACY CONSENSUS FACTS/TRUTH
36. Who Does It Best?
37. Wikipedia is a “Super” Knowledge Graph
38. Contextual Internal Links to Related Pages
39. Contextual Internal Links to Related Pages context of “is a” and “published by” and “by name in” and “close proximity to”
40. That Then Link to Related Pages
41. Even Krypto Gets His Own Page!
42. With a Link Back to Superman J
43. Let’s Get Back to “Facts” TRUTH/FACTS TRUST/ACCURACY CONSENSUS
44. Page (Entity) Focus • Consensus isn’t just an LLM thing • (It’s important and we’ll cover later) • Internal links provide consensus • Specific links point to specific pages • We can define the entity canonical • Supported by sub-topics entity canonical entity supporting topics
45. How Do We Tell Search Engines?
46. sameAs = Entity Canonical Signal SameAs Schema
47. sameAs = Entity Canonical Signal SameAs Schema
50. About
51. Mentions
52. Reinforce With Relevant Internal Links Takeaway: Map Entities to Pages
53. Quick Question When was the first Superman comic published?
54. Perplexity Claude Gemini Google AIO ChatGPT
55. Perplexity Claude Gemini Google AIO ChatGPT
56. Perplexity Claude Gemini Google AIO ChatGPT
57. Interesting. But Can That Help Us?
58. LLMs and Sources
59. Sources • Wikipedia • Others Perplexity
60. Knowledge graph to LLM pipeline (simplified) Search Query > NLP > “What are they asking for?”
61. Knowledge graph to LLM pipeline (simplified) Search → Knowledge Graph What Sources Are Relevant (From What I Know)
62. Knowledge graph to LLM pipeline (simplified) Search → Knowledge Graph → AI Answer What’s the Best Way of Presenting the Answer?
63. This is the Key Bit - Knowledge Graph → AI Answer What bits of what sources should I show? What’s the Best Way of Presenting the Answer?
64. Tough Question!
66. LLMs Look for Truth For an LLM, truth is whatever it believes from the data it has. It doesn’t know objective truth - it only knows what’s represented in its training data and (potentially) retrieval sources. So if Krypton is well-documented as Superman’s home planet, that becomes its ‘truth.’
67. LLMs Look for Accuracy Accuracy depends on whether the model connects that truth correctly in the context of the prompt. If there are missing links or gaps, the model might hallucinate. If Superman’s connection to Krypton isn’t graphed, the AI might fill the gap incorrectly and say ‘Superman is from Mars.’
68. LLMs Look for Consensus When multiple possible answers exist, the model looks for the strongest statistical signal. It’s essentially asking: ‘Across all the data I’ve seen, what do most of the sources agree on?’ If 80% of the sources say ‘Superman is from Gotham City’ - even though it’s wrong - consensus will often win.
69. Great Content & Connections Win
70. Great Content = Sources Covering Topics
71. And... Great Connections J
72. Quick Question Who Played Lois Lane in the Superman 2025 Movie?
73. Quick Question Who Played Lois Lane in the Superman 2025 Movie? Rachel Brosnahan
75. Spotting Gaps & “Kryptonite” to Build Better Knowledge Graphs
76. Finding Weakness in Your Knowledge Graph
77. You’ve Built Your Knowledge Graph • Identified the entities in your pages • Identified relevant sub- pages • Built topic hubs / pillars / core • Added internal links with relevant anchors
78. Something’s Missing?
79. Wikipedia is Your Friend
80. InLinks Also J - Entity Relevant
81. Look at your Competitors! Source: Waikay.io Knowledge Gaps for “Lois Lane”
82. Look at your Competitors! Find Gaps! Source: Waikay.io Knowledge Gaps for “Lois Lane”
83. Keywords? Schmewords!
84. Keywords? Schmewords!
85. Keywords are Not How People Search Not in the AI world – there’s a lot of “long tail” J Correlation ≠ Causation
86. Keywords Are Dead
87. Since 2013 J
88. So We’re Not Optimizing for Keywords We’re Optimizing for Topical Coverage Source: InLinks.com Clusters for “Superman”
89. Not Keywords. Queries. Topics. We’re Optimizing for Topical Coverage Source: InLinks.com Clusters for “Superman”
90. Yes. There is a Superman Burrito at Taco Bell.
91. Covering the Topic *Completely*
92. And Expertly
93. Topical Coverage -> Topical Authority • Entities connected with clear contextual links • Relevant page to relevant page • Relationships defined (girl friend <> boyfriend) • Clear anchor text for context • In copy blocks for context (surrounding copy) • HELP search engines and LLMs • By covering the topic completely & expertly
94. Topical Coverage • Topical authority is the outcome • Topical coverage is the process • Smarter research • Looking at every angle • Finding gaps in coverage • Understand audience problems • Satisfy those queries • Answer all questions
96. Entities are Building Blocks
97. Internal Links Connect Entities
98. Entities + Links = “The Graph”
99. Find Gaps. Fill Gaps.
100. A Better Graph Covers a Topic Completely
101. Topical Coverage = Topical Authority
102. Topical Authority Inspires Discovery
103. Discovery Leads to Visibility
104. And That’s “Super”
105. Thank You
106. Questions?
107. InLinks.com | waikay.io Grant Simmons
108. LinkedIn/in/simmonet x.com/simmonet