Everything Google Lied to Us About

Slide 1

Slide 1 text

1 1

Slide 2

Slide 2 text

2 2 Download this deck: https://speakerdeck.com/ipullrank

Slide 3

Slide 3 text

3 Salutations! I’m Mike King (@iPullRank)

Slide 4

Slide 4 text

Slide 5

Slide 5 text

5 Things Google Has Lied to Us About

Slide 6

Slide 6 text

6 6

Slide 7

Slide 7 text

7 7 Google Has Owned Everything It Would Need to Become SkyNet

Slide 8

Slide 8 text

8 Just Create Great Content, We’ll Figure it Out He said it so many times, I’m not going to cite a source.

Slide 9

Slide 9 text

9 9 We Don’t Understand Content, We Fake it

Slide 10

Slide 10 text

10 10 We Don’t Use Clicks in Rankings

Slide 11

Slide 11 text

11 11

Slide 12

Slide 12 text

12 12 We Use Clicks in Navboost aka GLUE -Pandu Nayak

Slide 13

Slide 13 text

13 13 Here’s Google Telling On Itself

Slide 14

Slide 14 text

14 14

Slide 15

Slide 15 text

15 15 Y’all owe @randﬁsh an apology. Go ahead and send it right now.

Slide 16

Slide 16 text

16 16 Don’t Make Machine-Generated Content For Search

Slide 17

Slide 17 text

17 17 Hey News People, Use this Generative AI Tool For Your Articles

Slide 18

Slide 18 text

18 18 Ads and Organic Search are Separated like Church and State

Slide 19

Slide 19 text

19 19 Oh, Here’s a Thread Where the Ads Team is Asking the Search Team to Juice Ads

Slide 20

Slide 20 text

20 20 Using Expired Domains Don’t Give You an Advantage …you can get that domain into Google; you just won’t get credit for any pre-existing links. -Matt Cutts So if the content was gone for a couple of years, probably we need to ﬁgure out what this site is, kind of essentially starting over fresh. So from that point of view I wouldn’t expect much in terms of kind of bonus because you had content there in the past. I would really assume you’re going to have to build that up again like any other site. -John Mueller

Slide 21

Slide 21 text

21 21 Then Why are Just NOW Making Expired Domains Spam?

Slide 22

Slide 22 text

22 22 Google is not here for us; we are here for Google.

Slide 23

Slide 23 text

23 23 Google’s Search guidelines are aspirational and we help them the enforce them.

Slide 24

Slide 24 text

24 24

Slide 25

Slide 25 text

25 25 SEOs are Google’s unpaid workforce.

Slide 26

Slide 26 text

26 Things SEO Lies to Itself About

Slide 27

Slide 27 text

27 27 There’s A Bigger Generational Shift Happening Too..

Slide 28

Slide 28 text

28 28

Slide 29

Slide 29 text

29 29

Slide 30

Slide 30 text

30 30 40% of People Leaving ChatGPT Go to Google

Slide 31

Slide 31 text

31 31 Google Still Dwarfs Everything, but More People are Using More Channels

Slide 32

Slide 32 text

32 32 It means there is fragmentation in how information needs are being met.

Slide 33

Slide 33 text

33 33 Our understanding of how Google works is out of date…

Slide 34

Slide 34 text

34 34 In SEO we still think Google is Here

Slide 35

Slide 35 text

35 35

Slide 36

Slide 36 text

36 36 Search Engines Work based on the Vector Space Model Documents and queries are plotted in multidimensional vector space. The closer a document vector is to a query vector, the more relevant it is.

Slide 37

Slide 37 text

37 37 TF-IDF Vectors The vectors in the vector space model were built from TF-IDF. These were simplistic based on the Bag-of-Words model and they did not do much to encapsulate meaning.

Slide 38

Slide 38 text

38 38 Relevance is a Function of Cosine Similarity When we talk about relevance, it’s the question of similar is determined by how similar the vectors are between documents and queries. This is a quantitative measure, not the qualitative idea of how we typically think of relevance.

Slide 39

Slide 39 text

39 39 Google Shifted from Lexical to Semantic a Decade Ago

Slide 40

Slide 40 text

40 40 Very very few SEO tools are offering analysis that aligns with how Google works today.

Slide 41

Slide 41 text

41 41 Google Has Been More Like This Since Hummingbird

Slide 42

Slide 42 text

42 Semantic Search is Fueled by High Density Embeddings …just like large language models. A lot of what Google has always been trying to do is more real.

Slide 43

Slide 43 text

43 43 This Allows for Mathematical Operations Comparisons of content and keywords become linear algebraic operations.

Slide 44

Slide 44 text

44 44 Words are Converted to Multi-dimensional Coordinates in Vector Space

Slide 45

Slide 45 text

45 45 We Went from Sparse Embeddings to Dense Embeddings

Slide 46

Slide 46 text

46 46 Word2Vec Gave Us Hummingbird

Slide 47

Slide 47 text

47 47 I Wrote About How You Can Get These With Screaming Frog and OpenAI Vectorize your content as you crawl. https://ipullrank.com/vector-embeddings-is-all-you-need

Slide 48

Slide 48 text

48 48 I Talked About How You Use Google Sheets to Do the Analysis Cosine Similarity is the measure of relevance: https://ipullrank.com/cosine-similarity-knn-in-google-sheets

Slide 49

Slide 49 text

49 49 Since the Introduction of BERT, Google Has Looked More Like This

Slide 50

Slide 50 text

50 50 Google Has Been Using Public About Models Since 2020 This is why some of the search results feel so weird. A re-ranking of documents with a mix of lexical and semantic. https://arxiv.org/pdf/2010.01195.pdf

Slide 51

Slide 51 text

51 51 The way SEO build links doesn’t work as well we think it does anymore.

Slide 52

Slide 52 text

52 52 Google Has Always Aspired to Leveraging Link Relevance in Meaningful Ways

Slide 53

Slide 53 text

53 Dense Retrieval You remember “passage ranking?” This is built on the concept of dense retrieval wherein there are more embeddings representing more of the query and the document to uncover deeper meaning.

Slide 54

Slide 54 text

54 54 Dense Retrieval is Scoring down to the Sentence Level

Slide 55

Slide 55 text

55 55 You need to focus on building more relevant links than higher volumes of links.

Slide 56

Slide 56 text

56 56 Under the SGE Model, Google is Structured Liked This

Slide 57

Slide 57 text

57 57 All of this is a huge problem because SEO software still operates on the lexical model.

Slide 58

Slide 58 text

58 58 No one is actually improving content when they do the skyscraper technique

Slide 59

Slide 59 text

59 Indexing is Also Harder It’s not being talked about as much, but indexing has gotten a lot harder since the Helpful Content update. You’ll see a lot more pages in the “Discovered - currently not indexed” and “Crawled - currently not indexed” than you did previously because the bar is higher for what Google deems worth capturing from the web.

Slide 60

Slide 60 text

60 60 I Believe This is a Function of Information Gain Conceptually, as it relates to search engines, Information Gain is the measure of how much unique information a given document adds to the ranking set of documents. In other words, what are you talking about that your competitors are not?

Slide 61

Slide 61 text

61 61 In conclusion: “More content” is no longer inherently the most effective approach because there’s no guarantee of traﬃc from Google.

Slide 62

Slide 62 text

62 62 The only content you should be making

Slide 63

Slide 63 text

63 63 I’m Leaving Y’all with Four Actions Today 1. How to Prune Your Content 2. How to Use LLMs 3. How to Appear in LLM based Search Engines 4. How to Think About Relevance

Slide 64

Slide 64 text

64 The Content Pruning Process

Slide 65

Slide 65 text

65 65 Pruning and Optimization Work Quite Well Together

Slide 66

Slide 66 text

66 Aleyda Has a Process Aleyda’s workﬂow is a great place to work through whether your content should be pruned or not. https://www.aleydasolis.com/en/crawli ng-mondays/how-to-prune-your-website- content-in-an-seo-process-crawlingmon days-16th-episode/

Slide 67

Slide 67 text

67 67 We like automate to get to a Keep. Revise. Kill. (Review.)

Slide 68

Slide 68 text

68 68 Content Decay The web is a rapidly changing organism. Google always wants the most relevant content, with the best user experience, and most authority. Unless you stay on top of these measures, you will see traﬃc fall off over time. Measuring this content decay is as simple comparing page performance period over period in analytics or GSC. Just knowing content has decayed is not enough to be strategic.

Slide 69

Slide 69 text

69 69 It’s not enough to know that the page has lost traﬃc.

Slide 70

Slide 70 text

70 70

Slide 71

Slide 71 text

71 71 The Content Potential Rating (CPR).

Slide 72

Slide 72 text

72 72 Content Potential Score

Slide 73

Slide 73 text

73 73 Interpreting the Content Potential Rating 80 - 100: High Priority for Optimization 60 - 79: Moderate Priority for Optimization 40 - 59: Selective Optimization 20 - 39: Low Priority for Optimization 0 - 19: Minimal Benefit from Optimization If you want quick and dirty, you can prune everything below a 40 that is not driving significant traffic.

Slide 74

Slide 74 text

74 74 Combining CPR with pages that lost traﬃc helps you understand if it’s worth it to optimize.

Slide 75

Slide 75 text

75 75 Step 1. Pull the Rankings Data from Semrush Organic Research > Positions > Export

Slide 76

Slide 76 text

76 76 Step 2: Pull the Decaying Content from GSC Google Search Console is a great source to spot Content Decay by comparing the last three months year over year. Filter for those pages where the Click Difference is negative (smaller than 0) then export.

Slide 77

Slide 77 text

77 77 Step 3: Drop them in the Spreadsheet and Press the Magic Button

Slide 78

Slide 78 text

78 The Output is List of URLs Prioritized by Action Each URL is marked as Keep, Revise, Kill or Review based on the keyword opportunities available and the effort required to capitalize on them. Sorting the URLs marked as “Revise” by Aggregated SV and CPR will give you the best opportunities ﬁrst.

Slide 79

Slide 79 text

79 79 Get your copy of the Content Pruning Workbook : https://ipullrank.com/cpr-sheet

Slide 80

Slide 80 text

80 How to Kill Content Content may be valuable for channels outside of Organic Search. So, killing it is about changing Google’s experience of your website to improve its relevance and reinforce its topical clusters. The best approach is to noindex the pages themselves, nofollow the links pointing to them, and submit an XML sitemap of all the pages that have changed. This will yield the quickest recrawling and reconsideration of the content.

Slide 81

Slide 81 text

81 81 How to Revise Content Review content across the topic cluster Use co-occurring keywords and entities in your content Add unique perspectives that can’t be found on other ranking pages Answer common questions Answer the People Also Ask Questions Restructure your content using headings relevant to the above Add relevant Structured markup Expand on previous explanations Add authorship Update the dates Make sure the needs of your audiences are accounted for Add to an XML sitemap of only updated pages

Slide 82

Slide 82 text

82 How to Review Content The sheet marks content that has a low content potential rating and a minimum of 500 in monthly search volume as “Review” because they may be long tail opportunities that are valuable to the business. You should take a look at the content you have for that landing page and determine if you think the effort is worthwhile.

Slide 83

Slide 83 text

83 How to Use LLMs for Content

Slide 84

Slide 84 text

84 84 What is Retrieval Augmented Generation?

Slide 85

Slide 85 text

85 85 It’s Not Diﬃcult to Build with Llama Index sitemap_url = "[SITEMAP URL]" sitemap = adv.sitemap_to_df(sitemap_url) urls_to_crawl = sitemap['loc'].tolist() ... # Make an index from your documents index = VectorStoreIndex.from_documents(documents) # Setup your index for citations query_engine = CitationQueryEngine.from_args( index, # indicate how many document chunks it should return similarity_top_k=5, # here we can control how granular citation sources are, the default is 512 citation_chunk_size=155, ) response = query_engine.query("YOUR PROMPT HERE")

Slide 86

Slide 86 text

86 86

Slide 87

Slide 87 text

87 87

Slide 88

Slide 88 text

88 88

Slide 89

Slide 89 text

89 89

Slide 90

Slide 90 text

90 90

Slide 91

Slide 91 text

91 91

Slide 92

Slide 92 text

92 92

Slide 93

Slide 93 text

93 Integrate Promptitude with Zapier or Make

Slide 94

Slide 94 text

94 94 If you’re building your own custom stuff, use Gemma. Vertex is so expensive.

Slide 95

Slide 95 text

95 The threat of Google’s Search Generative Experience (SGE) aka AI Overviews

Slide 96

Slide 96 text

96 96 Queries are Longer and the Featured Snippet is Bigger 1. The query is more natural language and no longer Orwellian Newspeak. It can be much longer than the 32 words that is has been historically in order 2. The Featured Snippet has become the “AI snapshot” which takes 3 results and builds a summary. 3. Users can also ask follow up questions in conversational mode.

Slide 97

Slide 97 text

97 97 AI Overviews are a Function of Retrieval Augmented Generation (RAG)

Slide 98

Slide 98 text

98 98 The Search Demand Curve will Shift With the change in the level of natural language query that Google can support, we’re going to see a lot less head terms and a lot more long tail term.

Slide 99

Slide 99 text

99 99 The CTR Model Will Change With the search results being pushed down by the AI snapshot experience, what is considered #1 will change. We should also expect that any organic result will be clicked less and the standard organic will drop dramatically. However, this will likely yield query displacement.

Slide 100

Slide 100 text

10 0 Rank Tracking Will Be More Complex As an industry, we’ll need to decide what is considered the #1 result. Based on this screenshot positions 1-3 are now the citations for the AI snapshot and #4 is below it. However, the AI snapshot loads on the client side, so rank tracking tools will need to change their approach.

Slide 101

Slide 101 text

10 1 10 1 Context Windows Will Yield More Personalized Results SGE maintains the context window of the previous search in the journey as the user goes through predeﬁned follow questions. This will need to drive the composition of pages to ensure they remain in the consideration set for subsequent results.

Slide 102

Slide 102 text

10 2 SGE is Susceptible to Spam and Lower Quality Sites

Slide 103

Slide 103 text

10 3 10 3 Luckily Users Love it So Much

Slide 104

Slide 104 text

10 4 10 4 Ranking in Search Generative Experience is more about relevance than the other signals.

Slide 105

Slide 105 text

10 5 How to Appear in LLMs

Slide 106

Slide 106 text

10 6 10 6 Blocking LLMs is a Mistake. Appearing in these places will be recognized as brand awareness opportunities very soon.

Slide 107

Slide 107 text

10 7 10 7 What is Mitigation for AI Overviews? 1. Manage expectations on the impact 2. Understand the keywords under threat 3. Re-prioritize your focus to keywords that are not under threat 4. Optimize the passages for the keywords you want to save

Slide 108

Slide 108 text

10 8 10 8

Slide 109

Slide 109 text

10 9 10 9 We Can Also Show You Per Keyword How You Show Up

Slide 110

Slide 110 text

11 0 11 0 It’s all about the Fraggles.

Slide 111

Slide 111 text

11 1 11 1

Slide 112

Slide 112 text

11 2 11 2 The Fraggles Show What SGE Used for the AI Snapshot

Slide 113

Slide 113 text

11 3 11 3 Scroll to Text You can capture the copy used to inform the AI snapshots by scraping the Scroll to Text copy from the page.

Slide 114

Slide 114 text

11 4 11 4 There’s a Nearly Linear Relationship Between Fraggle Relevance and AI Snapshot Appearance Relevance against the chunks to keyword: Relevance against AI Snapshot:

Slide 115

Slide 115 text

115 11 5 Check out MarketBrew’s Free Tool to Help

Slide 116

Slide 116 text

Embrace Structured Data There are three models gaining popularity: 1. KG-enhanced LLMs - Language Model uses KG during pre-training and inference 2. LLM-augmented KGs - LLMs do reasoning and completion on KG data 3. Synergized LLMs + KGs - Multilayer system using both at the same time https://arxiv.org/pdf/2306.08302.pdf Source: Unifying Large Language Models and Knowledge Graphs: A Roadmap

Slide 117

Slide 117 text

117 11 7

Slide 118

Slide 118 text

118 11 8

Slide 119

Slide 119 text

119 11 9

Slide 120

Slide 120 text

120 12 0 They Share Their Prompts in Their Code The GEO team also shared the ChatGPT prompts that help them improve their visibility. You can augment them and put them to work right away. https://github.com/GEO-optim/ GEO/blob/main/src/geo_functi ons.py

Slide 121

Slide 121 text

121 12 1

Slide 122

Slide 122 text

12 2 Roll the Credits

Slide 123

Slide 123 text

12 3 12 3 Get your threat report: https://ipullrank.com/sge-report

Slide 124

Slide 124 text

Thank You | Q&A [email protected] Award Winning, #GirlDad Featured by Get Your SGE Threat Report: https://ipullrank.com/sge-report Play with Raggle: https://www.raggle.net Download the Slides: https://speakerdeck.com/ipullrank Mike King Chief Executive Oﬃcer @iPullRank