{"id":602022,"date":"2023-12-21T16:10:50","date_gmt":"2023-12-21T13:10:50","guid":{"rendered":"https:\/\/en.buradabiliyorum.com\/googles-shifting-approach-to-ai-content-an-in-depth-look\/"},"modified":"2023-12-21T16:10:50","modified_gmt":"2023-12-21T13:10:50","slug":"googles-shifting-approach-to-ai-content-an-in-depth-look","status":"publish","type":"post","link":"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/","title":{"rendered":"#Google\u2019s shifting approach to AI content: An in-depth look"},"content":{"rendered":"<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_85 counter-hierarchy ez-toc-counter ez-toc-custom ez-toc-container-direction\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">Table of Contents<\/p>\n<label for=\"ez-toc-cssicon-toggle-item-6a3493d419aaa\" class=\"ez-toc-cssicon-toggle-label\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #dd3333;color:#dd3333\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #dd3333;color:#dd3333\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/label><input type=\"checkbox\"  id=\"ez-toc-cssicon-toggle-item-6a3493d419aaa\" checked aria-label=\"Toggle\" \/><nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#A_deep_dive_into_the_proliferation_of_AI-generated_content_its_impact_on_search_quality_and_the_future_of_combating_spam\" >A deep dive into the proliferation of AI-generated content, its impact on search quality, and the future of combating spam.<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Spammy_AI_content_all_over_the_web\" >Spammy AI content all over the web<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Invisible_junk\" >Invisible junk<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Content_is_king_%E2%80%93_and_the_algorithm_is_the_Emperors_new_clothes\" >Content is king \u2013 and the algorithm is the Emperor\u2019s new clothes<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Google_relies_on_user_interactions_on_SERPs_to_judge_content_quality\" >Google relies on user interactions on SERPs to judge content quality<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Brands_and_the_cesspool\" >Brands and the cesspool<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#What_does_Google_consider_AI_spam\" >What does Google consider AI spam?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-8\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Patterns_of_AI_content_spam\" >Patterns of AI content spam<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-9\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#If_its_spam_why_does_it_rank_at_all\" >If it\u2019s spam, why does it rank at all?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-10\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#Google_Mind_the_gap\" >Google: Mind the gap<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-11\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#An_HCU_hop_to_UGC_to_beat_the_GPT\" >An HCU hop to UGC to beat the GPT?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-12\" href=\"https:\/\/buradabiliyorum.com\/en\/googles-shifting-approach-to-ai-content-an-in-depth-look\/#What_does_Googles_long-term_plan_look_like_for_AI_spam\" >What does Google\u2019s long-term plan look like for AI spam?<\/a><\/li><\/ul><\/nav><\/div>\n<h2 class=\"subhead\" itemprop=\"alternativeHeadline\"><span class=\"ez-toc-section\" id=\"A_deep_dive_into_the_proliferation_of_AI-generated_content_its_impact_on_search_quality_and_the_future_of_combating_spam\"><\/span>A deep dive into the proliferation of AI-generated content, its impact on search quality, and the future of combating spam. <span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p><\/p>\n<div class=\"bialty-container\">\nThe prevalence of mass-produced, AI-generated content is making it harder for Google to detect spam.\u00a0<\/p>\n<p>AI-generated content has also made judging what is quality content difficult for Google. <\/p>\n<p><!-- \/1038259\/SEL_Post-text --><\/p>\n<div id=\"div-gpt-ad-1693000027709-0\"><\/div>\n<div id=\"post-break\"><\/div>\n<p>However, indications are that Google is improving its ability to identify low-quality AI content algorithmically.\u00a0<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-spammy-ai-content-all-over-the-web\"><span class=\"ez-toc-section\" id=\"Spammy_AI_content_all_over_the_web\"><\/span>Spammy AI content all over the web<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>You don\u2019t need to be in SEO to know generative AI content has been finding its way into Google search results over the last 12 months.<\/p>\n<p>During that time, Google\u2019s attitude toward AI-created content evolved. The official position moved from \u201cit\u2019s spam and breaks our guidelines\u201d to \u201cour focus is on the quality of content, rather than how content is produced.\u201d<\/p>\n<p>I\u2019m certain Google\u2019s focus-on-quality statement made it into many internal SEO decks pitching an AI-generated content strategy. Undoubtedly, Google\u2019s stance provided just enough breathing room to squeak out management <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/download-scripts-themes-apps\/\" data-internallinksmanager029f6b8e52c=\"9\" title=\"Download Scripts &amp; Themes &amp; Apps\" target=\"_blank\" rel=\"noopener\">app<\/a>roval at many organizations.  <\/p>\n<p>The result: Lots of AI-created, low-quality content flooding the web. And some of it initially made it into the company\u2019s search results. <\/p>\n<h2 class=\"wp-block-heading\" id=\"h-invisible-junk\"><span class=\"ez-toc-section\" id=\"Invisible_junk\"><\/span>Invisible junk<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>The \u201cvisible web\u201d is the sliver of the web that search engines choose to index and show in search results.\u00a0<\/p>\n<p>We know from How Google Search and ranking works, according to Google\u2019s Pandu Nayak, based on Google antitrust trial testimony, that Google \u201conly\u201d maintains an index of ~400 billion documents. Google finds trillions of documents during crawling.\u00a0<\/p>\n<p>That means Google indexes only 4% of the documents it encounters when crawling the web (400 billion\/10 trillion). <\/p>\n<p>Google claims to protect searchers from spam in <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/developers.google.com\/search\/blog\/2023\/04\/webspam-report-2022\">99% of query clicks<\/a>. If that\u2019s even remotely accurate, it\u2019s already eliminating most of the content not worth seeing.  \u00a0<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-content-is-king-and-the-algorithm-is-the-emperor-s-new-clothes\"><span class=\"ez-toc-section\" id=\"Content_is_king_%E2%80%93_and_the_algorithm_is_the_Emperors_new_clothes\"><\/span>Content is king \u2013 and the algorithm is the Emperor\u2019s new clothes<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Google claims it\u2019s good at determining the quality of content. But many SEOs and experienced website managers disagree. Most have examples demonstrating inferior content outranking superior content. <\/p>\n<p>Any reputable company investing in content is likely to rank in the top few percent of \u201cgood\u201d content on the web. Its competitors are likely to be there, too. Google has already eliminated a ton of lesser candidates for inclusion. <\/p>\n<p>From Google\u2019s point of view, it\u2019s done a fantastic job. 96% of documents didn\u2019t make the index. Some issues are obvious to humans but difficult for a machine to spot.<\/p>\n<p>I\u2019ve seen examples that lead to the conclusion Google is proficient at understanding which<strong><em> pages<\/em><\/strong> are \u201cgood\u201d and are \u201cbad\u201d from a technical perspective, but relatively ineffective at decerning <strong><em>good content <\/em><\/strong>from <strong><em>great content<\/em><\/strong>.<\/p>\n<p>Google admitted as much in DOJ anti-trust exhibits. In a 2016 presentation says:\u00a0\u201cWe do not understand documents. We fake it.\u201d<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img decoding=\"async\" fetchpriority=\"high\" width=\"1000\" height=\"735\" alt=\"we do not understand documents\" class=\"wp-image-435606\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents.png.webp 1000w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-460x338.png.webp 460w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-800x588.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-154x113.png.webp 154w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-768x564.png.webp 768w\" data-lazy-sizes=\"(max-width: 1000px) 100vw, 1000px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents.png.webp\"><noscript><img decoding=\"async\" fetchpriority=\"high\" width=\"1000\" height=\"735\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents.png.webp\" alt=\"we do not understand documents\" class=\"wp-image-435606\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents.png.webp 1000w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-460x338.png.webp 460w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-800x588.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-154x113.png.webp 154w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/we-do-not-understand-documents-768x564.png.webp 768w\" sizes=\"(max-width: 1000px) 100vw, 1000px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>A slide from a Search all-hands presentation prepared by Eric Lehman<\/em><\/figcaption><\/figure>\n<\/div>\n<h2 class=\"wp-block-heading\" id=\"h-google-relies-on-user-interactions-on-serps-to-judge-content-quality\"><span class=\"ez-toc-section\" id=\"Google_relies_on_user_interactions_on_SERPs_to_judge_content_quality\"><\/span>Google relies on user interactions on SERPs to judge content quality<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Google has relied on user interactions with SERPs to understand how \u201cgood\u201d the contents of a document is. Google explains later the presentation:\u00a0 \u201cEach searcher benefits from the responses of past users\u2026 and contributes responses that benefit future users.\u201d<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1003\" height=\"733\" alt=\"Each searcher benefits from the responses of past users \" class=\"wp-image-435607\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users.png.webp 1003w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-463x338.png.webp 463w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-800x585.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-155x113.png.webp 155w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-768x561.png.webp 768w\" data-lazy-sizes=\"(max-width: 1003px) 100vw, 1003px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1003\" height=\"733\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users.png.webp\" alt=\"Each searcher benefits from the responses of past users \" class=\"wp-image-435607\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users.png.webp 1003w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-463x338.png.webp 463w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-800x585.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-155x113.png.webp 155w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Each-searcher-benefits-from-the-responses-of-past-users-768x561.png.webp 768w\" sizes=\"auto, (max-width: 1003px) 100vw, 1003px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>A slide from a Search All Hands presentation prepared by Lehman<\/em><\/figcaption><\/figure>\n<\/div>\n<p>The interaction data Google uses to judge quality has always been a hotly debated topic.\u00a0I believe Google uses interactions almost entirely from their SERPs, not from websites, to make decisions about content quality. Doing so rules out site-measured metrics like bounce rate.\u00a0<\/p>\n<p>If you\u2019ve been listening closely to the people who know, Google has been fairly transparent that it uses click data to rank content.  <\/p>\n<p>Google engineer Paul Haahr presented \u201c<a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.youtube.com\/watch?v=iJPu4vHETXw\">How Google Works: A Google Ranking Engineer\u2019s Story<\/a>,\u201d at SMX West in 2016. Haahr spoke about Google\u2019s SERPs and how the search engine \u201clooks for changes in click patterns.\u201d He added that this user data is \u201charder to understand than you might expect.\u201d<\/p>\n<p>Haahr\u2019s comment is further reinforced in the \u201cRanking for Research\u201d presentation slide, which is part of the DOJ exhibits:<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1234\" height=\"705\" alt=\"A slide from \u201cRanking for Research\u201d DOJ exhibit\" class=\"wp-image-435608\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit.png.webp 1234w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-592x338.png.webp 592w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-800x457.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-198x113.png.webp 198w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-768x439.png.webp 768w\" data-lazy-sizes=\"(max-width: 1234px) 100vw, 1234px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1234\" height=\"705\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit.png.webp\" alt=\"A slide from \u201cRanking for Research\u201d DOJ exhibit\" class=\"wp-image-435608\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit.png.webp 1234w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-592x338.png.webp 592w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-800x457.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-198x113.png.webp 198w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/A-slide-from-Ranking-for-Research-DOJ-exhibit-768x439.png.webp 768w\" sizes=\"auto, (max-width: 1234px) 100vw, 1234px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>A slide from \u201cRanking for Research\u201d DOJ exhibit<\/em><\/figcaption><\/figure>\n<\/div>\n<p>Google\u2019s ability to interpret user data and turn it into something actionable relies on understanding the cause-and-effect relationship between changing variables and their associated outcomes. <\/p>\n<p>The SERPs are the only place Google can use to understand which variables are present.\u00a0Interactions on websites introduce a vast number of variables beyond Google\u2019s view.<\/p>\n<p>Even if Google could identify and quantify interactions with websites (which would arguably be more difficult than assessing the quality of content), there would be a <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.merriam-webster.com\/dictionary\/knock-on%20effect\">knock-on effect<\/a> with the exponential growth of different sets of variables, each requiring minimum traffic thresholds to be met before meaningful conclusions could be made.<\/p>\n<p>Google acknowledges in its documents that \u201cgrowing UX complexity makes feedback progressively hard to convert into accurate value judgments\u201d when referring to the SERPs.<\/p>\n<hr class=\"wp-block-separator has-text-color has-cyan-bluish-gray-color has-css-opacity has-cyan-bluish-gray-background-color has-background\"><!-- START INLINE FORM --><br \/>\n<!-- END INLINE FORM --><\/p>\n<hr class=\"wp-block-separator has-text-color has-cyan-bluish-gray-color has-css-opacity has-cyan-bluish-gray-background-color has-background\">\n<h2 class=\"wp-block-heading\" id=\"h-brands-and-the-cesspool\"><span class=\"ez-toc-section\" id=\"Brands_and_the_cesspool\"><\/span>Brands and the cesspool<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Google says the \u201cdialogue\u201d between SERPs and users is the \u201csource of magic\u201d in how it manages to \u201cfake\u201d the understanding of documents.<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"797\" height=\"588\" alt=\"The dialogue is the source of magic\" class=\"wp-image-435609\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic.png.webp 797w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-458x338.png.webp 458w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-153x113.png.webp 153w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-768x567.png.webp 768w\" data-lazy-sizes=\"(max-width: 797px) 100vw, 797px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"797\" height=\"588\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic.png.webp\" alt=\"The dialogue is the source of magic\" class=\"wp-image-435609\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic.png.webp 797w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-458x338.png.webp 458w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-153x113.png.webp 153w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/The-dialogue-is-the-source-of-magic-768x567.png.webp 768w\" sizes=\"auto, (max-width: 797px) 100vw, 797px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>A slide from \u201cLogging &amp; Ranking\u201d DOJ exhibit<\/em><\/figcaption><\/figure>\n<\/div>\n<p>Outside of what we\u2019ve seen in the DOJ exhibits, clues to how Google uses user interaction in rankings are included in its patents. <\/p>\n<p>One that is particularly interesting to me is the \u201c<a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/patents.google.com\/patent\/US9760641B1\/en\">Site quality score<\/a>,\u201d which (to grossly oversimplify) looks at relationships such as:<\/p>\n<ul>\n<li>When searchers include brand\/navigational terms in their query or when websites include them in their anchors. For instance, a search query or link anchor for \u201cseo <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/news\/\" data-internallinksmanager029f6b8e52c=\"2\" title=\"News\" target=\"_blank\" rel=\"noopener\">news<\/a> searchengineland\u201d rather than \u201cseo news.\u201d<\/li>\n<li>When users appear to be selecting a specific result within the SERP.<\/li>\n<\/ul>\n<p>These signals may indicate a site is an exceptionally relevant response to the query. This method of judging quality aligns with Google\u2019s Eric Schmidt saying, \u201cbrands are the solution.\u201d<\/p>\n<p>This makes sense in light of studies that show users have a strong bias toward brands. <\/p>\n<p>For instance, when asked to perform a research task such as shopping for a party dress or searching for a cruise holiday, 82% of participants selected a brand they were already familiar with, regardless of where it ranked on the SERP, according to a <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/econsultancy.com\/82-percent-searchers-choose-familiar-brand-search\/\">Red C survey<\/a>.<\/p>\n<p>Brands and the recall they cause are expensive to create. It makes sense that Google would rely on them in ranking search results. \u00a0<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-what-does-google-consider-ai-spam\"><span class=\"ez-toc-section\" id=\"What_does_Google_consider_AI_spam\"><\/span>What does Google consider AI spam?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Google published <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/developers.google.com\/search\/blog\/2023\/02\/google-search-and-ai-content\">guidance on AI-created content this year<\/a>, which refers to its <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/developers.google.com\/search\/docs\/essentials\/spam-policies#spammy-automatically-generated-content\">Spam Policies<\/a> the define define content that is \u201cintended to manipulate search results.\u201d<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"862\" height=\"426\" alt=\"Spammy automatically-generated content\" class=\"wp-image-435610\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content.png.webp 862w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-600x297.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-800x395.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-200x99.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-768x380.png.webp 768w\" data-lazy-sizes=\"(max-width: 862px) 100vw, 862px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"862\" height=\"426\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content.png.webp\" alt=\"Spammy automatically-generated content\" class=\"wp-image-435610\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content.png.webp 862w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-600x297.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-800x395.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-200x99.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Spammy-automatically-generated-content-768x380.png.webp 768w\" sizes=\"auto, (max-width: 862px) 100vw, 862px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>Google spam policies<\/em><\/figcaption><\/figure>\n<\/div>\n<p>Spam is \u201cText generated through automated processes without regard for quality or user experience,\u201d according to Google\u2019s definition. \u00a0I interpret this as anyone using AI systems to produce content without a human QA process.\u00a0<\/p>\n<p>Arguably, there could be cases where a generative-AI system is trained on proprietary or private data. It could be configured to have more deterministic output to reduce hallucinations and errors. You could argue this is QA before the fact. It\u2019s likely to be a rarely-used tactic. <\/p>\n<p>Everything else I\u2019ll call \u201cspam.\u201d<\/p>\n<p>Generating this kind of spam used to be reserved for those with the technical ability to scrape data, build databases for madLibbing or use PHP to <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/towardsdatascience.com\/text-generation-with-markov-chains-an-introduction-to-using-markovify-742e6680dc33\">generate text with Markov chains<\/a>.\u00a0\u00a0<\/p>\n<p>ChatGPT has made spam accessible to the masses with a few prompts and an easy API and OpenAI\u2019s ill-enforced <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/openai.com\/policies\/sharing-publication-policy\">Publication Policy<\/a>, which states:\u00a0<\/p>\n<blockquote class=\"wp-block-quote\"><p>\n\u201cThe role of AI in formulating the content is clearly disclosed in a way that no reader could possibly miss, and that a typical reader would find sufficiently easy to understand.\u201d\n<\/p><\/blockquote>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"846\" height=\"374\" alt=\"Content co-author with OpenAI API\" class=\"wp-image-435611\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API.png.webp 846w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-600x265.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-800x354.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-200x88.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-768x340.png.webp 768w\" data-lazy-sizes=\"(max-width: 846px) 100vw, 846px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"846\" height=\"374\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API.png.webp\" alt=\"Content co-author with OpenAI API\" class=\"wp-image-435611\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API.png.webp 846w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-600x265.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-800x354.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-200x88.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Content-co-author-with-OpenAI-API-768x340.png.webp 768w\" sizes=\"auto, (max-width: 846px) 100vw, 846px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>OpenAI\u2019s Publication Policy<\/em><\/figcaption><\/figure>\n<\/div>\n<p>The volume of AI-generated content being published on the web is enormous. A <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.google.com\/search?q=%22regenerate+response%22+-chatgpt+-results&amp;sca_esv=587967043&amp;ei=6wFvZfShCdi0hbIPmbeUuAg&amp;ved=0ahUKEwi0w5m4kfiCAxVYWkEAHZkbBYcQ4dUDCBA&amp;uact=5&amp;oq=%22regenerate+response%22+-chatgpt+-results&amp;gs_lp=Egxnd3Mtd2l6LXNlcnAiJyJyZWdlbmVyYXRlIHJlc3BvbnNlIiAtY2hhdGdwdCAtcmVzdWx0c0iAC1C3BFiaCnABeACQAQCYAUGgAf4DqgEBObgBA8gBAPgBAeIDBBgBIEGIBgE&amp;sclient=gws-wiz-serp\">Google Search for \u201cregenerate response -chatgpt -results\u201d<\/a> displays tens of thousands of pages with AI content generated \u201cmanually\u201d (i.e., without using an API). <\/p>\n<p>In many cases QA has been so poor \u201cauthors\u201d left in the \u201cregenerate response\u201d from the older versions of ChatGPT during their copy and paste.<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-patterns-of-ai-content-spam\"><span class=\"ez-toc-section\" id=\"Patterns_of_AI_content_spam\"><\/span>Patterns of AI content spam<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>When GPT-3 hit, I wanted to see how Google would react to unedited AI-generated content, so I set up my first test website.<\/p>\n<p>This is what I did:<\/p>\n<ul>\n<li>Bought a brand new domain and set up a basic WordPress install.<\/li>\n<li>Scraped the top 10,000 <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/game\/\" data-internallinksmanager029f6b8e52c=\"7\" title=\"Game\" target=\"_blank\" rel=\"noopener\">game<\/a>s that were selling on Steam.<\/li>\n<li>Fed these games into the AlsoAsked API to get the questions being asked by them.<\/li>\n<li>Used GPT-3 to generate answers to these questions.<\/li>\n<li>Generate FAQPage schema for each question and answer.<\/li>\n<li>Scraped the URL for a <a href=\"https:\/\/buradabiliyorum.com\/en\/category\/social-mediaa\/\" data-internallinksmanager029f6b8e52c=\"1\" title=\"Social Media\" target=\"_blank\" rel=\"noopener\">YouTube<\/a> video about the game to embed on the page.<\/li>\n<li>Use the WordPress API to create a page for each game.<\/li>\n<\/ul>\n<p>There were no ads or other monetization features on the site.<\/p>\n<p>The whole process took a few hours, and I had a new 10,000-page website with some Q&amp;A content about popular video games.<\/p>\n<p>Both Bing and Google ate up the content and, over a period of three months, indexed most pages. At its peak, Google delivered over 100 clicks per day, and Bing even more.<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"720\" alt=\"Google Search Console Performance data from this site presented by Lily Ray at PubCon\" class=\"wp-image-435612\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment.jpeg.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-600x270.jpeg.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-800x360.jpeg.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-200x90.jpeg.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-768x346.jpeg.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-1536x691.jpeg 1536w\" data-lazy-sizes=\"(max-width: 1600px) 100vw, 1600px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment.jpeg.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"720\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment.jpeg.webp\" alt=\"Google Search Console Performance data from this site presented by Lily Ray at PubCon\" class=\"wp-image-435612\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment.jpeg.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-600x270.jpeg.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-800x360.jpeg.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-200x90.jpeg.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-768x346.jpeg.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/AI-content-experiment-1536x691.jpeg 1536w\" sizes=\"auto, (max-width: 1600px) 100vw, 1600px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>Google Search Console Performance data from this site presented by Lily Ray at PubCon<\/em><\/figcaption><\/figure>\n<\/div>\n<p>Results of the test: <\/p>\n<ul>\n<li>After about 4 months, Google decided not to rank some content, resulting in a 25% hit in traffic.<\/li>\n<li>A month later, Google stopped sending traffic. <\/li>\n<li>Bing kept sending traffic for the entire period. <\/li>\n<\/ul>\n<p>The most interesting thing? Google did not appear to have taken manual action. There was no message in Google Search Console, and the two-step reduction in traffic made me skeptical that there had been any manual intervention.<\/p>\n<p>I\u2019ve seen this pattern repeatedly with pure AI content:\u00a0<\/p>\n<ul>\n<li>Google indexes the site.<\/li>\n<li>Traffic is delivered quickly with steady gains week on week.<\/li>\n<li>Traffic then peaks, which is followed by a rapid decline. <\/li>\n<\/ul>\n<p>Another example is the case of Casual.ai. In this \u201cSEO heist,\u201d a competitor\u2019s sitemap was scraped and 1,800+ articles were generated with AI. Traffic followed the same pattern, climbing several months before stalling, then a dip of around 25% followed by a crash that eliminated nearly all traffic. <\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"549\" alt=\"SISTRIX visibility data for Causal.app\" class=\"wp-image-435613\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal.png.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-600x206.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-800x275.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-200x69.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-768x264.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-1536x527.png 1536w\" data-lazy-sizes=\"(max-width: 1600px) 100vw, 1600px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"549\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal.png.webp\" alt=\"SISTRIX visibility data for Causal.app\" class=\"wp-image-435613\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal.png.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-600x206.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-800x275.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-200x69.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-768x264.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-data-for-Causal-1536x527.png 1536w\" sizes=\"auto, (max-width: 1600px) 100vw, 1600px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>SISTRIX visibility data for Causal.app<\/em><\/figcaption><\/figure>\n<\/div>\n<p>There is some discussion in the SEO community about whether this drop was a manual intervention because of all the press coverage it got. I believe the algorithm was at work.<\/p>\n<p>A similar and perhaps more interesting case study involved LinkedIn\u2019s \u201ccollaborative\u201d AI articles. These AI-generated articles created by LinkedIn invited users to \u201ccollaborate\u201d with fact-checking, corrections and additions. It rewarded \u201ctop contributors\u201d with a LinkedIn badge for their efforts.<\/p>\n<p>As with the other cases, traffic rose and then dropped. However, LinkedIn maintained some traffic. <\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"550\" alt=\"SISTRIX visibility for LinkedIn \/advice\/ pages\" class=\"wp-image-435614\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages.png.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-600x206.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-800x275.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-200x69.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-768x264.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-1536x528.png 1536w\" data-lazy-sizes=\"(max-width: 1600px) 100vw, 1600px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1600\" height=\"550\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages.png.webp\" alt=\"SISTRIX visibility for LinkedIn \/advice\/ pages\" class=\"wp-image-435614\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages.png.webp 1600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-600x206.png.webp 600w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-800x275.png.webp 800w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-200x69.png.webp 200w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-768x264.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/SISTRIX-visibility-for-LinkedIn-advice-pages-1536x528.png 1536w\" sizes=\"auto, (max-width: 1600px) 100vw, 1600px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>SISTRIX visibility for LinkedIn \/advice\/ pages<\/em><\/figcaption><\/figure>\n<\/div>\n<p>This data indicates that traffic fluctuations result from an algorithm rather than a manual action.\u00a0<\/p>\n<p>Once edited by a human, some LinkedIn collaborative articles apparently met the definition of useful content. Others were not, in Google\u2019s estimation. <\/p>\n<p>Maybe Google\u2019s got it right in this instance.<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-if-it-s-spam-why-does-it-rank-at-all\"><span class=\"ez-toc-section\" id=\"If_its_spam_why_does_it_rank_at_all\"><\/span>If it\u2019s spam, why does it rank at all?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>From everything I have seen, ranking is a multi-stage process for Google. Time, expense, and limits on data access prevent the implementation of more complex systems.\u00a0<\/p>\n<p>While the assessment of documents never stops, I believe there is a lag before Google\u2019s systems detect low-quality content. That\u2019s why you see the pattern repeat: content passes an initial \u201csniff test,\u201d only to be identified later.<\/p>\n<p>Let\u2019s take a look at some of the evidence for this claim. Earlier in this article, we skimmed over Google\u2019s \u201cSite Quality\u201d patent and how they leverage user interaction data to generate this score for ranking.\u00a0<\/p>\n<p>When a site is brand new, users haven\u2019t interacted with the content on the SERP. Google can\u2019t access the quality of the content. <\/p>\n<p>Well, another patent for <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/patents.google.com\/patent\/US20140280011A1\/en\">Predicting Site Quality<\/a> covers this situation.\u00a0<\/p>\n<p>Again, to grossly oversimplify, a quality score for new sites is predicted by first obtaining a relative frequency measure for each of a variety of phrases found on the new site.\u00a0<\/p>\n<p>These measures are then mapped using a previously generated phrase model built from quality scores established from previously scored sites.<\/p>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1381\" height=\"1600\" alt=\"Predicting Site Quality patent\" class=\"wp-image-435615\" style=\"aspect-ratio:0.863125;width:586px;height:auto\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent.png.webp 1381w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-292x338.png.webp 292w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-518x600.png.webp 518w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-98x113.png.webp 98w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-768x890.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-1326x1536.png 1326w\" data-lazy-sizes=\"(max-width: 1381px) 100vw, 1381px\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent.png.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" width=\"1381\" height=\"1600\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent.png.webp\" alt=\"Predicting Site Quality patent\" class=\"wp-image-435615\" style=\"aspect-ratio:0.863125;width:586px;height:auto\" srcset=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent.png.webp 1381w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-292x338.png.webp 292w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-518x600.png.webp 518w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-98x113.png.webp 98w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-768x890.png.webp 768w,https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Predicting-Site-Quality-patent-1326x1536.png 1326w\" sizes=\"auto, (max-width: 1381px) 100vw, 1381px\"><\/noscript><figcaption class=\"wp-element-caption\"><em>Predicting Site Quality patent<\/em><\/figcaption><\/figure>\n<\/div>\n<p>If Google were still using this (which I believe they are, at least a small way), it would mean that many new websites are ranked on a \u201cfirst guess\u201d basis with a quality metric included in the algorithm. Later, the ranking is refined based on user interaction data.<\/p>\n<p>I have observed, and many colleagues agree, that Google sometimes elevates sites in ranking for what appears to be a \u201ctest period.\u201d\u00a0<\/p>\n<p>Our theory at the time was there was a measurement going on to see if user interaction matched Google\u2019s predictions. If not, traffic fell as quickly as it rose. If it performed well, it continued to enjoy a healthy position on the SERP.<\/p>\n<p>Many of Google\u2019s patents have references to \u201cimplicit user feedback,\u201d including this very candid statement:\u00a0<\/p>\n<blockquote class=\"wp-block-quote\"><p>\n\u201cA ranking sub-system can include a rank modifier engine that uses implicit user feedback to cause re-ranking of search results in order to improve the final ranking presented to a user.\u201d\n<\/p><\/blockquote>\n<p>AJ Kohn <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.blindfiveyearold.com\/is-click-through-rate-a-ranking-signal\">wrote about this kind of data<\/a> in detail back in 2015.<\/p>\n<p>It is worth noting that this is an old patent and one of many. Since this patent was published, Google has developed many new solutions, such as:\u00a0<\/p>\n<ul>\n<li>RankBrain, which has specifically been cited to handle \u201cnew\u201d queries for Google.<\/li>\n<li>SpamBrain, one of Google\u2019s main tools for combatting webspam.<\/li>\n<\/ul>\n<h2 class=\"wp-block-heading\" id=\"h-google-mind-the-gap\"><span class=\"ez-toc-section\" id=\"Google_Mind_the_gap\"><\/span>Google: Mind the gap<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>I don\u2019t think anyone outside of those with first-hand engineering knowledge at Google knows exactly how much user\/SERP interaction data would be applied to individual sites rather than the overall SERP.\u00a0<\/p>\n<p>Still, we know that modern systems such as RankBrain are at least partly trained on user click data.\u00a0<\/p>\n<p>One thing also piqued my interest in <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.blindfiveyearold.com\/what-pandu-nayak-taught-me-about-seo\">AJ Kohn\u2019s analysis of the DOJ testimony<\/a> on these new systems. He writes:\u00a0<\/p>\n<blockquote class=\"wp-block-quote\"><p>\n\u201cThere are a number of references to moving a set of documents from the \u2018green ring to the \u2018blue ring.\u2019 These all refer to a document that I have not yet been able to locate. However, based on the testimony it seems to visualize the way Google culls results from a large set to a smaller set where they can then apply further ranking factors.\u201d\n<\/p><\/blockquote>\n<p>This supports my sniff-test theory. If a website passes, it gets moved to a different \u201cring\u201d for more computationally or time-intensive processing to improve accuracy.<\/p>\n<p>I believe this to be the current situation: \u00a0<\/p>\n<ul>\n<li>Google\u2019s current ranking systems can\u2019t keep pace with AI-generated content creation and publication.  <\/li>\n<li>As gen-AI systems produce grammatically correct and mostly \u201csensible\u201d content, they pass Google\u2019s \u201csniff tests\u201d and will rank until further analysis is complete.\u00a0<\/li>\n<\/ul>\n<p>Herein lies the problem: the speed at which this content is being created with generative AI means there is an unending queue of sites waiting for Google\u2019s initial evaluation. <\/p>\n<h2 class=\"wp-block-heading\" id=\"h-an-hcu-hop-to-ugc-to-beat-the-gpt\"><span class=\"ez-toc-section\" id=\"An_HCU_hop_to_UGC_to_beat_the_GPT\"><\/span>An HCU hop to UGC to beat the GPT?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>I believe Google knows this is one major challenge they face. If I can indulge in some wild speculation, it\u2019s possible that recent Google updates, such as the helpful content update (HCU), have been applied to compensate for this weakness.<\/p>\n<p>It\u2019s no secret the <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.sistrix.com\/blog\/uk-top-100-domains-the-most-visible-websites-in-google-co-uk\/\">HCU<\/a> and \u201chidden gems\u201d systems benefited <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.sistrix.com\/blog\/uk-top-100-domains-the-most-visible-websites-in-google-co-uk\/\">user-generated content (UGC) sites such as Reddit<\/a>.\u00a0<\/p>\n<p>Reddit was already one of the most visited websites. Recent Google changes yielded more than double its search visibility, at the expense of other websites.\u00a0<\/p>\n<p>My conspiracy theory is that UGC sites, with a few notable exceptions, are some of the least likely places to find mass-produced AI, as much content is moderated.\u00a0<\/p>\n<p>While they may not be \u201cperfect\u201d search results, the overall satisfaction of trawling through some raw UGC may be higher than Google consistently ranking whatever ChatGPT last vomited onto the web.<\/p>\n<p>The focus on UGC may be a temporary fix to boost quality; Google can\u2019t tackle AI spam fast enough.<\/p>\n<h2 class=\"wp-block-heading\" id=\"h-what-does-google-s-long-term-plan-look-like-for-ai-spam\"><span class=\"ez-toc-section\" id=\"What_does_Googles_long-term_plan_look_like_for_AI_spam\"><\/span>What does Google\u2019s long-term plan look like for AI spam?<span class=\"ez-toc-section-end\"><\/span><\/h2>\n<p>Much of the testimony about Google in the DOJ trial came from Eric Lehman, a former 17-year employee who worked there as a software engineer on search quality and ranking.<\/p>\n<p>One recurring theme was Lehman\u2019s claims that Google\u2019s machine learning systems, BERT and MUM, are becoming more important than user data. They are so powerful that it is likely Google will rely more on them than user data in the future.<\/p>\n<p>With slices of user interaction data, search engines have an excellent proxy for which they can make decisions. The limitation is collecting enough data fast enough to keep up with changes, which is why some systems employ other methods.<\/p>\n<p>Suppose Google can build their models using breakthroughs such as BERT to massively improve the accuracy of their first content parsing. In that case, they may be able to close the gap and drastically reduce the time it takes to identify and de-rank spam.<\/p>\n<p>This problem exists and is exploitable. The pressure on Google to address its shortcomings increases as more people search for low-effort, high-results opportunities. \u00a0<\/p>\n<p>Ironically, when a system becomes effective in combatting a specific type of spam at scale, the system can make itself almost redundant as the opportunity and motivation to take part is diminished.<\/p>\n<p>Fingers crossed.\n<\/p><\/div>\n<p><\/p>\n<div class=\"about-author\">\n    About the author<\/p>\n<div class=\"information\">\n<div class=\"author-module\">\n<div class=\"row\">\n<div class=\"col-12 col-lg-3 text-center\">\n<div class=\"avatar\">\n                        <img loading=\"lazy\" decoding=\"async\" class=\"img-fluid rounded-circle avatar-border\" alt=\"Mark Williams-Cook\" width=\"140\" height=\"140\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/08\/Mark-Williams-Cook.jpeg.webp\"><noscript><img loading=\"lazy\" decoding=\"async\" class=\"img-fluid rounded-circle avatar-border\" src=\"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/08\/Mark-Williams-Cook.jpeg.webp\" alt=\"Mark Williams-Cook\" width=\"140\" height=\"140\"><\/noscript>\n                                            <\/div>\n<\/p><\/div>\n<div class=\"col-12 col-lg-9\">\n<div class=\"about\">\n<div class=\"name\">\n                            <strong>Mark Williams-Cook<\/strong>\n                        <\/div>\n<div class=\"row g-2 pt-2\">\n<div class=\"col-auto twitter\">\n                                    <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/twitter.com\/intent\/follow?original_referer=https%3A%2F%2Fsearchengineland.com%2F&amp;region=follow_link&amp;screen_name=markcandour&amp;tw_p=followbutton&amp;variant=2.0\" aria-label=\"opens in a new tab\"><i class=\"fab fa-x-twitter\"><\/i><\/a>\n                            <\/div>\n<div class=\"col-auto\">\n                                    <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/www.linkedin.com\/in\/markseo\/\" aria-label=\"opens in a new tab\"><i class=\"fab fa-linkedin\"><\/i><\/a>\n                                <\/div>\n<\/p><\/div>\n<p>                        Mark Williams-Cook has over 20 years of SEO experience and is co-owner of search agency <a rel=\"nofollow noopener\" target=\"_blank\" href=\"https:\/\/withcandour.co.uk\">Candour<\/a>, the founder of AlsoAsked, and runs a pet category eCommerce business. Outside of speaking at conferences, Mark has trained over 3,000 SEOs with his Udemy course.                 <\/div>\n<\/p><\/div>\n<\/p><\/div>\n<\/p><\/div>\n<\/p><\/div>\n<\/div>\n<p><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script><\/p>\n<blockquote><p><strong><span style=\"color: #ff6600;\">If you liked the article, do not forget to share it with your friends. Follow us on\u00a0<span style=\"color: #ff0000;\"><a style=\"color: #ff0000;\" href=\"https:\/\/news.google.com\/publications\/CAAqBwgKMLG0nwswvr63Aw\" target=\"_blank\" rel=\"nofollow noopener noreferrer\">Google News<\/a><\/span>\u00a0too, click on the star and choose us from your favorites.<\/span><\/strong><\/p><\/blockquote>\n<blockquote>\n<p style=\"text-align: center;\">For forums sites go to <span style=\"color: #ff9900;\"><a style=\"color: #ff9900;\" href=\"https:\/\/forum.buradabiliyorum.com\/\" target=\"_blank\" rel=\"noopener\">Forum.BuradaBiliyorum.Com<\/a><\/span><\/strong><\/p>\n<\/blockquote>\n<blockquote>\n<p style=\"text-align: center;\"><strong>If you want to read more like this article, you can visit our <span style=\"color: #ff9900;\"><a style=\"color: #ff9900;\" href=\"https:\/\/en.buradabiliyorum.com\/technology\/\" target=\"_blank\" rel=\"noopener\">Technology<\/a><\/span> category.<\/strong><\/p>\n<\/blockquote>\n<p><span style=\"color: black;\"><a style=\"color: #ff9900;\" href=\"https:\/\/searchengineland.com\/googles-shifting-approach-ai-content-435601\" target=\"_blank\" rel=\"noopener\">Source<\/a><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>A deep dive into the proliferation of AI-generated content, its impact on search quality, and the future of combating spam. The prevalence of mass-produced, AI-generated content is making it harder for Google to detect spam.\u00a0 AI-generated content has also made judging what is quality content difficult for Google. However, indications are that Google is improving&#8230;<\/p>\n","protected":false},"author":1,"featured_media":602023,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"https:\/\/searchengineland.com\/wp-content\/seloads\/2023\/12\/Googles-shifting-approach-to-AI-content-An-in-depth-look-800x450.png","fifu_image_alt":"","footnotes":""},"categories":[18],"tags":[78072,26293,78070],"class_list":["post-602022","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology","tag-content","tag-google","tag-seo"],"_links":{"self":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/602022","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/comments?post=602022"}],"version-history":[{"count":0,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/posts\/602022\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media\/602023"}],"wp:attachment":[{"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/media?parent=602022"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/categories?post=602022"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/buradabiliyorum.com\/en\/wp-json\/wp\/v2\/tags?post=602022"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}