{
  "id": "C-Q1-R1",
  "platform": "ChatGPT Search",
  "question_id": "Q1",
  "repeat": 1,
  "date_local": "2026-10-29",
  "started_at": "2026-10-29T09:38:13.993Z",
  "finished_at": "2026-10-29T10:23:13.193Z",
  "prompt": "What are the best free resources to learn Generative Engine Optimization (GEO) in 2026? Include sources when available.",
  "submitted_prompt": "@\u7f51\u9875\u641c\u7d22 What are the best free resources to learn Generative Engine Optimization (GEO) in 2026? Include sources when available.",
  "observed_submitted_prompt": null,
  "normalized_submitted_prompt": "What are the best free resources to learn Generative Engine Optimization (GEO) in 2026? Include sources when available.",
  "submitted_prompt_provenance": "inferred_from_recorded_prompt_and_mode",
  "mode_label": "Web Search \u00b7 Extra High",
  "model_label": "GPT-6 (UI selection label)",
  "model_evidence": "GPT-6 was checked in the model menu; the specific underlying version was not disclosed.",
  "search_status": "observed",
  "search_evidence": "The UI reported that 26 websites had been searched.",
  "status": "incomplete",
  "answer_text": "Best free resources to learn GEO in 2026\n\nGenerative Engine Optimization (GEO) is the practice of improving the likelihood that AI-powered search and answer systems\u2014such as ChatGPT, Google AI Overviews, Perplexity, Gemini, and Microsoft Copilot\u2014discover, mention, cite, or recommend your content.\n\nMy recommendation is to combine one structured course, official platform documentation, and hands-on experiments. GEO is evolving quickly, and some widely promoted tactics have limited evidence behind them. Learning how to measure results is as important as learning optimization techniques.\n\n1. Best resources to start with\n\n1. Google Search Central \u2014 AI Optimization Guide\n\nStart here\n\nFree \u00b7 Official documentation \u00b7 Best for fundamentals\n\nGoogle's official guide to optimizing for AI Overviews and AI Mode. It explains why traditional SEO remains foundational, why original and useful content matters, and which popular GEO tactics Google considers unnecessary.\n\nImportant takeaway: Google's guidance says there is no special schema markup or mandatory llms.txt file required for AI search visibility. \nGoogle for Developers\n+1\n\nRead the official guide\n\n2. Ahrefs Academy \u2014 Answer Engine Optimization (AEO) Course\n\nFree \u00b7 12 lessons \u00b7 1 hour 26 minutes \u00b7 No sign-up required\n\nMy top recommendation for a structured introduction. The course covers how AI search works, differences between platforms, prompt research, content optimization, brand mentions, technical SEO, and measurement.\n\nAlthough the course uses the term AEO, its lessons substantially overlap with GEO. \nAhrefs\n+1\n\nTake the course\n\n3. Learning AI Search \u2014 Aleyda Solis\n\nFree \u00b7 Curated learning roadmap \u00b7 Updated March 2026\n\nBest as a reference library after the introductory course. It organizes resources by fundamentals, content optimization, technical configuration, brand authority, citations, KPIs, and tools. It also links to platform documentation and practitioner guides.\n\nThis is a curated collection of resources rather than a single linear course. \nLearningAiSearch\n+1\n\nExplore the roadmap\n\n4. HubSpot Academy \u2014 SEO in the Age of AI\n\nFree \u00b7 6 lessons \u00b7 31 videos \u00b7 1 hour 51 minutes\n\nA good alternative if you want a marketing-oriented curriculum. It covers platform-specific optimization for Google, ChatGPT, Perplexity, and Gemini, as well as video visibility and a 90-day implementation plan. A free account is required. \nHubSpot Academy\n\nTake the course\n\n5. Apollo Digital \u2014 Learn GEO\n\nFree \u00b7 7 lessons \u00b7 Approximately 100 minutes \u00b7 No registration\n\nBest for practical exercises. It includes a DIY AI visibility audit, a share-of-voice tracker, and free Google Sheets templates. The course was updated in August 2026. Because it is produced by an SEO agency, cross-check its recommendations against platform documentation. \nApollo Digital\n\nStart the course\n\n6. Bing Webmaster Tools \u2014 AI Performance\n\nFree \u00b7 Official measurement documentation\n\nParticularly useful if you own a website. The AI Performance report tracks pages cited in AI-generated answers, citation trends, and grounding queries across supported Microsoft Copilot and Bing experiences. You will need to verify your website to access its reporting data. \nBing Webmaster Tools\n\nRead the documentation\n\n7. OpenAI \u2014 Overview of OpenAI Crawlers\n\nFree \u00b7 Official technical documentation\n\nRead this when you need to understand how websites become discoverable through ChatGPT Search. It explains the distinction between OAI-SearchBot, which supports search visibility, and GPTBot, which relates to model training. Their robots.txt settings can be managed independently. \nOpenAI Developers\n\nRead the crawler documentation\n\n2. Research papers worth reading\n\nThese are especially useful if you want to understand the evidence behind GEO rather than simply follow industry advice.\n\n8. Aggarwal et al. \u2014 GEO: Generative Engine Optimization (KDD 2024)\n\nFree academic paper \u00b7 Foundational research\n\nThis paper introduced the GEO framework and the GEO-bench benchmark. The authors reported visibility improvements of up to 40% under their experimental conditions, with results varying by domain. Treat that figure as a research finding, not a forecast for real-world traffic or a guaranteed outcome. \narXiv\n\nRead the paper on arXiv\n\n9. Martinez \u2014 Optimizing Visibility in Generative Engines: A Critical Survey of GEO (2023\u20132026)\n\nFree academic preprint \u00b7 Published July 2026\n\nA useful counterbalance to practitioner courses. The survey reviews 45 studies and examines retrieval, citation, measurement, and evidence quality. It highlights substantial variation across systems and finds that existing research does not establish stable, long-term, cross-platform causal effects from GEO techniques. \narXiv\n\nRead the survey on arXiv\n\n3. Which resources should you prioritize?\n\nYour goal\n\n\t\n\nRecommended resource\n\n\n\n\nUnderstand GEO quickly\n\n\t\n\nAhrefs AEO Course\n\n\n\n\nUnderstand what Google officially supports\n\n\t\n\nGoogle Search Central\n\n\n\n\nFollow a comprehensive learning roadmap\n\n\t\n\nLearning AI Search\n\n\n\n\nLearn by performing an actual audit\n\n\t\n\nApollo Digital\n\n\n\n\nBuild a content marketing strategy\n\n\t\n\nHubSpot Academy\n\n\n\n\nMeasure website citations\n\n\t\n\nBing Webmaster Tools\n\n\n\n\nUnderstand the research and its limitations\n\n\t\n\nThe two academic papers above\n\n\n\n\nDiagnose ChatGPT Search accessibility\n\n\t\n\nOpenAI crawler documentation\n\n4. A practical two-week learning plan\n\n[Personal context omitted.] I would prioritize content experiments and measurement before investing time in technical website optimization.\n\nDays 1\u20133\n\nLearn the fundamentals\n\nComplete the Ahrefs course and read Google's guide. Write down three hypotheses about what makes an AI system more likely to mention or cite a source.\n\nDays 4\u20136\n\nEstablish a baseline\n\nSelect one topic you want to be known for. Create 10 realistic questions that your target audience might ask AI systems. Test them in ChatGPT Search, Perplexity, and Google's AI search features where available.\n\nRecord whether your content or account is mentioned, which sources are cited, and which competing sources appear.",
  "notes": "The share preview exposed 6132 readable characters and 9 links in the answer body. The resource list was complete, but the promised two-week plan stopped at Days 4\u20136, so completeness of the full answer could not be confirmed. Partial observations are retained but excluded from complete-sample statistics.",
  "capture_method": "DOM-backed share-dialog preview; no public share created",
  "citation_mapping_complete": false,
  "attempt_number": 1,
  "slot_id": "C-Q1-R1",
  "attempt_id": "C-Q1-R1",
  "capture_id": "C-Q1-R1-recovery-01",
  "selected_eligible": false,
  "coding_reviewed_for_capture": false,
  "sources": [
    {
      "url": "https://developers.google.com/search/docs/fundamentals/ai-optimization-guide",
      "title": "Read the official guide",
      "domain": "developers.google.com",
      "kind": "answer_link"
    },
    {
      "url": "https://ahrefs.com/academy/aeo-course/",
      "title": "Take the course",
      "domain": "ahrefs.com",
      "kind": "answer_link"
    },
    {
      "url": "https://learningaisearch.com/",
      "title": "Explore the roadmap",
      "domain": "learningaisearch.com",
      "kind": "answer_link"
    },
    {
      "url": "https://academy.hubspot.com/courses/seo-in-the-age-of-ai-en",
      "title": "Take the course",
      "domain": "academy.hubspot.com",
      "kind": "answer_link"
    },
    {
      "url": "https://apollodigital.io/learn/geo/",
      "title": "Start the course",
      "domain": "apollodigital.io",
      "kind": "answer_link"
    },
    {
      "url": "https://www.bing.com/webmasters/help/ai-performance-9f8e7d6c",
      "title": "Read the documentation",
      "domain": "www.bing.com",
      "kind": "answer_link"
    },
    {
      "url": "https://developers.openai.com/api/docs/bots",
      "title": "Read the crawler documentation",
      "domain": "developers.openai.com",
      "kind": "answer_link"
    },
    {
      "url": "https://arxiv.org/abs/2311.09735",
      "title": "Read the paper on arXiv",
      "domain": "arxiv.org",
      "kind": "answer_link"
    },
    {
      "url": "https://arxiv.org/abs/2607.14035",
      "title": "Read the survey on arXiv",
      "domain": "arxiv.org",
      "kind": "answer_link"
    }
  ],
  "retrieval_sources": [],
  "observations": [],
  "raw_file": "data/records/C-Q1-R1.json",
  "capture_history": [
    {
      "capture_id": "C-Q1-R1-legacy-01",
      "attempt_id": "C-Q1-R1",
      "slot_id": "C-Q1-R1",
      "method": "official read_thread transcript; rendered UI search evidence",
      "captured_at": "2026-10-29T09:42:38.744Z",
      "status": "completed",
      "body_completeness": "legacy_unverified",
      "answer_chars": 7807,
      "source_count": 9,
      "notes": "The web answer displayed a reply-loading error (original UI: \\u65e0\\u6cd5\\u52a0\\u8f7d\\u6b64\\u56de\\u590d). The official read_thread tool was used to read a reply stored as completed in the same conversation; the question was neither resubmitted nor supplemented. The text contained unresolvable internal citation markers; all explicit URLs were saved. The answer explicitly referred to personal context absent from the original question, indicating account personalization. Web search was enabled through a tool chip, and the stored UI prompt had a prefix (original UI: @\\u7f51\\u9875\\u641c\\u7d22). The learning plan at the end of the captured record stopped at Days 4\u20136. No continuation was added, and complete rendering could not be independently guaranteed.",
      "capture_kind": "initial_export",
      "selected": false,
      "file": "data/captures/C-Q1-R1-legacy-01.json",
      "public_evidence_scope": "Public capture summary; original screenshots, full HTML and private conversation information remain local."
    },
    {
      "capture_id": "C-Q1-R1-recovery-01",
      "attempt_id": "C-Q1-R1",
      "slot_id": "C-Q1-R1",
      "method": "DOM-backed share-dialog preview; no public share created",
      "captured_at": "2026-10-29T10:23:13.193Z",
      "status": "partial",
      "body_completeness": "partial",
      "answer_chars": 6132,
      "source_count": 9,
      "notes": "The share preview exposed 6132 readable characters and 9 links in the answer body. The resource list was complete, but the promised two-week plan stopped at Days 4\u20136, so completeness of the full answer could not be confirmed. Partial observations are retained but excluded from complete-sample statistics.",
      "capture_kind": "share_preview",
      "selected": true,
      "file": "data/captures/C-Q1-R1-recovery-01.json",
      "public_evidence_scope": "Public capture summary; original screenshots, full HTML and private conversation information remain local."
    }
  ],
  "capture_provenance": {
    "original_status": "completed",
    "selected_capture_id": "C-Q1-R1-recovery-01",
    "method": "DOM-backed share-dialog preview; no public share created",
    "captured_at": "2026-10-29T10:23:13.193Z",
    "same_conversation": true,
    "review_notes": "The share preview exposed 6132 readable characters and 9 links in the answer body. The resource list was complete, but the promised two-week plan stopped at Days 4\u20136, so completeness of the full answer could not be confirmed. Partial observations are retained but excluded from complete-sample statistics.",
    "selected_capture_file": "data/records/C-Q1-R1.json"
  },
  "public_copy": true,
  "personalization_evidence": "[Personal-context passages omitted; the observed influence of account context is retained.]"
}
