[{"data":1,"prerenderedAt":552},["ShallowReactive",2],{"blog-\u002Fblog\u002Fhow-to-rank-in-chatgpt":3,"blog-siblings":500},{"id":4,"title":5,"author":6,"body":7,"category":482,"date":483,"description":484,"draft":485,"extension":486,"image":487,"meta":488,"navigation":489,"path":490,"readingTime":491,"seo":492,"stem":493,"tags":494,"updated":498,"__hash__":499},"blog\u002Fblog\u002Fhow-to-rank-in-chatgpt.md","How to rank in ChatGPT: what decides which sites it cites","devin",{"type":8,"value":9,"toc":460},"minimark",[10,14,22,25,30,41,50,58,62,71,132,139,148,157,161,164,173,182,191,195,198,207,210,214,219,232,236,245,249,252,256,265,268,272,279,283,286,339,347,350,358,365,373,377,390,396,406,410,414,417,421,427,431,434,438,444,448,456],[11,12,13],"p",{},"ChatGPT has no fixed list of results, so you can't rank in it the way you rank in Google. When it uses the web, it runs searches, opens some of the pages it finds, and writes an answer that cites what it read. So how to rank in ChatGPT comes down to being one of those pages, which is what people mean by ChatGPT SEO. Three things come first: whether OpenAI's search crawler can reach you, whether you turn up in the searches it runs, and whether the page it opens shows the answer without needing JavaScript.",[11,15,16,17,21],{},"I checked five guides that rank for \"how to rank in ChatGPT\" on 18 September 2026 (US, desktop). One cites OpenAI's own crawler documentation. None explains what a ",[18,19,20],"code",{},"ChatGPT-User"," hit in your server log means, and two still describe ChatGPT as ad-free. This guide stays close to what OpenAI has published and to studies that show their numbers.",[23,24],"cta-inline",{},[26,27,29],"h2",{"id":28},"how-chatgpt-search-works","How ChatGPT search works",[11,31,32,33,40],{},"Most ChatGPT conversations don't touch the web, by one measure. Profound ",[34,35,39],"a",{"href":36,"rel":37},"https:\u002F\u002Fwww.tryprofound.com\u002Fblog\u002Fchatgpt-citation-sources",[38],"nofollow","studied about 730,000 ChatGPT conversations"," from late 2025 that contained a web citation. The same study reports: \"About 18% of ChatGPT conversations trigger at least one web search.\" The rest are answered without a search, mostly from what the model learned in training, so a page you publish this week is unlikely to show up in them.",[11,42,43,44,49],{},"For searches made through its API, OpenAI's ",[34,45,48],{"href":46,"rel":47},"https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fguides\u002Ftools-web-search",[38],"developer documentation"," describes a loop. The model \"can perform web searches as part of its chain of thought, analyze results, and decide whether to keep searching.\" In plain terms, it can search, open a page, and look for something inside that page. OpenAI adds: \"The number of sources is often greater than the number of citations.\" A page can be fetched and still go uncited. That documentation covers the API. I haven't found the same detail published for the app.",[11,51,52,53,57],{},"Answers also differ from one run to the next. SparkToro and Gumshoe found less than a 1 in 100 chance of getting the same list of brands twice. So a rank in ChatGPT is a rate: the share of answers that include you. ",[34,54,56],{"href":55},"\u002Fblog\u002Fai-visibility","How to measure AI visibility"," covers the arithmetic.",[26,59,61],{"id":60},"oai-searchbot-chatgpt-user-and-gptbot-which-one-matters","OAI-SearchBot, ChatGPT-User and GPTBot: which one matters",[11,63,64,65,70],{},"OpenAI runs separate crawlers and ",[34,66,69],{"href":67,"rel":68},"https:\u002F\u002Fdevelopers.openai.com\u002Fapi\u002Fdocs\u002Fbots",[38],"documents each one",". Three matter here:",[72,73,74,90],"table",{},[75,76,77],"thead",{},[78,79,80,84,87],"tr",{},[81,82,83],"th",{},"Crawler",[81,85,86],{},"What it does",[81,88,89],{},"What blocking it costs you",[91,92,93,107,119],"tbody",{},[78,94,95,101,104],{},[96,97,98],"td",{},[18,99,100],{},"OAI-SearchBot",[96,102,103],{},"Surfaces sites in ChatGPT's search results",[96,105,106],{},"Your site \"will not be shown in ChatGPT search answers, though can still appear as navigational links\"",[78,108,109,113,116],{},[96,110,111],{},[18,112,20],{},[96,114,115],{},"Fetches a page when a person's request calls for it",[96,117,118],{},"OpenAI says robots.txt rules \"may not apply\", because a user started the action",[78,120,121,126,129],{},[96,122,123],{},[18,124,125],{},"GPTBot",[96,127,128],{},"Collects content that may be used to train models",[96,130,131],{},"No documented effect on search. It governs training use",[11,133,134,135,138],{},"A fourth, ",[18,136,137],{},"OAI-AdsBot",", checks the landing pages of ads.",[11,140,141,142,147],{},"OpenAI's ",[34,143,146],{"href":144,"rel":145},"https:\u002F\u002Fhelp.openai.com\u002Fen\u002Farticles\u002F12627856-publishers-and-developers-faq",[38],"publisher FAQ"," puts the requirement in one line: \"For your site content to be included in summaries and snippets in ChatGPT, make sure you aren't blocking OAI-SearchBot.\" It also says \"Any public website can appear in ChatGPT search.\" After you change robots.txt, the crawler documentation says it \"can take ~24 hours\" for OpenAI's systems to adjust.",[11,149,150,151,153,154,156],{},"Blocking ",[18,152,125],{}," to stay out of training data is a fair choice. Blocking ",[18,155,100],{}," by the same rule takes you out of ChatGPT's search answers.",[26,158,160],{"id":159},"where-chatgpt-searches","Where ChatGPT searches",[11,162,163],{},"OpenAI's publisher FAQ doesn't name a search provider. It mentions URLs obtained \"from a third-party search provider or by crawling other pages\", which points to at least two routes: someone else's search results, and OpenAI's own crawl.",[11,165,166,167,172],{},"The best public test I know of points at Bing. In February 2025 Seer Interactive took 100 questions, collected the pages ChatGPT search cited, and ",[34,168,171],{"href":169,"rel":170},"https:\u002F\u002Fwww.seerinteractive.com\u002Finsights\u002F87-percent-of-searchgpt-citations-match-bings-top-results",[38],"looked for each one in the search results"," for the same question. 87% turned up in Bing's top 20. 56% turned up in Google. Seer calls the result directional, and I haven't found a 2026 re-run.",[11,174,175,176,181],{},"Google may be in the mix as well. The Information reported in August 2025 that OpenAI was using Google results scraped by a third party for some current-events queries, according to ",[34,177,180],{"href":178,"rel":179},"https:\u002F\u002Fwww.seroundtable.com\u002Fchatgpt-serpapi-scrape-google-39989.html",[38],"Search Engine Roundtable",". I haven't found a confirmation from OpenAI.",[11,183,184,185,190],{},"In practice: make sure your pages show up in Bing. ",[34,186,189],{"href":187,"rel":188},"https:\u002F\u002Fwww.bing.com\u002Fwebmasters\u002Fabout",[38],"Bing Webmaster Tools"," is where you check.",[26,192,194],{"id":193},"what-chatgpt-cites-and-how-fast-that-changes","What ChatGPT cites, and how fast that changes",[11,196,197],{},"In Profound's study, Wikipedia took 5% of ChatGPT's citations and Reddit 3%. The more useful figure is the tail: \"Top 10 domains capture only 12% of all citations.\" The other 88% are spread across everyone else.",[11,199,200,201,206],{},"The mix also moves without warning. Semrush ",[34,202,205],{"href":203,"rel":204},"https:\u002F\u002Fwww.semrush.com\u002Fblog\u002Fmost-cited-domains-ai\u002F",[38],"tracked 230,000 prompts"," across ChatGPT search, Google AI Mode and Perplexity from July to October 2025, and watched Reddit's presence fall: \"ChatGPT cited Reddit in close to 60% of prompt responses in early August before collapsing to around 10% by mid-September.\" Semrush says \"The data doesn't tell us exactly why these shifts happened.\"",[11,208,209],{},"So don't build a plan on whichever site ChatGPT favours this quarter. Measure which sources it cites for your own questions, and measure again next month.",[26,211,213],{"id":212},"how-to-rank-in-chatgpt-and-get-cited","How to rank in ChatGPT and get cited",[215,216,218],"h3",{"id":217},"_1-let-oai-searchbot-in","1. Let OAI-SearchBot in",[11,220,221,222,226,227,231],{},"Check three places: your robots.txt, the bot rules on your CDN or firewall, and any server rule that blocks visitors by the name they announce, called the user agent. Shruwd's ",[34,223,225],{"href":224},"\u002Fscan","free crawler access check"," reads the first of those for you. ",[34,228,230],{"href":229},"\u002Fblog\u002Fllm-seo","LLM SEO: five levers with evidence"," has a PowerShell request that shows how your site answers OpenAI's search crawler.",[215,233,235],{"id":234},"_2-put-the-answer-in-the-raw-html-at-a-url-that-still-works","2. Put the answer in the raw HTML, at a URL that still works",[11,237,238,239,244],{},"Vercel and MERJ ",[34,240,243],{"href":241,"rel":242},"https:\u002F\u002Fvercel.com\u002Fblog\u002Fthe-rise-of-the-ai-crawler",[38],"analysed crawler traffic across Vercel's network"," in December 2024. OpenAI's crawlers didn't run JavaScript, so content that only appears after scripts run may not be there for them. The same study found that 34.82% of ChatGPT's fetches hit a 404 (page not found). Redirect pages you've moved, so an old link still lands somewhere.",[215,246,248],{"id":247},"_3-be-findable-in-bing-as-well-as-google","3. Be findable in Bing as well as Google",[11,250,251],{},"In Seer's test above, most cited pages also ranked in Bing's top 20 for the same question. So aim to rank for the specific question your buyer asks, in Bing as well as Google.",[215,253,255],{"id":254},"_4-get-mentioned-on-the-pages-chatgpt-already-cites","4. Get mentioned on the pages ChatGPT already cites",[11,257,258,259,264],{},"Ahrefs ",[34,260,263],{"href":261,"rel":262},"https:\u002F\u002Fahrefs.com\u002Fblog\u002Fai-brand-visibility-correlations\u002F",[38],"studied 75,000 brands"," across ChatGPT, AI Mode and AI Overviews in December 2025. Mentions on YouTube had the strongest correlation with AI visibility, at about 0.74, where 1 would be a perfect match, and web mentions came in between 0.66 and 0.71. Link metrics were \"very weak\". A correlation shows two things rising together, not that one causes the other, so treat these as a direction.",[11,266,267],{},"Ask ChatGPT your buyers' questions, several times each, and note which pages it cites. Then work on getting included there.",[215,269,271],{"id":270},"_5-answer-the-question-where-it-can-be-found","5. Answer the question where it can be found",[11,273,274,275,278],{},"Put the question in a heading and the answer in the first lines beneath it, with real numbers and names. The evidence for this one is thin, and ",[34,276,277],{"href":229},"the LLM SEO guide"," explains why.",[26,280,282],{"id":281},"how-to-tell-whether-its-working","How to tell whether it's working",[11,284,285],{},"Your server logs come first, because they show what OpenAI's crawlers received. This counts their visits in your web server's access log, the file that records every request, then lists any visit that got an error:",[287,288,293],"pre",{"className":289,"code":290,"language":291,"meta":292,"style":292},"language-powershell shiki shiki-themes github-light github-dark","$log = \".\\access.log\"\n\"OAI-SearchBot\", \"ChatGPT-User\", \"GPTBot\" | ForEach-Object {\n  \"{0,-14} {1}\" -f $_, @(Select-String -Path $log -Pattern $_ -SimpleMatch).Count\n}\nSelect-String -Path $log -Pattern 'OAI-SearchBot|ChatGPT-User|GPTBot' |\n  Where-Object { $_.Line -match '\" (4|5)\\d\\d ' } |\n  Select-Object -ExpandProperty Line\n","powershell","",[18,294,295,303,309,315,321,327,333],{"__ignoreMap":292},[296,297,300],"span",{"class":298,"line":299},"line",1,[296,301,302],{},"$log = \".\\access.log\"\n",[296,304,306],{"class":298,"line":305},2,[296,307,308],{},"\"OAI-SearchBot\", \"ChatGPT-User\", \"GPTBot\" | ForEach-Object {\n",[296,310,312],{"class":298,"line":311},3,[296,313,314],{},"  \"{0,-14} {1}\" -f $_, @(Select-String -Path $log -Pattern $_ -SimpleMatch).Count\n",[296,316,318],{"class":298,"line":317},4,[296,319,320],{},"}\n",[296,322,324],{"class":298,"line":323},5,[296,325,326],{},"Select-String -Path $log -Pattern 'OAI-SearchBot|ChatGPT-User|GPTBot' |\n",[296,328,330],{"class":298,"line":329},6,[296,331,332],{},"  Where-Object { $_.Line -match '\" (4|5)\\d\\d ' } |\n",[296,334,336],{"class":298,"line":335},7,[296,337,338],{},"  Select-Object -ExpandProperty Line\n",[11,340,341,343,344,346],{},[18,342,100],{}," visits mean your pages are being collected for search. A ",[18,345,20],{}," visit usually means a person's request sent ChatGPT to that page just then, which is the closest thing in your logs to a citation. Lines in the second list are pages that returned an error to one of the crawlers.",[11,348,349],{},"The second part assumes the common log layout, where the status code follows the quoted request. If your log is laid out differently, an empty list proves nothing.",[11,351,352,353,357],{},"A user agent is easy to fake. OpenAI publishes the IP ranges its crawlers use, so a visit can be checked. Shruwd, the tool we are building, checks every hit against those ranges, and its docs list ",[34,354,356],{"href":355},"\u002Fdocs\u002Fcrawler-logs\u002Fcrawlers-tracked","the crawlers it tracks",".",[11,359,360,361,364],{},"Traffic is the second signal. OpenAI's publisher FAQ says ChatGPT adds ",[18,362,363],{},"utm_source=chatgpt.com"," to the links it sends, and since May 2026 Google Analytics has had an AI Assistant channel for these visits.",[11,366,367,368,372],{},"The third is asking. Keep ",[34,369,371],{"href":370},"\u002Fdocs\u002Ftracking\u002Fprompts","a fixed list of questions",", put them to ChatGPT repeatedly, and track how often you're named, with a range on the number.",[26,374,376],{"id":375},"what-hasnt-been-shown-to-move-it","What hasn't been shown to move it",[11,378,379,383,384,389],{},[380,381,382],"strong",{},"Ads."," ChatGPT shows ads now. OpenAI's ",[34,385,388],{"href":386,"rel":387},"https:\u002F\u002Fopenai.com\u002Findex\u002Fexpanding-access-to-ai-with-chatgpt-ads\u002F",[38],"announcement"," says they are \"always clearly labeled and separate from ChatGPT's answers\", and that \"Advertising does not influence the answers ChatGPT provides.\"",[11,391,392,395],{},[380,393,394],{},"Schema markup."," Ahrefs tracked 1,885 pages that added schema. Their ChatGPT citations rose by about 2%, which was statistically indistinguishable from zero.",[11,397,398,401,402,405],{},[380,399,400],{},"llms.txt."," OpenAI's crawler documentation doesn't mention it. ",[34,403,404],{"href":229},"The LLM SEO guide"," covers both of these in full.",[26,407,409],{"id":408},"frequently-asked-questions","Frequently asked questions",[215,411,413],{"id":412},"can-you-rank-in-chatgpt","Can you rank in ChatGPT?",[11,415,416],{},"Not in the sense of holding a position. ChatGPT writes each answer fresh, and the same question can produce a different set of sources and brands each time. What you can raise is the share of answers that cite or name you, by being reachable, findable in search and readable without JavaScript. Mentions on the pages ChatGPT already cites go with being named more often.",[215,418,420],{"id":419},"how-do-you-get-cited-by-chatgpt","How do you get cited by ChatGPT?",[11,422,423,424,426],{},"Let ",[18,425,100],{}," crawl your site, and check that no firewall rule turns it away. Put your answers in the page's HTML so they don't depend on JavaScript. Rank for the specific questions your buyers ask, in Bing as well as Google. Then get your brand mentioned on the pages ChatGPT already cites for those questions.",[215,428,430],{"id":429},"does-chatgpt-use-bing-or-google","Does ChatGPT use Bing or Google?",[11,432,433],{},"OpenAI's publisher FAQ doesn't name one. It refers to \"a third-party search provider\" and to its own crawling. In Seer Interactive's February 2025 test, 87% of ChatGPT search citations matched a page in Bing's top 20, against 56% for Google. A press report says OpenAI also uses scraped Google results. I haven't found a confirmation.",[215,435,437],{"id":436},"how-long-does-it-take-to-show-up-in-chatgpt","How long does it take to show up in ChatGPT?",[11,439,440,441,443],{},"The one number I found from OpenAI is about 24 hours for its systems to pick up a robots.txt change. I found nothing from it on how fast a new page is collected. Your logs will tell you when ",[18,442,100],{}," first fetches it.",[215,445,447],{"id":446},"what-is-seo-for-chatgpt-called","What is SEO for ChatGPT called?",[11,449,450,451,455],{},"ChatGPT SEO, LLM SEO, GEO and AEO all describe it. ",[34,452,454],{"href":453},"\u002Fblog\u002Fgeo-vs-seo","GEO vs SEO"," explains the terms, and the work is mostly the same under each name.",[457,458,459],"style",{},"html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":292,"searchDepth":311,"depth":311,"links":461},[462,463,464,465,466,473,474,475],{"id":28,"depth":305,"text":29},{"id":60,"depth":305,"text":61},{"id":159,"depth":305,"text":160},{"id":193,"depth":305,"text":194},{"id":212,"depth":305,"text":213,"children":467},[468,469,470,471,472],{"id":217,"depth":311,"text":218},{"id":234,"depth":311,"text":235},{"id":247,"depth":311,"text":248},{"id":254,"depth":311,"text":255},{"id":270,"depth":311,"text":271},{"id":281,"depth":305,"text":282},{"id":375,"depth":305,"text":376},{"id":408,"depth":305,"text":409,"children":476},[477,478,479,480,481],{"id":412,"depth":311,"text":413},{"id":419,"depth":311,"text":420},{"id":429,"depth":311,"text":430},{"id":436,"depth":311,"text":437},{"id":446,"depth":311,"text":447},"Guides","2026-09-19","Which OpenAI crawler to let in, where ChatGPT searches, what it can read on your page, and how to check your server logs for its visits.",false,"md","\u002Fimgs\u002Fhow-to-rank-in-chatgpt.jpg",{},true,"\u002Fblog\u002Fhow-to-rank-in-chatgpt",8,{"title":5,"description":484},"blog\u002Fhow-to-rank-in-chatgpt",[495,100,496,497],"ChatGPT SEO","GEO","AI search",null,"e4N9jBo6DmHKJY_QoXkYcD8432JW0davW7qg6E_zWZk",[501,512,524,534,542,544],{"path":55,"title":502,"description":503,"date":504,"category":505,"image":506,"tags":507,"draft":485,"readingTime":511},"AI visibility: how to measure it, and why one number lies","AI visibility is a rate across repeated runs, so one score from one check tells you little. What to count, and how to tell a real change from noise.","2026-09-18","Measurement","\u002Fimgs\u002Fai-visibility.jpg",[508,509,510,496],"AI visibility","measurement","share of voice",9,{"path":513,"title":514,"description":515,"date":516,"category":517,"image":518,"tags":519,"draft":485,"readingTime":523},"\u002Fblog\u002Fbest-ai-visibility-tools","Best AI visibility tools in 2026, and the catch with each","Nine AI visibility tools compared on price, AI engines, how often they run and how they collect answers, from each vendor's own pages, checked September 2026.","2026-09-21","Tools","\u002Fimgs\u002Fbest-ai-visibility-tools.jpg",[508,520,521,522],"AEO tools","GEO tools","tools",11,{"path":525,"title":526,"description":527,"date":516,"category":528,"image":529,"tags":530,"draft":485,"readingTime":511},"\u002Fblog\u002Fcloudflare-block-ai-bots","Cloudflare Block AI Bots: what changed, and what to check","Cloudflare replaced its Block AI Bots toggle on 15 September 2026. What the new settings do to Googlebot and ChatGPT, and how to see what it turns away.","Diagnostics","\u002Fimgs\u002Fcloudflare-block-ai-bots.jpg",[531,532,533,100],"Cloudflare","AI crawlers","Bot Fight Mode",{"path":453,"title":535,"description":536,"date":504,"category":482,"image":537,"tags":538,"draft":485,"readingTime":541},"GEO vs SEO: what changed, what didn't, what to do first","How GEO differs from SEO, which GEO advice has evidence behind it, and two checks you can run on your own site today.","\u002Fimgs\u002Fgeo-vs-seo.jpg",[496,539,540,497],"SEO","AEO",13,{"path":490,"title":5,"description":484,"date":483,"category":482,"image":487,"tags":543,"draft":485,"readingTime":491},[495,100,496,497],{"path":229,"title":545,"description":546,"date":504,"category":482,"image":547,"tags":548,"draft":485,"readingTime":523},"LLM SEO: five levers with evidence, and two without","Five LLM SEO levers with evidence behind them, graded by strength, and the reason schema markup and llms.txt aren't among them.","\u002Fimgs\u002Fllm-seo.jpg",[549,496,497,550,551],"LLM SEO","llms.txt","schema",1790003440526]