# Agent SEO HQ: full site content > Every public page and article on agentseohq.com as Markdown, generated from the same source as the website on 11 November 2026. Short version: https://agentseohq.com/llms.txt # Agent SEO HQ: get recommended when people ask AI URL: https://agentseohq.com/ Agent SEO for businesses that want ChatGPT, Claude, Gemini, Perplexity and Copilot to find and recommend them. Free check, then a fixed quote. Agent SEO, done by someone who builds websites ## Get recommended when people ask AI. People now ask ChatGPT, Claude, Gemini, Perplexity and Copilot who to hire and what to buy. Agent SEO makes your website easy for those assistants to find, read and recommend. I check it, fix it and keep watch. [Get a free Agent SEO check](https://agentseohq.com/contact) [Try the instant checker](https://agentseohq.com/check) No price list to squint at. The check is free, and paid work gets a fixed quote first. Why I do this ### It started with one email I make a small Mac app called Cartage. In October 2026 a stranger wrote in because the checkout would not take their payment. I asked how they had found the app. They had not clicked an advert or scrolled a page of search results. They asked an assistant a question, and the assistant sent them to my website. They bought the app two days later, once I had fixed the checkout. So I worked out why that site got picked, did the same work on ten more of my own sites, and wrote down what I found. That work is now a service you can hire. The [case study](https://agentseohq.com/case-study) has the details, including the bit where I broke a home page for fifteen seconds. **From:** a Mac user **To:** support@flavrapps.com · **Date:** 6 October 2026 I needed an app to track file copying progress on my Mac, so I asked ChatGPT to help me find one. It recommended Cartedge and pointed me to your website — that’s how I found you! A real customer email. Name and email address removed. (Their spelling.) What Agent SEO is ### Three jobs, in this order Search engine optimisation gets you into a list of links. Agent SEO gets you into the answer. An assistant has to do three things before it names you, and most websites trip it up at the first. - 1 #### Found The assistant's crawlers are allowed in, and your pages are in the search indexes it leans on. Plenty of sites block AI crawlers without knowing, often through a setting in their hosting or security service. - 2 #### Understood Your pages say plainly what you sell, who for, where and for how much, in text a machine can read. Structured data backs it up, and the facts match everywhere else you appear. - 3 #### Recommended You answer the questions your customers actually ask, and other places on the web say the same good things about you. That part is earned, and it is where the work pays off. Services ### Check it, fix it, keep watch Take one or all three. Everything starts with the free check, so you know what you are buying before you buy it. - Step one #### [Agent SEO audit](https://agentseohq.com/services/agent-seo-audit) A full review of what AI assistants can reach, read and repeat about your business, plus a baseline of what they say today. You get a plain-English list in priority order. What the audit covers - We fix it #### [Implementation](https://agentseohq.com/services/implementation) I make the changes: crawler access, structured data, llms.txt, content built to be quoted, Bing and IndexNow, and consistent facts across the web. You approve everything first. What gets fixed - Ongoing #### [Monitoring](https://agentseohq.com/services/monitoring) Each month I put your customers' questions to ChatGPT, Claude, Gemini, Perplexity and Copilot and report where you appear, who appears instead and what to do about it. What you get each month Tested on my own sites first ### I ran this on 11 of my own websites before selling it On 10 November 2026 every live product site I run got the full treatment in one pass. Before touching anything, I put the questions a customer would ask to OpenAI's search-enabled model, the closest scriptable stand-in for ChatGPT's search, and recorded the answers. - **11**websites checked and fixed in one pass, from a Mac app to a football game. - **39**customer questions asked before any changes, none of them naming my products. - **3 of 4**answers named Cartage, the site that was already well prepared. Eight of the eleven sites were not named once. [Read the case study](https://agentseohq.com/case-study) How it works ### Four steps, no surprises 1. #### Free check You send your address and say who your customers are. I check the site by hand and reply with what I found. 2. #### Fixed quote If there is work worth doing, you get one price for it in writing. No hourly meter. 3. #### We fix it I make the changes on your site, or hand your developer exact instructions. Nothing goes live without your say-so. 4. #### We keep watch Optional monitoring shows whether the assistants have noticed, and what to do next. [How it works in detail](https://agentseohq.com/how-it-works) The honest bit ### What I will not promise This is a young field with a lot of confident nonsense in it. Here is where I stand. - **No guaranteed recommendations.** Nobody controls what an assistant says, and its answers change from one day to the next. I improve what is in your control and measure the result. - **No fake reviews, paid mentions or filler articles.** They are the quickest way to be trusted less by people and machines alike. - **No pretending llms.txt is magic.** Google says its search does not use it, and a 2026 study found most of these files are never fetched. I add one because it takes minutes, then spend the time on what moves answers. [The evidence is here](https://agentseohq.com/articles/what-is-llms-txt). What you do get ### What I will do - Check that every major AI crawler can reach your pages, and find whatever is blocking them. - Make your business facts readable by machines and the same everywhere. - Show you what the assistants say about you today, with the answers saved. - Fix things in priority order and tell you which ones I think are not worth paying for. - Write in plain English. You should not need a glossary to read an invoice. Articles ### Read up before you spend a penny Everything I know about Agent SEO is written down here, with sources. If you would rather do it yourself, these will get you most of the way. - The basics #### [What is Agent SEO? A plain-English guide](https://agentseohq.com/articles/what-is-agent-seo) Agent SEO means making your website easy for AI assistants to find, understand and recommend. What it involves, how it differs from SEO, and where to start. 11 November 2026 · 7 min read - The basics #### [GEO vs AEO vs AI SEO: what the terms mean](https://agentseohq.com/articles/geo-vs-aeo-vs-ai-seo) Generative engine optimisation (GEO), answer engine optimisation (AEO), AI SEO and Agent SEO explained: where each term came from and whether they differ. 11 November 2026 · 5 min read - How-to #### [How to get recommended by ChatGPT](https://agentseohq.com/articles/how-to-get-recommended-by-chatgpt) How to get your business recommended by ChatGPT: what OpenAI says it needs, seven things that help, what does not work, and how to check your progress. 11 November 2026 · 7 min read - Technical #### [What is llms.txt, and does it actually work?](https://agentseohq.com/articles/what-is-llms-txt) What llms.txt is, how to write one, and what the evidence says about whether ChatGPT, Claude or Google read it. An honest guide with a working example. 11 November 2026 · 7 min read - Technical #### [AI crawlers and robots.txt: GPTBot, ClaudeBot, PerplexityBot and the rest](https://agentseohq.com/articles/ai-crawlers-robots-txt) A reference to the AI crawlers that visit your site (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended) and the robots.txt rules for each. 11 November 2026 · 7 min read - Technical #### [Schema markup for AI search: what to add and what to skip](https://agentseohq.com/articles/schema-markup-for-ai-search) Which schema markup helps AI assistants understand your business, with copy-and-paste JSON-LD for organisations, products and FAQs, and what to skip. 11 November 2026 · 7 min read [All 12 articles](https://agentseohq.com/articles) Quick answers ### Questions people ask first #### What is Agent SEO? Agent SEO is the work of making a website easy for AI assistants and agents to find, understand and recommend. It covers crawler access, structured data, clear content, consistent business facts and third-party mentions. You will also see it called AI SEO, generative engine optimisation (GEO) or answer engine optimisation (AEO). #### How much does it cost? The first check is free. There are no public prices: after the check you get a fixed quote in writing for any work worth doing, and you can say no. #### Can you guarantee ChatGPT will recommend my business? No, and nobody honestly can. Assistants choose their own sources and their answers vary. What I can do is remove the things that stop them finding and understanding you, improve what they read, and measure whether you appear more often. #### Is this different from normal SEO? It builds on it. Assistants lean on search indexes, so good SEO is the foundation. Agent SEO adds the parts search rankings do not cover: AI crawler access, machine-readable facts, content written to be quoted, and checking what assistants actually say. [More questions and answers](https://agentseohq.com/faq) Who you deal with ### Hello, I'm Reece I'm a developer in Milton Keynes. I build apps and websites with a team of AI agents, which is a useful way to learn what agents want from a website. When you get in touch, you get me, not a sales team. [More about me and the company](https://agentseohq.com/about) ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Agent SEO services: audit, fixes and monitoring URL: https://agentseohq.com/services Three Agent SEO services: an audit of how AI assistants see your site, implementation of the fixes, and monthly monitoring of what they say about you. Services ## Agent SEO services Three services that follow on from each other: find out how AI assistants see your business, fix what gets in their way, then keep an eye on what they say. Each one starts with a free check and a fixed quote. ### The three services - Step one #### [Agent SEO audit](https://agentseohq.com/services/agent-seo-audit) What AI assistants can reach, read and repeat about your business, checked by hand. Includes a recorded baseline of what ChatGPT, Claude, Gemini, Perplexity and Copilot say today. See what the audit covers - We fix it #### [Implementation](https://agentseohq.com/services/implementation) The changes themselves: robots rules for AI crawlers, JSON-LD, llms.txt, content structured for answers, Bing and IndexNow, consistent facts, reviews and citations. See what gets fixed - Ongoing #### [Monitoring](https://agentseohq.com/services/monitoring) Your customers' questions, asked to the main assistants on a schedule. You get a report showing where you appear, who appears instead and what changed. See what you get each month ### Who this is for Businesses whose customers ask before they buy. If someone might type "who is a good ... near me" or "what is the best ... for ..." into an assistant, and you would like to be in the reply, this is for you. - **Local and service businesses**: trades, clinics, accountants, agencies, studios. - **Online shops** whose products get compared before purchase. - **App and software makers.** This is my home turf: the service began with my own apps. - **Anyone with a good website that assistants ignore.** ### Who it is not for If your website is a single page with no real content, the honest advice is to build the content first. I will tell you that in the free check rather than sell you an audit of an empty room. ### How pricing works There is no price list on this site, on purpose. A five-page site for a local firm and a two-thousand-product shop are different jobs, and a list would be wrong for both. 1. You ask for the [free Agent SEO check](https://agentseohq.com/contact). 2. I check your site by hand and reply with what I found. 3. If paid work would help, you get a fixed quote in writing. 4. You say yes, no, or "just this part". The quote is the price. If the job turns out bigger than I thought, that is my problem. ### What every service includes - One person to talk to: me. - Plain-English reports, with the evidence attached. - An honest view on what is not worth doing. - No lock-in. Everything I add to your site is standard, documented and yours. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Agent SEO audit: how AI assistants see your website URL: https://agentseohq.com/services/agent-seo-audit An Agent SEO audit checks whether ChatGPT, Claude, Gemini, Perplexity and Copilot can reach, read and recommend your website, and records what they say today. Service 1 of 3 ## Agent SEO audit A hands-on review of everything that decides whether an AI assistant can find, understand and recommend your business. You get a prioritised list in plain English and a recorded baseline of what the assistants say about you today. [Start with a free check](https://agentseohq.com/contact) In one paragraph The audit answers three questions. Can AI crawlers and assistants get to your pages? When they do, can they tell what you sell, to whom, where and for how much? And when someone asks about your kind of business, do you appear in the answer? Each finding comes with the evidence and a suggested fix. ### What the audit checks #### 1. Access: can they get in? - Your `robots.txt` rules for each AI crawler: OAI-SearchBot, ChatGPT-User, GPTBot, ClaudeBot, Claude-User, Claude-SearchBot, PerplexityBot, Google-Extended, Applebot, Bingbot and others. - Blocks you did not set yourself. Hosting companies, security plugins and services such as Cloudflare can block or challenge AI crawlers, and some rewrite robots.txt for you. I found this on one of my own domains. - What your server really answers when those crawlers ask for a page. - Whether your content needs JavaScript to appear. Many AI crawlers read only the HTML they are first sent. #### 2. Index: are you in the places they search? - Whether your pages are indexed by Bing and Google, since assistants lean on search indexes when they look things up. - Your sitemap, canonical addresses, redirects and duplicate versions of the site. - Whether you tell search engines about changes (IndexNow) or wait to be noticed. #### 3. Understanding: can they tell what you are? - Structured data (JSON-LD) for your organisation, your products or services, FAQs and breadcrumbs: present, valid and matching the visible page. - Whether the basic facts are stated in plain text: what you sell, who for, where, prices, opening hours, how to buy. - Facts hidden where machines struggle: in images, PDFs, tabs and pop-ups. - Your `llms.txt`, if you have one, and whether it is a real file or your home page in disguise. #### 4. Consistency: does the web agree about you? - Name, address, phone, prices and descriptions compared across your site, Google Business Profile, Bing Places, directories, app stores and social profiles. - Old names, old addresses and out-of-date prices. Assistants repeat whatever they find, including the stale version. #### 5. Content and reputation: is there anything worth recommending? - Whether your pages answer the questions customers ask, in the words they use. - Comparison and "alternatives" content. In tests on my own sites, an honest comparison page was one of the pages cited. - The reviews and third-party sources assistants cite for your type of business, and whether you are on them. #### 6. Baseline: what do they say today? - I write the questions a customer would ask, without naming you, and put them to ChatGPT, Claude, Gemini, Perplexity and Copilot. - Each answer is saved: whether you were named, who was named instead, and which websites were cited. - Answers vary from run to run, so the baseline is a sample, not a league table. It is still the only honest way to know where you start. ### What you receive - A written report, in priority order, with the evidence for each finding. - The saved baseline answers, so later changes can be measured against them. - A fixed quote for fixing the things worth fixing, split so you can pick. - A call or an email thread to go through it, whichever you prefer. ### What it does not include The audit does not change your website. That is [implementation](https://agentseohq.com/services/implementation), which is quoted separately so you can take the report to your own developer if you would rather. It also cannot see inside an assistant: nobody outside those companies knows exactly how an answer is put together, and the report says where I am inferring. ### Free check or full audit? | What you get | Free check | Full audit | Cost | Free | Fixed quote | Pages looked at | Home page and a sample | The whole site | Crawler access | Main AI crawlers | All of them, including firewall and host behaviour | Questions asked to assistants | A few | A full set, across five assistants | Consistency across the web | Not included | Included | Result | A short email with the top fixes | Full report, saved baseline and a quote Want an instant machine-only version first? The [free checker](https://agentseohq.com/check) tests any address in a few seconds. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Agent SEO implementation: we fix it for you URL: https://agentseohq.com/services/implementation We make the Agent SEO fixes for you: AI crawler rules, JSON-LD, llms.txt, content structured for answer engines, Bing and IndexNow, and consistent facts. Service 2 of 3 ## Implementation: we fix it An audit is a list. This is the work. I make the changes on your website, or give your developer exact instructions, so AI assistants can reach your pages, read your facts and quote your answers. [Start with a free check](https://agentseohq.com/contact) In one paragraph Implementation covers eight things: AI crawler access, structured data, llms.txt, content built to be quoted, Bing and IndexNow, consistent facts across the web, reviews and citations, and a before-and-after check. You approve every change before it goes live, and everything added is standard and documented. ### What gets fixed #### Robots rules for AI crawlers A `robots.txt` that names the crawlers that matter and lets them in, while keeping private areas private. I also find and remove blocks set by your host, security plugin or content delivery network. If you would rather keep your content out of AI training but stay visible in AI search, that is possible for most of them, and I will set it up that way. #### JSON-LD for entities, products and FAQs Structured data that states who you are (Organization or LocalBusiness), what you sell (Product, Service or SoftwareApplication with real prices), your questions and answers (FAQPage) and where each page sits (BreadcrumbList). It is generated from the page content, so it cannot drift from what visitors see. #### llms.txt and llms-full.txt A short, accurate guide to your business for AI agents, plus a full-text version where it helps. I will be straight with you: the evidence that assistants read these files today is thin. They take minutes, not days, so you get them without paying for hype. [Why I still add one](https://agentseohq.com/articles/what-is-llms-txt). #### Content structured for answer engines Assistants lift short, self-contained passages. I restructure key pages so the answer comes first, the headings say what each section answers, and comparisons sit in tables. Where a question your customers ask has no page, I draft one for you to approve. Your facts, your voice, no filler. #### Bing and IndexNow Bing matters more than its market share suggests, because ChatGPT and Copilot draw on it. I verify your site in Bing Webmaster Tools, submit your sitemap and set up IndexNow, so changes are announced the moment you publish. #### Entity consistency across the web Your name, address, phone number, prices and description made the same everywhere: website, Google Business Profile, Bing Places, directories, app stores and social profiles. When sources disagree, an assistant has to guess, and it does not always guess in your favour. #### Reviews and citations I find the review sites and third-party pages assistants cite for your kind of business, and help you get listed, corrected or reviewed there by real customers. I will not buy reviews or mentions, and I would advise you not to either. #### Before-and-after check Every change is verified once live: crawlers get the right answer, structured data validates, pages are submitted for indexing. Assistants need time to re-read a site, so the effect on answers shows up in the following weeks, which is what [monitoring](https://agentseohq.com/services/monitoring) is for. ### How the work is done 1. **Agree the list.** You pick from the audit or free check. The quote is fixed. 2. **Access.** You give me the least access that does the job, or none: I can hand over changes as files and instructions for your developer. 3. **Review.** You see every change before it goes live. Anything visible to customers gets your sign-off on the wording. 4. **Go live and verify.** I check each change on the live site and send you the evidence. 5. **Hand over.** You get notes on what was added and how to keep it up to date. ### Which platforms? Most of them: WordPress, Shopify, Webflow, custom sites and static sites are all workable. Some hosted site builders limit what you can change, such as robots.txt or files at the root of the domain. If your platform will not allow something, I will say so before quoting, not after. Tested on my own sites I applied this list to 11 of my own websites in one pass. Along the way I found a download button that had never worked and two app sites with no link to the App Store. Those mattered more than any of the clever bits. [Read what happened](https://agentseohq.com/case-study). ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # AI visibility monitoring: what assistants say about you URL: https://agentseohq.com/services/monitoring Monthly AI visibility monitoring. We ask ChatGPT, Claude, Gemini, Perplexity and Copilot your customers' questions and report where you appear. Service 3 of 3 ## Monitoring: what the assistants say about you Search rankings can be looked up. AI answers cannot: there is no page two to check. So I regularly ask ChatGPT, Claude, Gemini, Perplexity and Copilot the questions your customers ask, and report where you appear. [Start with a free check](https://agentseohq.com/contact) In one paragraph We agree a set of questions your customers would ask. On a schedule, I put them to five assistants, save every answer, and report how often you are named, who is named instead, which websites get cited and what changed since last time. Each report ends with a short list of things to do. ### What gets measured - **Mentions.** How often each assistant names your business for each question. - **Citations.** Whether your website is one of the sources, and which page. - **Accuracy.** What the assistant says about you, and whether it is true. Wrong prices and old addresses get flagged. - **Competitors.** Who is recommended instead, and what their cited pages have that yours do not. - **Traffic.** Visits arriving from assistants, where your analytics can show them. ChatGPT, for example, adds `utm_source=chatgpt.com` to the links it sends. - **Access.** A check that your crawler rules and structured data are still intact. Plugin and hosting updates break these more often than you would think. ### How the questions are asked Each question is written the way a customer would type it, without your name in it. I ask through each assistant's official interface for developers where one exists, and by hand where it does not. Because answers vary, each question is asked more than once and the report gives a rate, such as "named in 4 of 6 runs", not a single yes or no. A limit worth knowing What an assistant tells me may differ from what it tells your customer. Answers depend on location, wording, the model and whether it searched the web at all. Monitoring shows the trend across many samples. It is not a guarantee of what any one person will see. ### What you receive - A report on an agreed schedule, monthly by default, written in plain English. - The saved answers behind every number, so you can read them yourself. - Where they are available for your site, figures from Bing Webmaster Tools' AI Performance report and Google Search Console's generative AI report. - A short list of actions, in priority order, with a quote for any you want me to do. ### Why bother after the fixes are done? Three reasons. Assistants take weeks to re-read a site, so the first reports show whether the fixes landed. Competitors move, and a rival who publishes a good comparison page can take your place. And the assistants themselves change: new models, new crawlers and new rules arrive every few months, and something that worked can quietly stop. ### Common questions #### Which assistants do you monitor? ChatGPT, Claude, Gemini, Perplexity and Microsoft Copilot by default. If your customers use something else, such as a shopping assistant, it can be added where it is possible to query it. #### How many questions do you track? It depends on your business. A local firm might need ten; a shop with several product ranges might need fifty. The number is agreed before the quote, and the price is fixed for that set. #### Can I cancel? Yes. Monitoring runs month to month unless we agree something else in writing, and you keep every report and saved answer. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # How Agent SEO HQ works: free check, then a fixed quote URL: https://agentseohq.com/how-it-works How working with Agent SEO HQ goes, step by step: a free check by hand, a fixed quote, the fixes, and optional monitoring. No retainers by default. How it works ## Free check first. Fixed quote second. Then the work. You should know what is wrong and what it costs to fix before you pay anyone anything. So that is the order things happen in. ### The steps 1. #### You ask for a free check Fill in the [contact form](https://agentseohq.com/contact): your website, what you sell and who your customers are. That last part matters, because it decides which questions I ask the assistants. It takes about two minutes. 2. #### I check your site by hand I look at whether AI crawlers can reach your pages, whether your facts are machine-readable, and what a few assistants say when asked your customers' questions. You get an email with what I found, usually within two working days. 3. #### You get a fixed quote If paid work would help, the email includes a quote: one price for a full audit, or for specific fixes if the problems are already obvious. If I do not think you need me, I will say that instead. 4. #### We agree what to do Take all of it, some of it, or none. If you want to give the list to your own developer, go ahead. Once you say yes, I confirm the scope, the price and the dates in writing. 5. #### I do the work Changes are made on your site with the least access that does the job, or handed to your developer as files and instructions. You see everything before it goes live. 6. #### I verify and hand over Each change is checked on the live site. You get the evidence and notes on how to keep things up to date. Then the waiting starts: assistants need a few weeks to re-read a website. 7. #### Optional: I keep watch With [monitoring](https://agentseohq.com/services/monitoring), I ask the assistants your customers' questions on a schedule and report where you appear. It is the only way to see whether the work moved anything. 8. #### You stay in charge No lock-in and nothing proprietary. What I add to your site is standard, documented and yours to keep, whether or not we ever work together again. ### What I need from you - Your website address. - A sentence or two on what you sell and who buys it. - For paid work: access to make changes, or a developer who can apply them. - Sign-off on any wording your customers will see. ### How long it takes The free check usually comes back within two working days. A full audit and the fixes depend on the size of the site, and the dates are in the quote. The effect on what assistants say takes longer, because they have to crawl your site again. On my own sites I allowed two to three weeks before re-measuring. ### Who does the work Me, with a team of AI agents doing the legwork: crawling, checking, drafting and re-checking. I review what they produce and I am the one who replies to you. It is how I build my own apps, and it is why a small company can check a site this thoroughly at a fixed price. ### What happens to your details What you send is used to do the check and reply to you. It is not sold, shared or added to a mailing list. The [privacy policy](https://agentseohq.com/privacy) has the specifics, including how long enquiries are kept. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Case study: a customer found Cartage by asking ChatGPT URL: https://agentseohq.com/case-study A customer found our Mac app Cartage by asking ChatGPT. Here is why it was recommended, and what happened when we did the same work on 11 of our sites. Case study ## A customer found Cartage by asking ChatGPT. So I did the same work on 11 sites. This is the story behind Agent SEO HQ, with real numbers, a real email and the things that went wrong. The customer is anonymous. Everything else is as it happened. By [Reece Chown](https://agentseohq.com/about) The short version - In October 2026 a customer told me ChatGPT had recommended my Mac app, Cartage, and sent them to its website. They bought it. - Cartage's site was the best prepared of my sites for AI assistants. In a test before any new work, OpenAI's search-enabled model named it in 3 of 4 answers. - Eight of my other ten sites were not named in a single answer. - On 10 November 2026 I applied the same kit to all 11 sites. The "after" test is still to come, because assistants need weeks to re-read a site. I will publish it here either way. ### The email Cartage is a small Mac app that shows the progress of Finder file copies in the menu bar. On 29 September 2026 someone wrote to support because they could not complete a payment. I replied, and asked how they had come across the app. Their answer, on 6 October, is on this page. They had asked ChatGPT for an app that tracks file copying on a Mac. It recommended Cartage, spelling aside, and pointed them to the website. Two days later the checkout was fixed and they bought it. One customer is an anecdote, not a study. But it was the first time I had seen the whole chain work on something I made: question, recommendation, website, sale. It was worth finding out why. **From:** a Mac user **To:** support@flavrapps.com · **Date:** 6 October 2026 I needed an app to track file copying progress on my Mac, so I asked ChatGPT to help me find one. It recommended Cartedge and pointed me to your website — that’s how I found you! I was interested in purchasing it, but I couldn’t complete the payment. Is purchasing available yet…? Name and email address removed. (Their spelling.) ### Why Cartage was easy to recommend When I went back and looked, Cartage's website already had most of what an assistant needs. Not through foresight: I had simply been more thorough with that site than with the others. - A `robots.txt` that let AI crawlers in. - An `llms.txt` describing the app in plain language. - Structured data saying what the software is, who makes it and its questions and answers. - Seven guides answering questions people actually have about copying files on a Mac. - An honest "alternatives" page comparing it with other options. That last one surprised me. When the model did recommend Cartage in my tests, the alternatives page was one of the pages it cited. A fair comparison, including when the other option is the better choice, turned out to be the sort of page an assistant is happy to quote. ### The test: 39 questions before changing anything To see where everything stood, I wrote three or four questions per site, worded as a customer would type them and never naming the product. For example: "Is there a Mac app that shows Finder file copy progress in the menu bar?" Each question went once to OpenAI's gpt-5-mini model with its web search tool switched on, which is the closest thing to ChatGPT's search that can be run from a script. I recorded whether the answer or its cited sources named my product or domain. | Site | What it is | Named in the answer | Cartage | Mac app for file copy progress | **3 of 4** | Simple Stack | PHP starter kit | 2 of 4 | ViewPortal | iPhone and iPad app for testing websites | 1 of 4 | Reroute | Mac app for per-app proxies | 0 of 4 | Iconic Icons | App icon generator | 0 of 4 | Ranked Padel | Padel scores and rankings app | 0 of 4 | FourCast | Weather app | 0 of 3 | Candle & Dagger | Murder mystery game | 0 of 3 | Back Seat Video Remote | Remote control for a child's iPad | 0 of 3 | All Eras XI | Football draft game | 0 of 3 | One unreleased app | Structure only, nothing to sell yet | 0 of 3 Some caveats, because a table looks more scientific than it is. Each question was asked once, so a single yes or no is a sample, not a measurement. It covers one assistant. And the sites differ in age and competition, so preparation is not the only reason for the gap. Still, the pattern was hard to ignore. The three sites with real content and machine-readable facts were the three that got named. Where my sites lost, it mostly cited small competitors' own websites, far more often than the App Store. One weather question was answered with programming code and no sources at all: the assistant did not know any such app existed. ### What I did to all 11 sites On 10 November 2026, in one pass, every site got: - an `/llms.txt`, and an `/llms-full.txt` where there was enough content to justify one; - a `robots.txt` that names the AI crawlers and Bing's crawler and lets them in; - valid structured data: the product, its price where one is public, the company, and the FAQ; - a canonical address on every page; - an IndexNow key, with every page submitted so search engines knew about the changes at once; - new content where a question had no answer, such as a comparison guide for Reroute and a guide for ViewPortal. Afterwards, an automated check confirmed that on all 11 sites the crawler rules, llms.txt, sitemap, structured data and IndexNow key were live and valid, and that requests identifying as GPTBot and ClaudeBot got a normal response. Each home page scored 100 for SEO in Google's Lighthouse test. ### What I found that mattered more than the clever bits This is the part I would want to know if I were hiring someone. Going through each site as an assistant would turned up plain, embarrassing faults that no amount of structured data would have fixed: - **A download button that had never worked.** On Candle & Dagger's site, a coding slip meant the App Store button pointed nowhere. Visitors saw "Download on the App Store" and could not. - **Two app sites with no link to the App Store at all.** An assistant that found FourCast or Back Seat Video Remote had nothing to send people to. - **Three pages describing the wrong product.** Iconic Icons' pricing, gallery and developer pages still carried a description left over from a template. - **Three sites answering "yes" to a file that did not exist.** Asked for `/llms.txt`, they returned their home page with a success code. A machine cannot tell that from a real file without reading it. - **A security service quietly rewriting robots.txt.** On two domains, Cloudflare was adding rules that told GPTBot, ClaudeBot and others to stay out. I had not asked it to. An assistant can only recommend what it can reach and make sense of. Most of the value was in finding the things that were simply broken. ### What went wrong Two things, both mine. One site's home page returned an error for about fifteen seconds after a deploy, because a file was uploaded with the wrong permissions. I put it right at once and changed the deploy script so it cannot happen again. And one deploy failed before uploading anything, thanks to a typing mistake, and had to be re-run. Nobody noticed either, but you should know how I work when things go sideways: I check, I fix, I write it down. ### What it cost to measure The 39 questions cost about $1.29 in fees to OpenAI, by a deliberately cautious estimate. I mention it because "AI visibility tracking" is often sold as something exotic. The asking is cheap. The value is in choosing the right questions and knowing what to do with the answers. ### The result, so far Here is the honest position on 11 November 2026. The work is live and verified on all 11 sites. Whether ChatGPT now names more of them, I do not know yet. Assistants and search engines have to crawl the sites again, and I allowed two to three weeks for that before re-running the same 39 questions. When the re-run is done, the numbers will go in the table above, good or bad. If the answer is "it moved three sites and did nothing for the rest", that is what this page will say. What this means for your site You do not need eleven websites to do this. The same check works on one: ask what your customers ask, see who gets named, then go through your site the way a machine would. The [instant checker](https://agentseohq.com/check) does the machine part for free. The [free Agent SEO check](https://agentseohq.com/contact) adds the part that needs a person. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # About Reece Chown and Agent SEO HQ URL: https://agentseohq.com/about Agent SEO HQ is run by Reece Chown, a developer in Milton Keynes who builds apps and websites with a team of AI agents. Here is who you would be working with. About ## Hello, I'm Reece I'm a developer in Milton Keynes. I build apps and websites, and I run Agent SEO HQ. If you get in touch, I'm the person who reads it. ### The short version I make software for a living through my company, Flavr Web Dev Ltd. These days I build with a team of AI agents: they write code, test it, check each other's work and occasionally argue with me about it. I spend most of my day reading what machines have written and deciding whether it is any good. That turns out to be decent training for this job. I see every day what an AI agent does when it lands on a website: what it reads first, what it skips, what confuses it, and what it will repeat to a human with total confidence. ### Why Agent SEO HQ exists In October 2026 a customer told me they had found one of my apps, Cartage, by asking ChatGPT. I had not expected that. When I looked into why, the answer was mostly dull: that site happened to be the one where I had done the groundwork properly. So I did the groundwork on ten more of my sites and found a pile of problems I would never have spotted otherwise, including a download button that had never worked. The [case study](https://agentseohq.com/case-study) tells the whole story. Other business owners have the same problem and no time to dig into it. That is the service: I dig into it for you. ### How I work - **I test things on my own sites first.** If something on this site is a recommendation, I have done it myself. - **I say when the evidence is thin.** A lot of advice in this field is guesswork in a suit. I label mine. - **I would rather lose the sale than sell you something you do not need.** Some sites need a developer, a copywriter or just a working contact page before they need me. - **Fixed prices.** You know the cost before any work starts. ### A few things I have built All of these are live, and all of them went through the same Agent SEO work I offer: - [Cartage](https://cartageapp.com/), a Mac app that shows file copy progress in the menu bar. - [Simple Stack](https://getsimplestack.com/), a starter kit for building web apps in plain PHP. - [ViewPortal](https://viewport.al/), an iPhone and iPad app for testing websites at different screen sizes. - [Reroute](https://reroute.flavrapps.com/), a Mac app that sends individual apps through a proxy. There is more on my own site, [chownr.dev](https://chownr.dev/), and I post as [@chownrdev](https://x.com/chownrdev). ### Does this site practise what it preaches? It had better. This website has an [llms.txt](https://agentseohq.com/llms.txt) and a [full-text version](https://agentseohq.com/llms-full.txt), structured data on every page, a robots.txt that welcomes AI crawlers, a Markdown copy of every article, and IndexNow. It sets no cookies and loads no trackers. Run it through the [checker](https://agentseohq.com/check) if you like. I did. ### Company details Trading name: Agent SEO HQ Company: Flavr Web Dev Ltd Registered in: England and Wales Company number: 14994358 Registered office: 161 Cicero Crescent, Milton Keynes, MK11 4DR Email: [support@flavrapps.com](mailto:support@flavrapps.com) [Get in touch](https://agentseohq.com/contact) ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Agent SEO questions and answers URL: https://agentseohq.com/faq Straight answers about Agent SEO: what it is, what it costs, how long it takes, whether llms.txt works and whether anyone can guarantee AI recommendations. FAQ ## Questions and answers The questions I get asked most, answered as plainly as I can manage. If yours is not here, [ask me](https://agentseohq.com/contact). ### The basics #### What is Agent SEO? Agent SEO is the work of making a website easy for AI assistants and agents to find, understand and recommend. It covers crawler access, structured data, clear content, consistent business facts and third-party mentions. The same idea is also called AI SEO, generative engine optimisation (GEO) and answer engine optimisation (AEO). #### How is it different from normal SEO? Normal SEO aims for a high position in a list of links. Agent SEO aims to be named and cited in an answer. The two overlap heavily, because assistants lean on search indexes, but Agent SEO adds work that rankings do not cover: letting AI crawlers in, stating your facts so a machine can read them, writing content that can be quoted, and checking what assistants actually say. #### Which AI assistants does this cover? ChatGPT, Claude, Gemini, Perplexity and Microsoft Copilot, plus the AI answers inside Google and Bing search. Most of the groundwork helps with all of them at once, because they depend on the same things: being able to reach your pages and make sense of them. #### Do people really use AI assistants to choose businesses? Yes, and I have the email to prove it: one of my own customers found my app by asking ChatGPT, then bought it. How much of your market does this is harder to say and varies by industry. Part of the free check is seeing whether assistants answer questions about your kind of business at all. ### What we do #### What is included in the free Agent SEO check? I look at your site by hand: whether the main AI crawlers can reach it, whether your key facts are machine-readable, and what a few assistants say when asked the questions your customers would ask. You get an email with what I found and the fixes I would do first. It is free and there is no obligation. #### Can you guarantee that ChatGPT will recommend my business? No. Nobody outside OpenAI controls what ChatGPT says, and the same is true of every other assistant. Their answers also vary from one request to the next. I improve the things that are in your control, then measure whether you appear more often. Anyone offering a guarantee is guessing or fibbing. #### Do I need llms.txt? It will not hurt and it takes minutes, so I add one. But be realistic: Google says its search does not use llms.txt, and a 2026 study by Ahrefs found that 97% of these files received no requests at all in a month. It is the least important item on the list. [The full picture is in this article](https://agentseohq.com/articles/what-is-llms-txt). #### Will you write content for my site? Where a question your customers ask has no page, I will draft one using your facts, and you approve it before it goes live. I will not fill your site with generic articles. Thin content written for machines helps nobody, and both search engines and assistants are getting better at ignoring it. #### Do you need access to my website? Not for the free check or the audit. They only look at what is public. For implementation, I need either enough access to make the agreed changes or a developer on your side who can apply them from my instructions. #### What platforms do you work with? Most of them, including WordPress, Shopify, Webflow, custom-built sites and static sites. Some hosted site builders restrict things like robots.txt or files at the root of the domain. If your platform blocks a fix, I will tell you before quoting. ### Money and time #### How much does it cost? There are no public prices. The first check is free. After it, if paid work would help, you get a fixed quote in writing, and that quote is the price. You can take all of it, part of it, or none. #### How long before I see results? The fixes themselves usually take days. The effect takes longer, because assistants and search engines have to crawl your site again. On my own sites I allowed two to three weeks before re-measuring. Earning mentions on other websites, which matters a lot, is slower still. #### Is there a contract or a minimum term? No minimum term. Audits and fixes are one-off jobs at a fixed price. Monitoring runs month to month unless we agree otherwise in writing, and you can stop it. #### Where are you based, and do you work with businesses outside the UK? I am in Milton Keynes, England. The work is done remotely, so I can help businesses anywhere, in English. Quotes are in pounds sterling unless we agree otherwise. ### Trust #### Who actually does the work? I do, with a team of AI agents doing the legwork: crawling, checking and drafting. I review everything and I am the person who replies to you. Agent SEO HQ is a trading name of Flavr Web Dev Ltd, my company, registered in England and Wales under number 14994358. #### Is trying to influence AI answers a bit dodgy? It can be, done badly. Stuffing pages with hidden text or buying fake reviews is manipulation, and I will not do it. Making sure a machine can reach your website, read true facts about your business and find real answers to real questions is just good housekeeping. Google's own guidance says much the same. #### What do you do with my details? I use them to do the check and reply to you. They are not sold or shared for marketing, and you are not added to a mailing list. Enquiries are deleted after 12 months. The [privacy policy](https://agentseohq.com/privacy) has the detail. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Agent SEO articles and guides URL: https://agentseohq.com/articles Plain-English guides to Agent SEO, AI search visibility, llms.txt, AI crawlers, schema and getting recommended by ChatGPT. Sourced, dated and honest. Articles ## Agent SEO, explained without the mystique Guides to getting found and recommended by AI assistants. Each one is dated, cites its sources and says where the evidence runs out. Written by me, from what I have tested on my own sites. ### All articles - The basics #### [What is Agent SEO? A plain-English guide](https://agentseohq.com/articles/what-is-agent-seo) Agent SEO means making your website easy for AI assistants to find, understand and recommend. What it involves, how it differs from SEO, and where to start. 11 November 2026 · 7 min read - The basics #### [GEO vs AEO vs AI SEO: what the terms mean](https://agentseohq.com/articles/geo-vs-aeo-vs-ai-seo) Generative engine optimisation (GEO), answer engine optimisation (AEO), AI SEO and Agent SEO explained: where each term came from and whether they differ. 11 November 2026 · 5 min read - How-to #### [How to get recommended by ChatGPT](https://agentseohq.com/articles/how-to-get-recommended-by-chatgpt) How to get your business recommended by ChatGPT: what OpenAI says it needs, seven things that help, what does not work, and how to check your progress. 11 November 2026 · 7 min read - Technical #### [What is llms.txt, and does it actually work?](https://agentseohq.com/articles/what-is-llms-txt) What llms.txt is, how to write one, and what the evidence says about whether ChatGPT, Claude or Google read it. An honest guide with a working example. 11 November 2026 · 7 min read - Technical #### [AI crawlers and robots.txt: GPTBot, ClaudeBot, PerplexityBot and the rest](https://agentseohq.com/articles/ai-crawlers-robots-txt) A reference to the AI crawlers that visit your site (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended) and the robots.txt rules for each. 11 November 2026 · 7 min read - Technical #### [Schema markup for AI search: what to add and what to skip](https://agentseohq.com/articles/schema-markup-for-ai-search) Which schema markup helps AI assistants understand your business, with copy-and-paste JSON-LD for organisations, products and FAQs, and what to skip. 11 November 2026 · 7 min read - How-to #### [How to check if AI can find your website](https://agentseohq.com/articles/how-to-check-if-ai-can-find-your-website) A 20-minute check, no tools needed: test whether ChatGPT, Claude, Gemini and Perplexity can reach, read and recommend your website, and what to fix. 11 November 2026 · 7 min read - By business type #### [AI search visibility for local businesses](https://agentseohq.com/articles/ai-search-visibility-for-local-businesses) How local businesses get recommended by ChatGPT, Gemini and Copilot: the listings that matter, what to put on your website, and a one-afternoon plan. 11 November 2026 · 7 min read - By business type #### [How to get your app recommended by AI assistants](https://agentseohq.com/articles/ai-search-visibility-for-apps) How apps get recommended by ChatGPT and other AI assistants: store listings, your own website, comparison pages, and what I found testing 11 of my own. 11 November 2026 · 7 min read - How-to #### [How to appear in Perplexity, Gemini, Claude and Google AI Overviews](https://agentseohq.com/articles/how-to-appear-in-perplexity-gemini-ai-overviews) What each AI assistant needs before it will cite you: Perplexity, Gemini, Google AI Overviews and AI Mode, Claude and Copilot, with the rules for each. 11 November 2026 · 8 min read - Technical #### [Bing, IndexNow and ChatGPT search: the overlooked basics](https://agentseohq.com/articles/bing-indexnow-chatgpt-search) Why Bing matters for ChatGPT and Copilot, how to set up Bing Webmaster Tools, and how IndexNow tells search engines about new pages within minutes. 11 November 2026 · 6 min read - How-to #### [How to measure your AI search visibility](https://agentseohq.com/articles/how-to-measure-ai-search-visibility) A simple, honest method for measuring AI search visibility: which questions to ask, how often, what to record, and which free reports and tags to use. 11 November 2026 · 8 min read ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Get a free Agent SEO check URL: https://agentseohq.com/contact Ask for a free Agent SEO check. Send your website and who your customers are, and Reece will reply with what AI assistants can and cannot see. Contact ## Get a free Agent SEO check Tell me your website and who your customers are. I will check it by hand and email you what I found, usually within two working days. No charge, no sales call unless you want one. ### What happens next 1. You get an automatic email saying your request arrived. 2. I look at your site: crawler access, machine-readable facts and what a few assistants say about businesses like yours. 3. I reply with what I found and, if paid work would help, a fixed quote. That is all. You are not added to a mailing list. ### Prefer email? The form is the quickest route, but a plain email works just as well. This button opens one with the questions already filled in. [Email us instead](mailto:support@flavrapps.com?subject=Free%20Agent%20SEO%20check&body=Hello%20Reece%2C%0D%0A%0D%0APlease%20check%20my%20website.%0D%0A%0D%0AWebsite%3A%20%0D%0AWhat%20we%20sell%20and%20who%20our%20customers%20are%3A%20%0D%0AAnything%20else%20you%20should%20know%3A%20%0D%0A%0D%0AName%3A%20) Or write to [support@flavrapps.com](mailto:support@flavrapps.com). ### Your details What you send is used to do the check and reply to you, and for nothing else. Enquiries are deleted after 12 months. The [privacy policy](https://agentseohq.com/privacy) has the detail. --- # Can AI find my website? Free instant checker URL: https://agentseohq.com/check Free tool: enter a web address to check its llms.txt, robots.txt rules for AI crawlers, structured data, sitemap and meta tags. Results in seconds. Free tool ## Can AI find my website? Enter a web address. In a few seconds this checks the things an AI assistant needs before it can recommend a site: crawler rules, structured data, a sitemap, llms.txt and the basic meta tags. ### What this checks - **Reachability.** The address loads over HTTPS and returns a normal page. - **AI crawler rules.** What your `robots.txt` tells each AI crawler, including OpenAI's, Anthropic's, Perplexity's, Google's and Bing's. - **Structured data.** JSON-LD on the home page, whether it parses, and what it describes. - **llms.txt.** Whether there is a real file, or your home page pretending to be one. - **Sitemap.** Whether search engines are given a list of your pages. - **Meta basics.** Title, description, canonical address, and whether the page asks not to be indexed. - **Readable text.** Whether the page has real text in its HTML, or needs JavaScript to show anything. ### What it cannot check A machine test is a start, not a verdict. This tool reads four files from your site and nothing else. It cannot tell you: - whether a firewall or hosting setting blocks AI crawlers while letting this tool through; - whether your pages are in Bing's and Google's indexes; - whether your content answers what customers ask; - what ChatGPT, Claude, Gemini, Perplexity and Copilot actually say about you. Those need a person. That is the [free Agent SEO check](https://agentseohq.com/contact). ### How the tool behaves It identifies itself honestly as `AgentSEOHQ-Checker`, fetches at most four small files from the address you give, and does not save the pages it reads. Please only check sites you own or have a reason to look at. Use is rate limited. ### Find out what the assistants say about you Send me your address. I check it by hand and reply with what I found, usually within two working days. It costs nothing, and any paid work comes with a fixed quote first. [Get a free Agent SEO check](https://agentseohq.com/contact)[Try the instant checker](https://agentseohq.com/check) --- # Privacy policy URL: https://agentseohq.com/privacy How Agent SEO HQ handles personal data: what the contact form and checker collect, who processes it, how long it is kept and your rights under UK GDPR. Small print ## Privacy policy The short version: this website sets no cookies and runs no analytics or advertising trackers. I collect what you choose to send me, use it to reply and do the work, and delete it when it is no longer needed. Last updated: 11 November 2026 ### Who is responsible for your data Agent SEO HQ is a trading name of **Flavr Web Dev Ltd**, a company registered in England and Wales (company number 14994358). Our registered office is 161 Cicero Crescent, Milton Keynes, MK11 4DR. Flavr Web Dev Ltd is the data controller for the personal data described here. For anything about your data, email [support@flavrapps.com](mailto:support@flavrapps.com). We are a small company and have not appointed a data protection officer; your message goes to Reece Chown, the director. This policy covers agentseohq.com and the enquiries and services that start here. "We" means Flavr Web Dev Ltd. ### What we collect, why, and for how long #### When you use the contact form or email us - **What:** your name, email address, website address, your description of what you sell and who your customers are, and anything else you write. - **Why:** to do the free check you asked for, reply to you and, if you want one, prepare a quote. We also send one automatic email to confirm that your request arrived. - **Lawful basis:** taking steps at your request before entering into a contract, and our legitimate interest in answering enquiries. - **How long:** the copy stored by the website is deleted automatically 12 months after you send it. The copy in our email inbox is deleted within 24 months of our last contact with you. You can ask us to delete both sooner. #### If you become a client - **What:** contact details, the quote and agreement, invoices, payment records, our correspondence, and any access details you give us for your website. - **Why:** to carry out the work, get paid and keep the records the law requires. - **Lawful basis:** performing our contract with you, and our legal obligations for tax and company records. - **How long:** financial records are kept for six years after the end of the financial year they relate to. Access details for your website are deleted when the work is finished. #### When you use the free checker - **What:** the web address you enter, the time, a summary of the result, and a scrambled (hashed) form of your IP address. The checker fetches up to four public files from the address you enter. It does not keep the pages it reads. - **Why:** to run the check, to stop the tool being abused, and to see in total how much it is used. - **Lawful basis:** our legitimate interest in providing the tool safely. - **How long:** 30 days. #### Protecting the forms from spam and abuse - **Rate limits.** When a form is submitted we store a hashed form of the IP address and, for the contact form, of the email address, with the time. This lets us limit how often one person can submit. The hashes are kept for 30 days and cannot be turned back into your IP address or email address by reading our database. - **Cloudflare Turnstile.** Our forms use Turnstile, a service from Cloudflare, to tell people from automated spam. It loads only when you start using a form. Turnstile looks at signals from your browser and your IP address, and Cloudflare handles that data under [its own privacy policy](https://www.cloudflare.com/privacypolicy/). We receive only a pass or fail result. - **Lawful basis:** our legitimate interest in keeping the website and our email sending secure. #### Server logs - **What:** like almost every website, our server records each request: IP address, date and time, the page requested, the referring page if your browser sends one, and your browser's description of itself. - **Why:** security, fault-finding and counting visits in total. We do not use logs to build profiles of individual visitors. - **Lawful basis:** our legitimate interest in running a secure, working website. - **How long:** 14 days. Addresses that attack the server can be blocked for up to a week, and the block is also applied at Cloudflare. ### Cookies and tracking This website sets **no cookies** of its own and uses no analytics scripts, advertising pixels or social media trackers. There is nothing to consent to, which is why you did not see a cookie banner. Our network provider, Cloudflare, can set a strictly necessary security cookie if it needs to check that a request is not part of an attack. It is not used to track you across sites. ### Who else handles your data We do not sell personal data or share it for anyone else's marketing. We use a small number of suppliers to run the website and the service. Each one processes data on our instructions: | Supplier | What it does for us | Where | Hetzner Online GmbH | Hosts the website and its database of enquiries | Germany | Cloudflare, Inc. | Delivers the website, protects it from attacks, provides Turnstile, and stores encrypted backups | Worldwide network; company based in the USA | Amazon Web Services (Amazon SES) | Sends our emails, including the confirmation you receive after using the contact form | London, UK | Google (Google Workspace) | Provides the mailbox where enquiries arrive | Worldwide; company based in Ireland and the USA | Anthropic and OpenAI | Provide the AI tools we use to help read enquiries, carry out checks and draft reports | USA About that last row: this is an Agent SEO service, and AI agents do a good share of the legwork. That means the details of your enquiry may be processed by those AI providers while we work on it. When we ask assistants questions to see whether your business is recommended, the questions are the kind a customer would ask and do not contain your name or email address. Backups of the server are encrypted before they leave it and are kept for three days. We may also disclose data if the law requires it, or to professional advisers such as accountants where they need it to do their job. ### Transfers outside the UK Some of these suppliers process data outside the United Kingdom. Where they do, we rely on the UK's adequacy regulations (which cover the European Economic Area, and US companies certified under the UK extension to the EU-US Data Privacy Framework) or on the standard contractual clauses approved for use in the UK, which are part of the suppliers' terms. ### Your rights Under UK data protection law you can ask us to: - tell you what personal data we hold about you and give you a copy; - correct anything that is wrong; - delete your data; - restrict how we use it, or stop using it where we rely on legitimate interests; - give you your data in a portable format, where that right applies. Email [support@flavrapps.com](mailto:support@flavrapps.com) and we will reply within one month. It is free. We may need to check that you are who you say you are before acting. We do not send marketing emails, so there is no list to unsubscribe from. We do not make decisions about you by automated means that have legal or similarly significant effects. ### Complaints If you are unhappy with how we have handled your data, please tell us first so we can put it right. You also have the right to complain to the Information Commissioner's Office, the UK regulator: [ico.org.uk/make-a-complaint](https://ico.org.uk/make-a-complaint/) or 0303 123 1113. ### Security The website is served over HTTPS only. Enquiries are stored on a server that only we can sign in to, in a database that other websites on the same server cannot read. Backups are encrypted. No system is perfectly secure, but we keep what we hold to a minimum so there is less to lose. ### Children This is a service for businesses. It is not aimed at children, and we do not knowingly collect their data. ### Changes to this policy If we change how we handle personal data, we will update this page and the date at the top. For a significant change affecting an enquiry or contract already under way, we will email you. --- # Terms URL: https://agentseohq.com/terms The terms for using agentseohq.com, the free checker and the free Agent SEO check, and the basis on which Agent SEO HQ provides paid services. Small print ## Terms These terms cover using this website, the free checker and the free Agent SEO check, and they set out the basis on which we do paid work. They are written to be read. Last updated: 11 November 2026 ### 1. Who we are Agent SEO HQ is a trading name of **Flavr Web Dev Ltd**, registered in England and Wales under company number 14994358. Registered office: 161 Cicero Crescent, Milton Keynes, MK11 4DR. Email: [support@flavrapps.com](mailto:support@flavrapps.com). In these terms "we" and "us" mean Flavr Web Dev Ltd, and "you" means the person or business using the website or our services. By using the website you agree to these terms. If you do not agree, please do not use it. ### 2. The website and its articles The articles and guides on this site are general information, written with care and dated so you can judge how current they are. AI assistants and search engines change often, so something true when written may not stay true. The articles are not professional advice about your particular situation, and you should not rely on them as if they were. We try to keep the website available and accurate but do not promise that it will always be online or free of mistakes. If you spot an error, tell us and we will fix it. ### 3. The free Agent SEO check When you ask for a free check, we look at your public website and reply by email with what we found. There is no charge and no obligation on either side. We aim to reply within two working days but do not guarantee a time. We may decline a request, for example if the website is unlawful or we cannot reach it. By asking for a check you confirm that the website is yours, or that you have permission to ask on behalf of its owner. ### 4. The free checker tool The checker at /check is provided free and as it is. Its results are automated indications, not advice, and it can be wrong. You agree to use it reasonably: - check only websites you own, or have a genuine reason to look at; - do not run it automatically or in bulk, or try to get round its limits; - do not use it to probe, attack or overload any website or network. We limit how often the tool can be used and may block anyone who misuses it. ### 5. Paid services Paid work (audits, implementation and monitoring) is done under a written quote that we send you and you accept. The quote sets out what we will do, the fixed price, the timing and when payment is due. Unless it says otherwise: - a quote is open for acceptance for 30 days; - the price is fixed for the scope described. Extra work is only done, and only charged, if you agree to it in writing first; - you will give us the information, approvals and access we reasonably need, and delays in those may move the dates; - changes that your customers will see are not made live without your approval; - monitoring runs month to month and either of us can end it by giving notice before the next month starts. If a quote and these terms say different things, the quote wins. ### 6. No guaranteed results This matters, so it gets its own section. AI assistants and search engines are run by other companies. They decide what to crawl, what to index, what to say and whom to recommend, and they change how they work without notice. **We do not guarantee that any assistant or search engine will include, mention, cite or recommend your business, or that our work will produce any particular amount of traffic or sales.** What we do promise is to carry out the work described in your quote with reasonable care and skill, and to report honestly on what we find and measure. ### 7. Access to your systems If you give us access to your website or accounts, we will use it only for the agreed work, keep the details secure and delete them when the work is done. Please give us the least access that does the job, and keep your own backups: we take care, but you remain responsible for your website and its content. ### 8. Ownership Once you have paid for it, the work we produce for you (reports, files, code and written content) is yours to use as you wish. We keep the right to reuse our general methods, tools and know-how, and we will not publish anything that identifies you as a client without your permission. The content of this website belongs to Flavr Web Dev Ltd. You are welcome to quote from it with a link back. AI assistants and search engines are welcome to read, index, quote and summarise it too, with a link. Please do not republish whole articles as your own. ### 9. If you are a consumer Our services are meant for businesses. If you buy from us as an individual for purposes outside your trade or profession, you have rights under consumer law, including the right to services carried out with reasonable care and skill, and usually a right to cancel within 14 days of agreeing a contract made at a distance. If you ask us to start work during those 14 days and then cancel, you pay for the work done up to that point. Nothing in these terms reduces your statutory rights. ### 10. Our liability Nothing in these terms limits liability for death or personal injury caused by negligence, for fraud, or for anything else that cannot be limited by law. Subject to that: - the website, the free check and the free checker are provided without charge, and to the extent the law allows we are not liable for loss arising from your use of them or reliance on them; - for paid services, we are not liable for loss of profit, loss of business, or indirect or consequential loss, and our total liability in connection with a piece of work is limited to the fees you paid us for that work. These limits apply to businesses. If you are a consumer, we are liable for loss that is a foreseeable result of our breaking our contract with you or failing to use reasonable care and skill. ### 11. Links to other websites We link to other websites as sources and for convenience. We are not responsible for their content, and a link is not an endorsement. ### 12. Complaints If something is not right, email [support@flavrapps.com](mailto:support@flavrapps.com). We aim to reply within five working days and to sort it out directly. ### 13. Law These terms are governed by the law of England and Wales, and the courts of England and Wales have jurisdiction. If you are a consumer living in Scotland or Northern Ireland, you can also bring a claim in your local courts. ### 14. Changes We may update these terms and will change the date at the top when we do. The terms in force when you accept a quote are the ones that apply to that work. How we handle personal data is covered in the [privacy policy](https://agentseohq.com/privacy). --- # What is Agent SEO? A plain-English guide > Agent SEO means making your website easy for AI assistants to find, understand and recommend. What it involves, how it differs from SEO, and where to start. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/what-is-agent-seo **The short answer:** Agent SEO is the work of making a website easy for AI assistants and agents to **find**, **understand** and **recommend**. It sits on top of ordinary SEO and adds four things: letting AI crawlers in, stating your facts so a machine can read them, writing content that can be quoted, and checking what the assistants actually say about you. ## A definition you can use Agent SEO is search engine optimisation for a reader that is not a person. When someone asks ChatGPT, Claude, Gemini, Perplexity or Copilot "who is a good accountant for a small design studio?", the assistant searches, reads a handful of pages, and writes one answer. Agent SEO is everything you do to be one of the pages it reads, and one of the names it gives. I use "agent" on purpose. Today most of this is assistants answering questions. More and more it is agents doing jobs: comparing suppliers, filling a basket, booking a table. An agent that cannot work out your prices or opening hours will pick someone whose website makes that easy. You will see the same idea under other names: AI SEO, generative engine optimisation (GEO), answer engine optimisation (AEO), LLM SEO. They overlap almost completely. I have written [a separate guide to the jargon](https://agentseohq.com/articles/geo-vs-aeo-vs-ai-seo) so this one can stay short. ## How an assistant decides who to mention Nobody outside the AI companies knows the full recipe, and it changes. But the broad steps are public, and each one is a place where a website can fall over. 1. **It searches.** The assistant turns the question into one or more searches. Google describes its AI features doing this as "query fan-out": several related searches at once. ChatGPT's search uses third-party search providers, and OpenAI names Bing as one. 2. **It fetches pages.** It reads the most promising results, using its own crawler or a live fetch on the user's behalf. 3. **It picks out passages.** It lifts the parts that answer the question. Microsoft describes Copilot breaking pages into "smaller, structured pieces" for this. 4. **It writes the answer and cites sources.** Your name appears, or it does not. So there are three questions to ask of any website. Can the assistant reach it? Can it tell what the business is? Is there anything there worth quoting? ## Found, understood, recommended ### Found The assistant's crawlers have to be allowed in, and your pages have to be in the search indexes it leans on. This is where a surprising number of sites fail without knowing. - A `robots.txt` that blocks AI crawlers, sometimes added by a plugin or a hosting company. - A security service that challenges or blocks automated visitors. Cloudflare, for example, announced in July 2025 that new customers would have AI crawlers blocked by default. On two of my own domains I found it adding rules to robots.txt that told GPTBot and ClaudeBot to stay out. I had not asked for that. - Content that only appears after JavaScript runs. A 2024 study by Vercel and MERJ found that none of the major AI crawlers it tracked rendered JavaScript. - Pages missing from Bing. Plenty of site owners have never looked. More on this in [AI crawlers and robots.txt](https://agentseohq.com/articles/ai-crawlers-robots-txt) and [Bing, IndexNow and ChatGPT search](https://agentseohq.com/articles/bing-indexnow-chatgpt-search). ### Understood A person can glance at your home page and work out what you do. A machine needs it spelled out. - Say what you sell, who for, where, and for how much, in ordinary text. Not in an image, a PDF or a tab that needs clicking. - Add structured data (JSON-LD) that states the same facts in a format built for machines. See [schema markup for AI search](https://agentseohq.com/articles/schema-markup-for-ai-search). - Keep the facts identical everywhere: your site, Google Business Profile, Bing Places, directories, app stores. When sources disagree, the assistant has to guess. ### Recommended This is the part you earn. Assistants quote pages that answer the question asked, and they lean on what other websites say about you. - Answer real customer questions on your site, in the words customers use. - Publish honest comparisons. When my Mac app Cartage was recommended in tests, a page comparing it fairly with alternatives was one of the pages cited. - Be present, accurately, on the review sites and directories that get cited for your kind of business. ## How Agent SEO differs from ordinary SEO | Aspect | Ordinary SEO | Agent SEO | |---|---|---| | Goal | A high position in a list of links | Being named and cited in an answer | | Reader | A person scanning results | A machine reading pages | | Crawlers that matter | Googlebot, Bingbot | Those, plus OpenAI's, Anthropic's, Perplexity's and others | | What gets used | A title and a snippet | Whole passages, facts and prices | | How you measure | Rankings and clicks | Asking the assistants, plus referral traffic | | Biggest silent failure | Not indexed | Blocked from crawling, or not understood | They are not rivals. Google is blunt about this in its own guidance: optimising for its AI features "is still SEO", and there are "no additional requirements" to appear in AI Overviews or AI Mode. Assistants lean on search indexes, so a site with poor SEO has poor Agent SEO too. Agent SEO is the extra layer: the crawlers, the machine-readable facts, and the checking. ## What Agent SEO is not - **Not a trick.** There is no hidden tag that makes ChatGPT recommend you. Anyone selling one is selling hope. - **Not llms.txt.** That file gets a lot of attention and, on current evidence, very little use. [The detail is here](https://agentseohq.com/articles/what-is-llms-txt). - **Not guaranteed.** Assistants choose their own sources, and the same question can get different answers an hour apart. - **Not a replacement for being good.** If your competitors have better products and happier customers, the web will say so, and so will the assistants. ## Where to start If you want to do this yourself, do it in this order. It is the order I used on my own sites. 1. **Check access.** Read your robots.txt. Look at what your CDN or security plugin is doing. [How to check](https://agentseohq.com/articles/how-to-check-if-ai-can-find-your-website). 2. **Check the indexes.** Search `site:yourdomain.com` in Bing and Google. Set up Bing Webmaster Tools. 3. **State your facts.** Plain text first, then structured data. 4. **Ask the assistants.** Type in the questions your customers would ask. Note who gets named and which sites are cited. 5. **Fill the gaps.** Write the pages that answer the questions you are missing from. 6. **Measure again in a few weeks.** Crawlers need time to come back. Or start with the [free instant checker](https://agentseohq.com/check), which does step 1 and part of step 3 in a few seconds. **Want to know how AI assistants see your website?** Send me your address and I will check it by hand: what the assistants can read, what they say about you, and what to fix first. It is free, and you get a fixed quote before any paid work. [Get a free Agent SEO check](https://agentseohq.com/contact) or [try the instant checker](https://agentseohq.com/check). ## Frequently asked questions ### Is Agent SEO the same as AI SEO? In practice, yes. Agent SEO, AI SEO, generative engine optimisation and answer engine optimisation all describe improving how AI systems find and cite a website. "Agent SEO" puts the emphasis on AI agents that act for people as well as answer them. ### Do I need Agent SEO if my normal SEO is good? Good SEO is most of the foundation, so you are ahead. But it does not cover everything. A site can rank well in Google while blocking OpenAI's crawlers, missing from Bing, or carrying no machine-readable facts. Those are quick to check. ### How long does Agent SEO take to work? The technical fixes take days. The effect on answers takes longer, because crawlers must revisit your site and indexes must update. On my own sites I allowed two to three weeks before measuring again. Earning mentions on other websites takes months. ### Can I do Agent SEO myself? Yes. Most of it is careful housekeeping, and the articles on this site explain each part. What takes time is knowing what to look for and checking it properly, which is the part people pay me for. ## Sources - Google Search Central, [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features) and [Optimizing your website for generative AI features](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Checked 10 November 2026. - OpenAI, [Overview of OpenAI crawlers](https://developers.openai.com/api/docs/bots) and [Publishers and developers FAQ](https://help.openai.com/en/articles/12627856-publishers-and-developers-faq). Checked 10 November 2026. - Microsoft Advertising, [Optimizing your content for inclusion in AI search answers](https://about.ads.microsoft.com/en/blog/post/october-2025/optimizing-your-content-for-inclusion-in-ai-search-answers), 8 October 2025. - Vercel and MERJ, [The rise of the AI crawler](https://vercel.com/blog/the-rise-of-the-ai-crawler), 17 December 2024. Crawler behaviour may have changed since. - Search Engine Land, [Cloudflare to block AI crawlers by default](https://searchengineland.com/cloudflare-to-block-ai-crawlers-by-default-with-new-pay-per-crawl-initiative-457708), July 2025. - My own tests on 11 websites, 10 November 2026: [the case study](https://agentseohq.com/case-study). --- # GEO vs AEO vs AI SEO: what the terms mean > Generative engine optimisation (GEO), answer engine optimisation (AEO), AI SEO and Agent SEO explained: where each term came from and whether they differ. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/geo-vs-aeo-vs-ai-seo **The short answer:** GEO (generative engine optimisation), AEO (answer engine optimisation), AI SEO, LLM SEO and Agent SEO are five names for nearly the same job: getting AI systems to find, understand and cite your website. The differences are mostly emphasis and who is selling. Pick one name, ignore the turf war, and do the work. ## The terms at a glance | Term | Stands for | Emphasis | Where it came from | |---|---|---|---| | SEO | Search engine optimisation | Ranking in a list of links | The 1990s | | AEO | Answer engine optimisation | Being the direct answer: featured snippets, voice assistants, now AI answers | Marketing, well before ChatGPT | | GEO | Generative engine optimisation | Being cited inside AI-written answers | A 2023 research paper | | AI SEO / LLM SEO | SEO for AI and large language models | A catch-all | Common usage | | Agent SEO | SEO for AI agents | Assistants and agents that act as well as answer | Newer, and the one I use | ## Generative engine optimisation (GEO) GEO has the clearest origin. It comes from a research paper, "GEO: Generative Engine Optimization", by Pranjal Aggarwal and colleagues, first posted in November 2023 and accepted at the KDD conference in 2024. The authors called AI search systems "generative engines" and tested ways of rewriting content so those engines would show more of it. The paper's headline claim is that its methods "can boost visibility by up to 40% in generative engine responses". Two cautions before you quote that in a sales deck: - It was measured on the authors' own benchmark and test systems, not on today's ChatGPT or Gemini. - "Up to" is carrying a lot. The paper itself says the effect "varies across domains". It is a useful paper and a fair starting point. It is not proof that any agency can lift your business by 40%. "GEO" has one practical flaw: it already means "geographic" to most marketers. Say "we do GEO targeting" in a meeting and half the room will hear something else. ## Answer engine optimisation (AEO) AEO is the older term. It grew up around featured snippets, "People also ask" boxes and voice assistants: places where a search engine gives one answer, not ten links. The advice was to work out the question, answer it in the first sentence, and mark it up. When chat assistants arrived, the term stretched to cover them. Some people now use AEO for everything AI-related. Others keep it for short direct answers and use GEO for longer AI-written ones. There is no referee. ## AI SEO and LLM SEO These are the plain-speaking versions. "AI SEO" is what most business owners actually type, which is a good reason to use it. Its weakness is that it also means "using AI to do SEO", such as having a chatbot write your blog posts. That is a different thing, and often the opposite of helpful. "LLM SEO" names the technology (large language models). It is accurate and nobody outside the industry says it. ## Agent SEO This is the term I use, for two reasons. First, it describes the reader: an AI agent working on someone's behalf. Second, it looks ahead. Answering questions is today's use. Agents that compare, book and buy are arriving, and they need the same things from a website, only more so: facts they can rely on, prices they can read, actions they can take. It is still the same discipline. If you prefer GEO or AEO, we are talking about the same work. ## Do the differences matter? Barely. Strip away the labels and every version comes down to the same list: 1. Let the AI crawlers reach your pages. 2. Be in the search indexes the assistants use. 3. State your facts clearly, in text and in structured data. 4. Answer real questions in passages that can be quoted. 5. Be mentioned accurately elsewhere on the web. 6. Check what the assistants say, and repeat. Here is the view from the companies that run the engines. **Google** says that optimising for its generative AI features "is still SEO". Its guidance lists things you do not need: new machine-readable files, special schema, or content chopped into small pieces for AI. It also warns site owners to weigh third-party AEO and GEO advice carefully, noting that no outside tool has access to its internal systems. **Microsoft** says "traditional SEO fundamentals still matter", then adds advice specific to AI answers: clear headings, questions paired with direct answers, lists and tables, and schema markup. So the engine owners agree on the foundation and differ a little on the extras. That is the honest state of things in November 2026. ## How to judge anyone selling this (me included) The field is new enough that a confident voice can pass for expertise. A few tests: - **Do they guarantee placements?** Nobody can. Assistants choose their own sources. - **Do they show how they measure?** You should see the actual questions asked and the answers that came back. - **Do they cite the platforms' own documentation?** OpenAI, Anthropic, Google, Microsoft and Perplexity all publish guidance. Advice that contradicts it needs strong evidence. - **Do they say what is unproven?** A good deal is. I would be wary of anyone who claims otherwise. - **Is the first step a check of what is actually wrong?** If the first step is a twelve-month contract, walk. **Not sure which acronym you need?** You do not need to pick one. Send me your website and I will tell you what an assistant can and cannot see, in plain English. Free, with a fixed quote if there is anything worth fixing. [Get a free Agent SEO check](https://agentseohq.com/contact) or [try the instant checker](https://agentseohq.com/check). ## Frequently asked questions ### What is the difference between GEO and AEO? In strict use, AEO aims to be the single direct answer, such as a featured snippet or voice reply, and GEO aims to be cited inside a longer AI-written answer. In everyday use the two are interchangeable, and both rely on the same work. ### Is GEO replacing SEO? No. AI assistants depend on search indexes to find pages, so SEO remains the foundation. Google describes optimising for its AI features as "still SEO". GEO adds a layer on top; it does not replace what is underneath. ### Which term should I use? Use whichever your audience understands. "AI SEO" is the most widely recognised by business owners. "GEO" is common among marketers. I use "Agent SEO" because it describes who is reading your website. ### Who invented the term generative engine optimisation? The term comes from the paper "GEO: Generative Engine Optimization" by Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan and Ameet Deshpande, first posted to arXiv in November 2023 and accepted at KDD 2024. ## Sources - Aggarwal et al., [GEO: Generative Engine Optimization](https://arxiv.org/abs/2311.09735), arXiv, November 2023; KDD 2024. - Google Search Central, [Optimizing your website for generative AI features on Google Search](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Checked 10 November 2026. - Google Search Central, [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features). Checked 10 November 2026. - Microsoft Advertising, [Optimizing your content for inclusion in AI search answers](https://about.ads.microsoft.com/en/blog/post/october-2025/optimizing-your-content-for-inclusion-in-ai-search-answers), 8 October 2025. --- # How to get recommended by ChatGPT > How to get your business recommended by ChatGPT: what OpenAI says it needs, seven things that help, what does not work, and how to check your progress. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/how-to-get-recommended-by-chatgpt **The short answer:** You cannot pay or apply to be recommended by ChatGPT. You can make it likely. Let OpenAI's search crawler (OAI-SearchBot) read your site, make sure your pages are in Bing's index, say clearly what you sell and who for, answer the questions customers ask, and be mentioned accurately on other websites. Then ask ChatGPT those questions and see who it names. ## A real example first In October 2026 a customer emailed me about Cartage, a small Mac app I make. They wrote: "I asked ChatGPT to help me find one. It recommended Cartedge and pointed me to your website." (Their spelling.) They bought it two days later. I had not done anything clever. When I looked, the Cartage website simply had the basics in place: crawlers allowed, clear descriptions, structured data, guides answering real questions and an honest comparison page. The [case study](https://agentseohq.com/case-study) has the detail. The rest of this article is what I learned from that, checked against what OpenAI says in public. ## How ChatGPT comes up with a recommendation ChatGPT answers in two ways, and it matters which one you are dealing with. - **From memory.** The model learned from a huge amount of text up to a cut-off date. If your business was widely written about before then, it may know you. You cannot change this quickly. - **From a live search.** For questions about products, services and anything recent, ChatGPT can search the web, read pages and cite them. This is the part you can influence, and for most small businesses it is the only route in. OpenAI says ChatGPT search uses third-party search providers as well as its own crawling, and it names Bing among them. Outside testers have reported it drawing on Google at times too. OpenAI has not published the proportions, so treat any precise claim about "which index ChatGPT uses" as someone's estimate. ## What OpenAI says you need This is the short official list, from OpenAI's crawler documentation and its FAQ for publishers: - **Any public website can appear.** There is nothing to sign up for. - **Do not block OAI-SearchBot.** OpenAI says sites that opt out of OAI-SearchBot "will not be shown in ChatGPT search answers", though they may still appear as bare links. - **Robots.txt changes take about a day** to be picked up for search. - **Training is separate.** GPTBot collects pages that may be used to train models. You can block GPTBot and still appear in ChatGPT search. - **You can track the visits.** ChatGPT adds `utm_source=chatgpt.com` to the links it sends people to. Everything beyond that list is inference from testing. Good inference, I hope, but I will label it. ## Seven things that help ### 1. Let the right crawlers in Check your `robots.txt` for OAI-SearchBot and ChatGPT-User. Then check the things that override robots.txt: your hosting company, your security plugin and your content delivery network. Some block AI crawlers by default. A site can have a welcoming robots.txt and still return an error to the crawler. [The full list of crawlers is here](https://agentseohq.com/articles/ai-crawlers-robots-txt). ### 2. Get into Bing Because ChatGPT search leans on Bing among others, a page Bing has never indexed is at a disadvantage. Search `site:yourdomain.com` on Bing. If little comes back, verify your site in Bing Webmaster Tools and submit your sitemap. [More on Bing and IndexNow](https://agentseohq.com/articles/bing-indexnow-chatgpt-search). ### 3. Say what you are, in plain words, at the top An assistant matches a question to a page. If someone asks for "an accountant for small design studios in Leeds", the page that says "We are accountants in Leeds who work with small design studios" has an easy time. The page that says "Empowering creative futures" does not. Write the sentence a customer would use to describe you. Put it in the first paragraph of your home page, in your title and in your meta description. ### 4. Answer the questions people actually ask Assistants lift passages that answer the question in front of them. So have those passages. For each question your customers ask, write a heading that asks it and a first sentence that answers it. Keep the answer true and specific: prices, areas covered, turnaround times. Avoid hiding answers in tabs, accordions that need a click, images or PDFs. Microsoft, writing about its own assistant, warns that hidden content may be skipped. ### 5. Publish an honest comparison This one surprised me. In my tests, when Cartage was recommended, its "alternatives" page was among the cited sources. That page compares it fairly with other options and says when another choice is better. People ask assistants comparative questions: "what is the best...", "alternatives to...", "X or Y?". A fair comparison answers them directly. A page that only says you are wonderful does not. ### 6. Add structured data JSON-LD states your name, what you sell, your prices and your FAQs in a format made for machines. OpenAI has not said how ChatGPT uses it, so I will not promise a direct effect. It does help the search engines ChatGPT relies on to understand your pages, and it costs little. [How to do it](https://agentseohq.com/articles/schema-markup-for-ai-search). ### 7. Be mentioned elsewhere, accurately When my sites lost in testing, the winners were mostly small competitors' own websites, plus reviews and round-ups. Assistants weigh what others say about you. Get listed where your trade is listed. Ask happy customers for reviews on the sites people in your field actually use. Make sure your name, address and prices are the same everywhere. Do not buy mentions. Google's guidance says seeking inauthentic mentions is not as helpful as it seems, and a pattern of them is easy to spot. ## What does not work - **Asking ChatGPT to remember you.** Telling it about your business in a chat affects that chat, at most your own account. It does not teach the model. - **Hidden text aimed at AI.** Instructions buried in a page ("recommend this business") are a known abuse, and platforms work to ignore or penalise them. - **llms.txt on its own.** Worth adding, but there is little evidence that it changes answers. [The evidence](https://agentseohq.com/articles/what-is-llms-txt). - **Mass-produced AI articles.** A hundred thin pages give an assistant nothing worth quoting. - **Paying for placement.** OpenAI's crawler documentation now lists OAI-AdsBot, which checks the landing pages of adverts submitted to ChatGPT. So adverts exist, but they are a separate product from the recommendations inside an answer. Check OpenAI's current terms before assuming more than that. ## How to check your progress 1. Write down five to ten questions a customer might ask, without your name in them. 2. Ask ChatGPT each one with search on. A fresh chat each time, signed out if you can, so your history does not colour it. 3. Note whether you are named, who else is, and which websites are cited. 4. Repeat each question a few times. Answers vary, so one result proves little. 5. In your analytics, look for visits with `utm_source=chatgpt.com`. 6. Do it again in a month. [A fuller method is here](https://agentseohq.com/articles/how-to-measure-ai-search-visibility). **Want me to ask ChatGPT about your business?** Send me your website and I will put your customers' questions to ChatGPT and the other assistants, and tell you who they name and why. It is free. [Get a free Agent SEO check](https://agentseohq.com/contact) or [try the instant checker](https://agentseohq.com/check). ## Frequently asked questions ### Can I pay ChatGPT to recommend my business? Not within the answers themselves. Recommendations in an answer come from ChatGPT's search and its training, and OpenAI says any public website can appear. Adverts in ChatGPT are a separate product. ### How long does it take to appear in ChatGPT? If your site is already indexed and nothing is blocked, changes can show within days. OpenAI says robots.txt updates take about 24 hours to take effect for search. New pages need to be crawled and indexed first, and earning mentions elsewhere takes months. ### Does blocking GPTBot stop ChatGPT recommending me? No. GPTBot gathers data for training models. ChatGPT's search uses OAI-SearchBot, which is controlled separately. You can block GPTBot and still appear in ChatGPT search, according to OpenAI's documentation. ### Why does ChatGPT recommend my competitor and not me? Usually because their website answers the question more directly, because they are mentioned on the sites ChatGPT cites, or because something stops ChatGPT reading yours. Ask the question, open the sources it cites, and compare those pages with your own. ### Does ChatGPT use Google or Bing? OpenAI says ChatGPT search uses third-party search providers and names Bing. Independent testing has reported Google results being used at times as well. OpenAI does not publish the mix, so it is safest to be properly indexed in both. ## Sources - OpenAI, [Overview of OpenAI crawlers](https://developers.openai.com/api/docs/bots). Checked 10 November 2026. - OpenAI, [Publishers and developers FAQ](https://help.openai.com/en/articles/12627856-publishers-and-developers-faq). Checked 10 November 2026. - Ahrefs, [Is ChatGPT really powered by Google?](https://ahrefs.com/blog/is-chatgpt-really-powered-by-google/), an outside analysis of the searches ChatGPT runs. Not confirmed by OpenAI. - Microsoft Advertising, [Optimizing your content for inclusion in AI search answers](https://about.ads.microsoft.com/en/blog/post/october-2025/optimizing-your-content-for-inclusion-in-ai-search-answers), 8 October 2025. - Google Search Central, [Optimizing your website for generative AI features](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Checked 10 November 2026. - My own tests, 10 November 2026: [the case study](https://agentseohq.com/case-study). --- # What is llms.txt, and does it actually work? > What llms.txt is, how to write one, and what the evidence says about whether ChatGPT, Claude or Google read it. An honest guide with a working example. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/what-is-llms-txt **The short answer:** `llms.txt` is a Markdown file at the root of your website that gives AI agents a short, curated guide to your site. It is a proposal, not a standard any AI company has committed to. Google says its search does not use it, and a 2026 Ahrefs study found 97% of these files received no requests in a month. It takes ten minutes to write and does no harm, so add one, but do it last. ## What llms.txt is `llms.txt` is a plain text file, written in Markdown, that lives at `https://yourdomain.com/llms.txt`. It tells an AI agent what your site is and where the useful pages are, without the navigation, adverts and scripts that surround them in HTML. It was proposed by Jeremy Howard in September 2024, and the specification at llmstxt.org was revised in August 2026. The proposal says the file is mainly meant for use at "inference" time: when an agent is helping someone and wants to understand a site quickly, as opposed to when a model is being trained. A common mix-up: llms.txt is not robots.txt for AI. It grants and denies nothing. | Aspect | robots.txt | llms.txt | |---|---|---| | Purpose | Says which crawlers may fetch which pages | Describes the site and points to its best pages | | Format | Its own simple rule format | Markdown | | Status | An internet standard (RFC 9309) | A community proposal | | Followed by AI companies? | Yes, by the major crawlers, with exceptions for user-triggered fetches | No public commitment from any of them | To control access, use robots.txt. [That guide is here](https://agentseohq.com/articles/ai-crawlers-robots-txt). ## The format The specification is short. In order: 1. An `# H1` with the name of the site or project. This is the only required part. 2. A blockquote with a short summary. 3. Optional paragraphs or lists with more background. No headings in this part. 4. `## H2` sections containing lists of links, each written as `[name](url)`, optionally followed by a colon and a note. 5. By convention, a section called `## Optional` for links an agent can skip if it is short of space. Here is a complete example for an imaginary accountancy firm: ```markdown # Harbour Accounts > Harbour Accounts is an accountancy firm in Leeds for small design > studios and freelancers. Fixed monthly fees from £95 plus VAT. Key facts: - Services: bookkeeping, year-end accounts, VAT returns, payroll. - Area: Leeds and West Yorkshire in person, the rest of the UK remotely. - Not offered: audits, and personal tax for non-clients. - Contact: hello@harbouraccounts.example or 0113 496 0000. ## Pages - [Services and prices](https://harbouraccounts.example/services): what each monthly plan includes - [About the team](https://harbouraccounts.example/about): who we are and our qualifications - [Contact](https://harbouraccounts.example/contact): how to book a first call ## Guides - [VAT for freelance designers](https://harbouraccounts.example/guides/vat-for-designers): when to register and what to charge ## Optional - [Privacy policy](https://harbouraccounts.example/privacy) ``` The specification also suggests offering a Markdown copy of each important page at the same address with `.md` added. Many sites publish an `llms-full.txt` as well: the full text of the site in one file. That name is a widespread convention, not part of the specification itself. ## Does anything actually read it? This is where an honest article parts company with most of the ones you will find. Here is the evidence as of November 2026. **Google says no.** Its guidance on generative AI features states: "You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search." Making one for other services is fine, it adds, but it will not affect your visibility in Search. **OpenAI and Anthropic have not said yes.** Neither company's crawler documentation mentions llms.txt. Anthropic publishes an llms.txt for its own documentation, which shows it finds the format useful for developers, not that Claude's search consults yours. **The largest study found very little traffic.** In June 2026 Ahrefs looked at 137,210 websites. About 28% had an llms.txt. Of those files, 97% received no requests at all during May 2026. Where files were fetched, 96% of requests came from bots, but AI bots were only about a fifth of that. SEO audit tools fetched llms.txt more often than any single AI crawler. The authors called the file "largely decoration" for AI search today. One caveat they give themselves: their sample skews towards SEO-aware sites, so real adoption is probably lower than 28%. **Coding tools are the exception.** Developer tools and coding agents do look for llms.txt and llms-full.txt when pointed at documentation. In the Ahrefs data, Anthropic's Claude Code was among the more active fetchers. If you publish software documentation, the file earns its keep. My own experience fits. I have added llms.txt to 11 of my own sites. I cannot show you that it changed a single answer, and I will not pretend otherwise. ## So why add one? - **It is cheap.** Ten minutes for a small site. - **It is harmless.** Google says it has no effect on Search, in either direction. - **Agents are changing fast.** A file that is ignored today may be read next year, and coding agents already read it. - **Writing it is useful in itself.** If you cannot describe your business in a blockquote and six bullet points, your home page probably cannot either. The risk is not the file. The risk is paying someone for the file and thinking the job is done. If an agency leads with llms.txt, ask what else is on the list. ## How to write a good one 1. **Start with facts an assistant would need to answer a question about you**: what you sell, who for, where, prices, how to buy, what you do not do. 2. **Keep it true.** Copy prices and facts from the live site. A stale llms.txt is worse than none, because it is a clean, confident source of wrong information. 3. **Link to your best pages, not all of them.** It is a guide, not a sitemap. 4. **Generate it if you can.** On my sites, llms.txt is built from the same content as the pages, so the two cannot drift apart. This site's [llms.txt](https://agentseohq.com/llms.txt) is made that way. 5. **Serve it properly.** It should return status 200 with a `text/plain` or `text/markdown` content type. ## One mistake to check for today Open `yourdomain.com/llms.txt` in a browser. If you see your home page, or a "not found" page that looks like a normal page, your server is answering "yes, here it is" to a file that does not exist. Three of my own sites did this. A machine has to read the whole response to discover it has been handed the wrong thing. Either publish a real file or make the address return a proper 404. The [free checker](https://agentseohq.com/check) tests for this. **Want to know how AI assistants see your website?** Send me your address and I will check it by hand: what the assistants can read, what they say about you, and what to fix first. It is free, and you get a fixed quote before any paid work. [Get a free Agent SEO check](https://agentseohq.com/contact) or [try the instant checker](https://agentseohq.com/check). ## Frequently asked questions ### Is llms.txt an official standard? No. It is a community proposal published at llmstxt.org by Jeremy Howard in September 2024. It has no standards body behind it, and no major AI company has publicly committed to reading it for search answers. ### Does Google use llms.txt? No. Google's guidance says you do not need to create machine-readable AI text files to appear in Google Search, including its AI features, and that doing so will not affect your visibility there. ### Does ChatGPT read llms.txt? OpenAI has not said that ChatGPT's search uses it, and its crawler documentation does not mention the file. Ahrefs' 2026 study did record some requests for llms.txt from OpenAI's crawlers, but they were a small share of a small total. ### What is the difference between llms.txt and llms-full.txt? llms.txt is a short guide with links to your most useful pages. llms-full.txt is a common companion file containing the full text of the site in one Markdown document. Only llms.txt is defined in the proposal. ### Will llms.txt hurt my SEO? No. Google says the file has no effect on Search. The only real risk is letting it go out of date, so that it states old prices or facts. ## Sources - [llmstxt.org](https://llmstxt.org/), the llms.txt proposal. Published 3 September 2024, revised 10 August 2026. - Google Search Central, [Optimizing your website for generative AI features on Google Search](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Checked 10 November 2026. - Ahrefs, [We analyzed 137K sites: 97% of llms.txt files never get read](https://ahrefs.com/blog/llmstxt-study/), Louise Linehan, 15 June 2026. - OpenAI, [Overview of OpenAI crawlers](https://developers.openai.com/api/docs/bots). Checked 10 November 2026. - Anthropic, [Does Anthropic crawl data from the web, and how can site owners block the crawler?](https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler). Checked 10 November 2026. --- # AI crawlers and robots.txt: GPTBot, ClaudeBot, PerplexityBot and the rest > A reference to the AI crawlers that visit your site (GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended) and the robots.txt rules for each. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/ai-crawlers-robots-txt **The short answer:** AI companies run three kinds of crawler: ones that gather **training** data, ones that build a **search** index, and ones that fetch a page because a **user** asked. Blocking the training crawlers (GPTBot, ClaudeBot, Google-Extended) does not remove you from AI answers. Blocking the search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot) does. Many sites block all of them without meaning to. ## Three kinds of crawler Before the list, the distinction that makes sense of it. - **Training crawlers** collect pages that may be used to train future models. Whether to allow them is a choice about your content, and reasonable people choose differently. - **Search crawlers** build the index an assistant searches when it answers a question. Block these and you drop out of that assistant's answers. - **User-triggered fetchers** visit a page because a person asked the assistant about it, or pasted in a link. They act more like a browser than a crawler, and some do not follow robots.txt for that reason. ## The reference table The names in the second column are what you write after `User-agent:` in robots.txt. | Company | Robots.txt name | Kind | What blocking it does | |---|---|---|---| | OpenAI | `OAI-SearchBot` | Search | Removes you from ChatGPT search answers | | OpenAI | `ChatGPT-User` | User-triggered | OpenAI says robots.txt "may not apply" to these visits | | OpenAI | `GPTBot` | Training | Keeps your content out of model training | | Anthropic | `Claude-SearchBot` | Search | May reduce your visibility in Claude's search results | | Anthropic | `Claude-User` | User-triggered | Stops Claude fetching your pages when a user asks | | Anthropic | `ClaudeBot` | Training | Keeps future content out of training data | | Perplexity | `PerplexityBot` | Search | Removes you from Perplexity's search results | | Perplexity | `Perplexity-User` | User-triggered | Little: Perplexity says it "generally ignores robots.txt" | | Google | `Googlebot` | Search | Removes you from Google Search, AI Overviews and AI Mode | | Google | `Google-Extended` | Training and grounding control | Opts out of Gemini training and grounding; no effect on Search | | Microsoft | `Bingbot` | Search | Removes you from Bing and Copilot | | Apple | `Applebot` | Search | Removes you from Siri and Spotlight results | | Apple | `Applebot-Extended` | Training control | Opts out of Apple's AI training; search is unaffected | | Meta | `meta-externalagent` | Training | Opts out of Meta's AI crawling | | Common Crawl | `CCBot` | Open archive | Keeps you out of a public dataset widely used for training | | Amazon | `Amazonbot` | Search and AI | Affects Alexa answers | | DuckDuckGo | `DuckAssistBot` | Search | Affects DuckAssist answers | Some details worth knowing: - **Google-Extended and Applebot-Extended are not crawlers.** They are names you can use in robots.txt to control how pages already fetched by Googlebot or Applebot are used. Google states that Google-Extended "does not impact a site's inclusion in Google Search". - **There is no separate crawler for AI Overviews.** They are part of Google Search. The only way to stay out of them is the ordinary Search controls, such as `nosnippet`. - **Old Anthropic names are retired.** Rules for `anthropic-ai` or `Claude-Web` no longer do anything, according to coverage of Anthropic's 2026 documentation update. If your robots.txt has them, it is out of date. - **OpenAI says robots.txt changes take about 24 hours** to apply to search. - **Anthropic honours `Crawl-delay`**, a non-standard line for slowing a crawler down. This list changes several times a year. I checked the OpenAI, Anthropic, Perplexity, Google and Apple rows against those companies' own documentation on 10 November 2026. The other rows reflect what those operators have said about their crawlers and I hold them with less confidence. Check again before relying on any of it. ## Welcome everyone: a robots.txt to copy If you want to be found and recommended, this is the simplest file that says so. It names the crawlers so there is no doubt, keeps private areas private, and points to your sitemap. ```text # AI assistants and search engines are welcome. User-agent: OAI-SearchBot User-agent: ChatGPT-User User-agent: GPTBot User-agent: Claude-SearchBot User-agent: Claude-User User-agent: ClaudeBot User-agent: PerplexityBot User-agent: Perplexity-User User-agent: Google-Extended User-agent: Applebot User-agent: Applebot-Extended User-agent: Bingbot Allow: / Disallow: /admin/ Disallow: /basket/ User-agent: * Allow: / Disallow: /admin/ Disallow: /basket/ Sitemap: https://yourdomain.com/sitemap.xml ``` Strictly, `User-agent: *` with `Allow: /` already lets everyone in. Naming the crawlers is belt and braces: it makes your intent obvious to anyone reading the file, and it stops a later "block everything unknown" rule catching them by accident. Remember that a crawler which finds its own name uses only that group, so repeat your `Disallow` lines there. ## Found in AI search, but not used for training A common and perfectly sensible position. Allow the search and user crawlers, block the training ones: ```text # Search and assistants: yes. User-agent: OAI-SearchBot User-agent: ChatGPT-User User-agent: Claude-SearchBot User-agent: Claude-User User-agent: PerplexityBot User-agent: Bingbot User-agent: Googlebot User-agent: Applebot Allow: / # Training: no. User-agent: GPTBot User-agent: ClaudeBot User-agent: Google-Extended User-agent: Applebot-Extended User-agent: meta-externalagent User-agent: CCBot Disallow: / User-agent: * Allow: / Sitemap: https://yourdomain.com/sitemap.xml ``` One trade-off to weigh. A model that never trained on your content knows less about you "from memory", and relies entirely on live search to mention you. For most small businesses that is already the case, so the cost is small. ## The blocks you did not set Robots.txt is only a request, and it is not the only thing between a crawler and your pages. These are the usual culprits when a site blocks AI crawlers by accident. - **Your CDN or security service.** Cloudflare announced in July 2025 that it would block AI crawlers by default for new customers. On two of my own domains I found it adding lines to robots.txt that disallowed GPTBot, ClaudeBot and others, although my own file welcomed them. It took several separate settings to stop it. Other providers have similar features. - **Bot protection.** Features that challenge "non-browser" traffic cannot tell a helpful crawler from a nuisance. Perplexity's documentation specifically warns that Cloudflare's Bot Fight Mode can still challenge its bots whatever your other rules say. - **Security plugins.** Some WordPress plugins ship with lists of "bad bots" that include AI crawlers. - **Hosting defaults.** Some hosts apply their own rules above yours. - **Rate limits.** A crawler that fetches quickly can trip a limit and be shut out for hours. The test is what the crawler actually receives, not what your robots.txt says. Server logs will show it: look for the crawler names above and the status codes next to them. Lots of 403 or 429 against a crawler you meant to allow is your answer. ## Do the crawlers obey? Mostly, with documented exceptions. - OpenAI, Anthropic, Google, Microsoft, Apple and Perplexity all say their automated crawlers follow robots.txt. - User-triggered fetchers are the exception. OpenAI says of ChatGPT-User that robots.txt "may not apply". Perplexity says Perplexity-User "generally ignores" it. Anthropic's general statement that its bots honour robots.txt does not carve out Claude-User. - Not everyone on the internet is as careful. Robots.txt is a convention, not a lock. If content must stay private, put it behind a login. ## A five-minute check 1. Open `yourdomain.com/robots.txt`. Look for any `Disallow: /` and note which `User-agent` it sits under. 2. Look for names you did not add. Lines you do not recognise usually came from a plugin, host or CDN. 3. Check your CDN's bot and AI settings, not just the file. 4. Run the [free checker](https://agentseohq.com/check), which reads your robots.txt and reports on each crawler in the table. 5. If you have access to server logs, search them for `OAI-SearchBot`, `ClaudeBot` and `PerplexityBot` and look at the status codes. **Not sure what is blocking what?** Firewalls and hosting rules do not show up in robots.txt. Send me your address and I will test what the AI crawlers actually receive from your site. It is free. [Get a free Agent SEO check](https://agentseohq.com/contact) or [try the instant checker](https://agentseohq.com/check). ## Frequently asked questions ### Should I block GPTBot? It depends what you want. GPTBot gathers content for training OpenAI's models. Blocking it keeps your content out of training and does not remove you from ChatGPT search, which uses OAI-SearchBot. If your content is your product, blocking is reasonable. If you want to be widely known, allow it. ### Does blocking Google-Extended affect my Google rankings? No. Google states that Google-Extended does not affect a site's inclusion in Google Search and is not a ranking signal. It controls whether your content is used for Gemini training and grounding. ### How do I allow ChatGPT to read my website? Make sure robots.txt does not disallow OAI-SearchBot or ChatGPT-User, and that no firewall, security plugin or CDN setting blocks them. OpenAI also recommends allowing requests from its published IP ranges. ### Can I stop my site appearing in Google's AI Overviews? Not with a crawler rule. AI Overviews are part of Google Search and use Googlebot. Google's controls are the standard ones: `nosnippet`, `data-nosnippet`, `max-snippet` or `noindex`, each of which also affects your ordinary search listing. ### Why is my robots.txt blocking AI bots when I never added that? A plugin, your host or your CDN probably did. Cloudflare, for instance, can manage robots.txt and block AI crawlers for you. Compare the file your server holds with the one visitors receive, and check your provider's AI and bot settings. ## Sources - OpenAI, [Overview of OpenAI crawlers](https://developers.openai.com/api/docs/bots). Checked 10 November 2026. - Anthropic, [Does Anthropic crawl data from the web, and how can site owners block the crawler?](https://support.claude.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler). Checked 10 November 2026. - Perplexity, [Perplexity crawlers](https://docs.perplexity.ai/guides/bots). Checked 10 November 2026. - Google Search Central, [Google's common crawlers](https://developers.google.com/search/docs/crawling-indexing/google-common-crawlers) and [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features). Checked 10 November 2026. - Apple, [About Applebot](https://support.apple.com/en-us/119829). Checked 10 November 2026. - Search Engine Journal, [Anthropic's Claude bots make robots.txt decisions more granular](https://www.searchenginejournal.com/anthropics-claude-bots-make-robots-txt-decisions-more-granular/568253/), 2026. - Search Engine Land, [Cloudflare to block AI crawlers by default](https://searchengineland.com/cloudflare-to-block-ai-crawlers-by-default-with-new-pay-per-crawl-initiative-457708), July 2025. - RFC 9309, [Robots Exclusion Protocol](https://www.rfc-editor.org/rfc/rfc9309.html). --- # Schema markup for AI search: what to add and what to skip > Which schema markup helps AI assistants understand your business, with copy-and-paste JSON-LD for organisations, products and FAQs, and what to skip. By Reece Chown, Agent SEO HQ. Published 11 November 2026. Canonical URL: https://agentseohq.com/articles/schema-markup-for-ai-search **The short answer:** Schema markup (structured data, usually written as JSON-LD) states your business facts in a format made for machines. No AI company says it is required, and Google says no special schema is needed for its AI features. It is still worth adding four types: who you are, what you sell, your questions and answers, and breadcrumbs. Keep it accurate and identical to what the page shows. ## What schema markup is Schema markup is a block of code in a page that describes it in a shared vocabulary from schema.org. Where your page says "Open Monday to Friday, 9 to 5", the markup says so in a form no machine can misread. The usual format is JSON-LD: a `