# See where your brand shows up in AI answers > Track whether ChatGPT, Claude, Perplexity and Google AI Overviews mention and cite your brand or your competitors. Every number links to the original answer. **AI visibility tracking for founders.** Find out whether ChatGPT, Claude, Perplexity and Google AI Overviews recommend you or your competitors when buyers ask for a product like yours. Every number links to the answer it came from. Runs in your own Chrome, signed in to your own AI accounts. Checks run while Chrome is open. [Track your first brand](https://aeometer.com/sign-up) or [see how it works](https://aeometer.com/how-it-works). Tracks answers from: [ChatGPT](https://aeometer.com/engines/chatgpt), [Claude](https://aeometer.com/engines/claude), [Perplexity](https://aeometer.com/engines/perplexity), [Google AI Overviews](https://aeometer.com/engines/google-ai-overviews). ## Your buyers ask AI before they google you A founder looking for a tool types the question into ChatGPT and reads a shortlist of three names. If you are left off, you never find out. - **The shortlist forms in a chat window.** Prompts like "best product analytics tool for a seed-stage startup" get a direct answer with a few named products. If you are not one of them, you are not on the shortlist. - **Nobody sends you a report.** ChatGPT does not email you when it recommends a competitor. Asking yourself tells you about one answer, on one day, in one account, and you forget it by next week. - **Every engine answers differently.** ChatGPT might name you while Perplexity cites a competitor's comparison page and Google's AI Overview skips you entirely. You need all four side by side. ## Every number traces back to a real answer AEOmeter saves the text each engine gave and the sources it cited. Your brand and your competitors are highlighted, so when a number moves you can read why. - Mentions are exact, case-insensitive matches on the names and aliases you set. A brand counts once per answer. - Sources the engine attributes to its answer are kept apart from links that only appear in the text. - Failed, blocked and "no AI Overview" checks show up as exactly that. They never count as a miss. ## From sign-up to your first answers in three steps 1. **Add your brand** Your brand, its domains and a few competitors. AEOmeter suggests starter prompts and queues a first ChatGPT run. 2. **Connect the Chrome extension** It sends each prompt to ChatGPT, Claude, Perplexity and Google in background tabs, signed in as you. Set a daily schedule in your timezone. 3. **Read what changed** Mention rate, citations and share of voice per engine, with the original answer one click behind every number. [Read the full walkthrough](https://aeometer.com/how-it-works). ## Three numbers, with the math in plain sight Each rate shows its denominator. When there are no answers yet you see a dash, not 0%. **Mention rate (shown as Visibility)**: Answers that mention your brand, divided by answers captured successfully. Example: 18 of 43 answers mention Acme Analytics → 42%. **Owned citation rate (shown as Cited)**: Answers that cite at least one page on your own domain or subdomains, divided by answers captured successfully. Example: 8 of 43 answers cite acme.io → 19%. **Tracked share of voice (against the competitors you track)**: Your mentions divided by mentions of every brand you track, in the same answers. It compares you with the brands you chose, so it is not market share. Example: 18 of 67 tracked mentions → 27%. [How every metric is calculated](https://aeometer.com/methodology). ## See the prompts where competitors show up and you don't Gaps lists answers where a competitor is mentioned and you are not, with the answer attached. Start with the prompts that hurt most and export the comparison to CSV when you want it in a spreadsheet. See [competitor tracking](https://aeometer.com/features/competitor-tracking). ## Ask your AI about your AI visibility Connect the AEOmeter MCP server to Claude, Cursor or any client that supports remote MCP with OAuth. Your assistant can read answers and analytics, add prompts and check on your browser using the same actions as the web app. The server lives at `https://aeometer.com/mcp`. Questions you can ask: - "Which prompts mention Mixpanel but not us on Perplexity?" - "Add these 15 prompts and tag them "pricing"." - "Is my browser connected and ready to run checks?" Prefer scripts? The REST API has an [OpenAPI spec](https://aeometer.com/openapi.json) and a [reference](https://aeometer.com/docs). Details: [MCP and API](https://aeometer.com/features/mcp). ## What you should know before you sign up AEOmeter makes a few deliberate choices. Here they are up front. - **Checks run only while your Chrome is open.** A schedule queues checks. Your browser runs them when it is open, connected and signed in to each engine. The app shows the next scheduled time and the last real capture separately, so a quiet week looks quiet. - **Four engines today.** ChatGPT, Claude, Perplexity and Google AI Overviews. Engines change their pages often, so our adapters are experimental and sometimes need a fix. - **CAPTCHAs stay with you.** If an engine asks you to sign in or verify you are human, the check pauses and the extension shows a red badge. AEOmeter never tries to get around it. - **Your sample, not the whole internet.** Results show what each engine told your browser, in private mode where available. That is close to what a buyer sees, but it does not represent every user. - **We measure and stop there.** AEOmeter tracks mentions and citations. It does not score sentiment, report ranking positions or write content for you. - **The extension is in beta.** It is not yet listed in Chrome Web Store search. After you sign up, the app walks you through installing it. ## Frequently asked questions ### How do I install AEOmeter? Create an account and add your brand. The app then walks you through installing the AEOmeter BETA Chrome extension and connecting it. The extension is not yet searchable in the Chrome Web Store, so start from [sign-up](https://aeometer.com/sign-up). ### Do you need my ChatGPT or Claude password? No. You sign in to each engine in your own Chrome, the way you already do. The extension works in those tabs and never captures your cookies or account tokens. ### Does Chrome need to be open? Yes. Checks run in your browser, so they run while Chrome is open and connected. If your laptop is closed all week, nothing is collected that week, and the app says so instead of showing zeros. ### Will it fill my chat history with test prompts? The extension turns on each engine's private mode where it can: temporary chat in ChatGPT, incognito in Claude and Perplexity, and non-personalized results on Google. This is best effort, because engines change these settings. ### What data do you store? Your brands, prompts and schedules, plus the latest answer text and cited sources for each check. We do not save the page itself, your cookies or your tokens. Site analytics load only after you consent and exclude prompts, answers and brand names. See the [privacy policy](https://aeometer.com/privacy). ### Which AI engines does it track? Four: [ChatGPT](https://aeometer.com/engines/chatgpt), [Claude](https://aeometer.com/engines/claude), [Perplexity](https://aeometer.com/engines/perplexity) and [Google AI Overviews](https://aeometer.com/engines/google-ai-overviews). You choose which ones each prompt goes to. ### Why does Chrome say an extension "started debugging this browser"? The extension uses Chrome's debugger API to type into each engine's own chat box like a person would. Chrome shows that banner whenever an extension uses this API. ### How much does it cost? Free for one brand. Founder is $19.99 a month for 5 brands and Studio is $49.99 a month for 20, with unlimited prompts on every plan. See the [pricing page](https://aeometer.com/pricing). ## Find out what AI tells your buyers about you. Add your brand, a few competitors and the questions your buyers ask. Your first answers arrive as soon as the extension is connected. [Track your first brand](https://aeometer.com/sign-up). Needs Chrome and your own AI accounts. --- - Web page: https://aeometer.com/ - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AI visibility tracking you can check line by line > A tour of AEOmeter, the LLM visibility tool for founders: brand setup, prompt tracking, saved answers, cited sources and browser monitoring across four AI engines. AEOmeter tracks how ChatGPT, Claude, Perplexity and Google AI Overviews answer the questions your buyers ask, and whether they name you or a competitor. Here is every screen, with example data for a fictional brand. Screens: Setup, Overview, Prompts, Answers, Sources, Monitoring, plus MCP and API access. ## Setup: Start with one brand and the rivals you lose deals to Add your brand, the names people actually use for it and the domains you own. Then add a few competitors. Aliases decide what counts as a mention, and domains decide what counts as your own citation. - One own brand plus competitors, each with aliases and owned domains. - Starter prompts are suggested as soon as you add a brand, so you are not staring at an empty list. - A first ChatGPT run is queued for you, ready for when the extension connects. More: [Choosing prompts](https://aeometer.com/guides/choosing-prompts). ## Overview: Where you stand, in one screen Visibility (mention rate), Cited (owned citation rate) and a competitor comparison for the period and engines you pick. Each number carries its denominator, like "18 of 43 answers". - Gaps lists answers where a competitor is mentioned and you are not. - Failed, blocked and "no AI Overview" checks are left out of the rates, never counted as misses. - Extension status and the time of the last real capture sit at the top, so stale data is obvious. More: [Competitor tracking](https://aeometer.com/features/competitor-tracking). ## Prompts: The questions your buyers ask, tracked over time Each prompt has its own history across engines, so you can see the day ChatGPT started naming you or Perplexity stopped. Bring your own list or start from the suggestions. - Import from CSV or plain text with a preview before anything is saved. - Tags, pause and bulk actions for keeping 20 or 200 prompts tidy. - Editing a prompt starts a new baseline, so trends compare like with like. More: [Prompt tracking](https://aeometer.com/features/prompt-tracking). ## Answers: The original answer behind every number Open any check and read what the engine said, word for word, with your brand and competitors highlighted and the sources it cited listed underneath. - Mentions are deterministic, case-insensitive alias matches on word boundaries. - Sources the engine cites are kept apart from links that only appear in the text. - AEOmeter keeps the latest answer and its sources for each check, not a copy of the page. More: [Methodology](https://aeometer.com/methodology). ## Sources: The pages AI leans on in your category Group every cited source by domain or by page and see which sites keep showing up next to your competitors. Your own domains are marked, so you can see how often AI cites you directly. - Group by domain for the big picture, by page for the exact URL. - Own, competitor and third-party sources are labelled. - When source extraction was incomplete for an answer, the app says so. More: [Citation tracking](https://aeometer.com/features/citation-tracking). ## Monitoring: Know whether your browser can collect Checks run in your Chrome, so Monitoring shows whether the extension is connected and which engine sessions are ready. When a check pauses for a login or verification, you see which one and why. - Daily schedules run in the timezone you choose. - The next scheduled time and the last actual capture are shown separately. - Pause, resume or skip from the extension's toolbar popup. More: [How it works](https://aeometer.com/how-it-works). ## MCP and API: Your data in Claude, Cursor and your own scripts The AEOmeter MCP server at `https://aeometer.com/mcp` gives AI assistants the same actions as the web app: read answers and analytics, import prompts, manage schedules and check browser status. It signs in with OAuth. A REST API with an [OpenAPI spec](https://aeometer.com/openapi.json) covers the rest. More: [MCP and API](https://aeometer.com/features/mcp). ## Questions ### Which AI engines does AEOmeter track? ChatGPT, Claude, Perplexity and Google AI Overviews. You pick which of the four each prompt goes to. See [all engines](https://aeometer.com/engines). ### What is the difference between a mention and a citation? A mention is your brand name (or an alias) appearing in the answer text. A citation is a source the engine attributes to its answer. AEOmeter counts them separately, and only citations to your own domains count toward your owned citation rate. ### Can I import the prompts I already have? Yes. Paste plain text or upload a CSV, check the preview, then save. You can tag, pause and bulk edit prompts afterwards. ### Does AEOmeter use the ChatGPT API? No. The Chrome extension asks the consumer apps your buyers use, signed in with your own accounts. [How it works](https://aeometer.com/how-it-works) explains the trade-offs of that choice. ### Can I export the data? Competitor comparisons export to CSV. Prompts, answers and analytics are also available through the [MCP server and REST API](https://aeometer.com/features/mcp). --- - Web page: https://aeometer.com/product - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # How AEOmeter collects AI answers in your browser > How the AEOmeter Chrome extension sends your prompts to ChatGPT, Claude, Perplexity and Google, reads the finished answers and reports them, and what pauses a check. AEOmeter asks ChatGPT, Claude, Perplexity and Google your prompts from your own Chrome, signed in as you, then reads the finished answers and turns them into reports. Here is every step, including what can stop a check. ## Step by step 1. **Set up in the web app** Add your brand, competitors and prompts, pick the engines for each prompt and set a daily schedule in your timezone. A schedule queues checks. It does not run them by itself. 2. **The extension connects** Install the AEOmeter BETA Chrome extension and click Connect. Sign in to ChatGPT, Claude and Perplexity in that same Chrome, as you normally would. Monitoring shows which engine sessions are ready. 3. **Checks run in background tabs** Queued checks open in background tabs grouped as "AEOmeter". Each prompt goes to all its selected engines at once, and the next prompt starts when all of them finish. Chrome shows a "started debugging this browser" banner while this happens. 4. **Private mode where available** Before asking, the extension turns on each engine's private mode: temporary chat in ChatGPT, incognito in Claude and Perplexity, and non-personalized results on Google. This is best effort, because engines move these settings around. 5. **Real input, finished answers** The prompt is typed into the engine's own chat box, the same way you would. The extension waits until the engine shows the answer is finished, up to 10 minutes, then reads it as Markdown with its cited sources. For Perplexity, sources come from the thread data. 6. **Upload the answer, nothing else** Only the answer text, its sources and run diagnostics are uploaded. The page itself, your cookies and your account tokens stay in your browser. 7. **Reports with the math shown** Mentions are matched against your aliases and citations against your domains. Visibility, owned citation rate and share of voice are calculated with explicit denominators, and every number links to its answer. ## What pauses a check Sometimes an engine wants a human. When that happens, AEOmeter stops and tells you instead of guessing. - **Login wall.** You were signed out of an engine. Sign back in in that Chrome and resume. - **Verification or CAPTCHA.** The engine asked you to prove you are human. You solve it yourself. AEOmeter never bypasses a CAPTCHA. - **Rate limit.** The engine asked you to slow down. The check waits until you resume it. A paused check shows a red "!" badge on the extension icon. Open the toolbar popup to pause, resume or skip. A prompt is never sent twice, and a failed or skipped check never counts as a miss in your numbers. ## Why collect from a browser - **The answers your buyers see.** Your buyers use the ChatGPT, Claude and Perplexity apps and Google search, not a developer API. AEOmeter reads answers from those same consumer apps. - **Signed in like a real person.** Checks run in a normal Chrome session with your own accounts, the way a buyer would ask. - **Low collection cost.** The work happens on your machine, so collecting answers costs very little to run. Other tools also collect from web interfaces. What AEOmeter adds is that it runs in your own browser and shows you the raw evidence. ## The trade-off: your Chrome has to be open Checks run only while Chrome is open, connected and signed in to each engine. If your laptop stays closed for a week, nothing is collected that week. AEOmeter shows the next scheduled time and the last actual capture side by side, so a gap in your data is visible instead of hidden behind zeros. ## Privacy - The extension never captures provider cookies or account tokens. - It uploads the latest answer and its sources, not the page. - Your prompts go only to the AI engines you selected for them. - Site analytics load only after you consent and exclude prompts, answers and brand identifiers. Read the [privacy policy](https://aeometer.com/privacy). ## Questions ### Can I keep working while checks run? Yes. Checks run in background tabs in their own "AEOmeter" tab group. You can use Chrome as normal, though closing those tabs or signing out of an engine can interrupt a check. ### Which browser do I need? Google Chrome on a computer. The extension is called AEOmeter BETA and is not yet searchable in the Chrome Web Store, so the app walks you through installing it after [sign-up](https://aeometer.com/sign-up). ### How long does a check take? As long as the engine takes to finish its answer. The extension waits for each engine's finished state and gives up after 10 minutes. ### Will a prompt ever be sent twice? No. AEOmeter never sends the same prompt twice for one check. If the answer cannot be read, the check shows as failed and stays out of your numbers. ### Why does Chrome show a debugging banner? The extension uses Chrome's debugger API to type into each engine's chat box like a person would. Chrome shows the banner whenever an extension uses that API. ### What if an engine changes its page? Engines redesign often, so the adapters are experimental. When one breaks, checks for that engine fail visibly and we fix the adapter. See [troubleshooting](https://aeometer.com/help/troubleshooting). --- - Web page: https://aeometer.com/how-it-works - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Simple pricing: pay per brand, not per prompt > AEOmeter pricing: free for one brand, $19.99/month for 5 brands, $49.99/month for 20. Unlimited prompts and all four AI engines, Claude included, on every plan. Checks run in your own Chrome, on your own AI accounts, so each answer costs us almost nothing. That is why every plan gets unlimited prompts, any schedule and all four engines, Claude included. You only pay for how many brands you track. > **Billing.** Paid plans are billed in advance through Polar, monthly or yearly. Cancel any time from the Billing page in the app, and your plan stays active until the end of the period you paid for. ## Plans as of 2026-10-16 | | Free | Founder | Studio | | --- | --- | --- | --- | | Monthly | $0 | $19.99 | $49.99 | | Yearly (4 months free) | - | $159.92 | $399.92 | | Own brands tracked | 1 | 5 | 20 | | Competitors per brand | Unlimited | Unlimited | Unlimited | | Prompts | Unlimited | Unlimited | Unlimited | | Schedule | Any | Any | Any | | Engines: ChatGPT, Claude, Perplexity, Google AI Overviews | All four | All four | All four | | History | Last 30 days | Full | Full | | CSV export, REST API, MCP | - | Yes | Yes | **More than 20 brands?** Talk to us. Past 20 brands, one browser cannot keep up with daily checks, so we plan the setup with you. Email [hello@aeometer.com](mailto:hello@aeometer.com). ## The fine print, in plain words - **Your browser sets the pace.** Checks run while Chrome is open and your engine accounts are signed in. One browser comfortably handles about 50 prompts a day across all four engines. Larger schedules still work, and the app shows how much fits and spreads the rest over the next days. - **No paying for checks that never ran.** We do not sell checks, so a closed laptop costs you nothing. A check that did not run shows as missing, never as a brand miss. - **What counts as a brand.** A brand is one of your own products that you track. Competitors inside it are unlimited and free. - **History on Free.** Free shows your last 30 days. Older answers are kept, not deleted, and come back when you upgrade. - **Collection without the extension.** We plan to offer collection that runs on our servers, with no Chrome needed, as a paid extra. It is not available yet and has no price. ## Questions ### How much does AEOmeter cost? Free for one brand. Founder is $19.99 a month for 5 brands, and Studio is $49.99 a month for 20 brands. Yearly billing gives you 4 months free. ### Why are prompts unlimited? Checks run in your Chrome, on your own ChatGPT, Claude and Perplexity accounts, so an extra prompt costs us almost nothing. The real limit is how many checks your browser and your engine accounts can run in a day, and the app shows you that. ### Do I need a credit card to sign up? No. Sign-up does not ask for payment details, and the Free plan never needs one. ### How do I cancel? Open Billing in the app and choose Manage subscription. Your plan stays active until the end of the period you paid for, then the organization moves to Free. Your data is kept. ### Is Claude extra? No. All four engines are included on every plan, Free too. ### What do I need to run AEOmeter? Google Chrome with the AEOmeter extension, and your own accounts for the engines you track. Free engine accounts work, but their message limits cap how many checks a day you can run. ### What if I track more than 20 brands? Email us at [hello@aeometer.com](mailto:hello@aeometer.com). Past 20 brands, one browser cannot keep up, so we work out the schedule and setup with you. ### Is there anything free I can use without an account? Yes. The [free AI visibility checker](https://aeometer.com/tools/ai-visibility-checker) helps you check a few prompts by hand. [Track your first brand for free](https://aeometer.com/sign-up). --- - Web page: https://aeometer.com/pricing - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Track the exact questions your buyers ask AI > Track the prompts your buyers ask ChatGPT, Claude, Perplexity and Google AI Overviews. Import a list, tag it, and read every answer per prompt. AEOmeter sends the prompts you choose to ChatGPT, Claude, Perplexity and Google AI Overviews and keeps every answer. Each prompt gets its own history, so you can see which engines mention you for "best product analytics tool for a seed-stage startup" and which ones leave you out. Think of it as an LLM visibility tracker organized around the questions themselves. Every number traces back to an answer you can read. ## Start with the questions that sell When you add a brand, AEOmeter suggests starter prompts and queues a first ChatGPT run, so answers start arriving as soon as your extension is connected. Claude Haiku writes the suggestions when it is available. Otherwise they come from templates. Keep the ones that sound like your buyers and delete the rest. Good prompts read like something a buyer types late at night: "Mixpanel alternatives with a generous free tier" or "which analytics tool should a two-person SaaS team use". Leave your brand name out of most of them. You want to know whether AI brings you up on its own. More on this in [how to choose prompts](https://aeometer.com/guides/choosing-prompts). ## Import a whole list in one paste Paste one prompt per line, or upload a CSV with a `text` column or a plain .txt file. Before anything is saved, a preview marks each row as ready, already added or invalid. Tags you type during import apply to the whole batch. - CSV with a `text` column, or one prompt per line - Duplicates of prompts you already track are flagged and skipped - Tags for the batch, such as `pricing` or `alternatives` ## Tags, pauses and bulk edits Tag prompts by topic or buying stage and filter the Prompts view by tag. Turn a prompt off to stop checking it without losing its history. Select a batch to turn prompts on or off or to add tags in one go. ## One prompt, every engine, over time The Prompts view shows the latest result for each prompt on every engine: mentioned, not mentioned, no AI Overview shown, or no answer yet. Open a prompt to see its visibility, how often answers cite your site, which competitors showed up, and the full answer history by date and engine. Every row opens the original answer text and the sources the engine cited. If a number looks wrong, you can read the answers behind it in two clicks. ## Edit a prompt and get a fresh baseline Rewording a prompt changes the question, so AEOmeter starts a new baseline when you edit one. Old answers stay saved. Trends compare answers to the same wording, which keeps a rewrite from showing up as a jump or a drop in your visibility. ## Where the answers come from Checks run in your own Chrome through the AEOmeter extension, signed in to your own AI accounts, using each engine's private mode where it has one. Each prompt goes to its selected engines at the same time, and the next prompt starts when all of them finish. Daily schedules queue checks in your timezone. They run when Chrome is open and connected, so the app shows the next scheduled time and the last actual capture separately. [How it works](https://aeometer.com/how-it-works) covers the details and the trade-offs. ## Questions ### How many prompts should I track? Start with 10 to 20 questions a buyer would really ask, spread across discovery, comparison and pricing. Each run asks every active prompt on every selected engine in your browser, so a focused list finishes sooner and is easier to read. ### Should my brand name be in the prompt? Mostly no. A prompt that names you will almost always mention you, which tells you little. Keep a few branded prompts such as "Acme Analytics vs Mixpanel" to see how AI describes you, and keep the rest neutral. ### What counts as a mention? A case-insensitive, whole-word match of any alias you set for the brand. A brand counts once per answer, however often it appears. The [methodology](https://aeometer.com/methodology) page has every definition. ### Why does editing a prompt start over? Different wording can get different answers. Starting a new baseline keeps old and new answers from being mixed into one trend line. The old answers remain available. ### Do checks run when my computer is off? No. Schedules queue checks, and the AEOmeter extension collects them when Chrome is open, connected and signed in to the engines. If Chrome was closed, the app shows when the last answer was actually captured. ### Does AEOmeter use the AI engines' APIs? No. Prompts go through the consumer apps your buyers use, in your browser, signed in like a real person. API responses can differ from what people see in the apps. --- - Web page: https://aeometer.com/features/prompt-tracking - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # See which pages AI answers cite, and whether any are yours > AI citation tracking for ChatGPT, Claude, Perplexity and Google AI Overviews. See every cited source, your owned citation rate and which pages keep showing up. When ChatGPT, Claude, Perplexity or a Google AI Overview shows sources for an answer, AEOmeter saves them next to the answer text. You see which sites shape answers in your category, how often your own site is one of them, and which competitor pages keep showing up. ## Citations and inline links are kept apart A citation is a source the engine attributes to its answer: a source pill in ChatGPT, a numbered source in Perplexity, a source card under an AI Overview. An inline link is a URL that appears inside the answer text. AEOmeter stores the two separately. The answer view lists citations first and "Linked in the answer" below them. Keeping them apart matters because a passing link to your pricing page is weaker evidence than being one of the sources an answer is built on. ## Owned citation rate Shown as **Cited** in the app. It is the share of successfully captured answers that cite at least one link on your domain or its subdomains. Add your domains in brand settings and both acme.io and docs.acme.io count. Failed checks and searches where Google showed no AI Overview are left out of the denominator. With zero answers in a period you see a dash instead of 0%. Example data | Answers captured | Answers citing acme.io | Owned citation rate | | --- | --- | --- | | 43 | 8 | 19% | ## Every cited site and page, grouped The Sources view lists everything AI answers cited in the period, grouped by site or by page. Each source is labelled as your site, a competitor site, community or other, with the engines that cited it and how many answers did. Click a source to read the exact answers that cited it. This is where you find the review directories, Reddit threads and comparison articles that keep appearing for your prompts. Those are the pages worth earning a mention on. ## When a source list may be incomplete Engines hide some sources behind grouped buttons and expanding panels, and they change their pages often. When AEOmeter cannot be sure it read every source, the answer says "Some sources may be missing" instead of presenting a partial list as complete. The engine adapters are experimental for this reason. A citation records what the engine attributed. It does not prove that the page caused the answer. ## Questions ### Which engines show citations? Perplexity shows sources on almost every answer. Google AI Overviews show source cards. ChatGPT shows sources when it searches the web for an answer. Claude adds citations when it uses web search, and many Claude answers have none. See [engines](https://aeometer.com/engines) for how each one is read. ### Is a citation the same as a mention? No. A mention is your brand name in the answer text. A citation is a source link to your domain. You can be mentioned without being cited, and an answer can cite your blog without naming your product. ### Do subdomains count as my site? Yes. Links on your owned domains and their subdomains count toward your owned citation rate. ### Does a cited page cause the answer? AEOmeter cannot tell you that. It records which sources the engine showed. Treat a frequently cited page as a strong hint about what the engine reads, nothing more. ### How do I get cited more often? AEOmeter measures citations. It does not write or publish content. Our [answer engine optimization guide](https://aeometer.com/guides/answer-engine-optimization) covers what tends to help. --- - Web page: https://aeometer.com/features/citation-tracking - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # See where competitors show up in AI answers and you don't > Track competitor AI visibility with aliases, tracked share of voice and gaps where a rival is named and you are not. Export the comparison as CSV. Add the competitors your buyers compare you with. AEOmeter checks every answer for each of them, so you can see who gets named for your prompts, on which engines, and where a competitor shows up without you. ## Competitors, with every name they go by Each project has one brand of your own and the competitors you choose. Give every brand its aliases and owned domains. Aliases catch the different ways an answer may name a product, and domains let AEOmeter tell a citation of posthog.com from one of yours. Matching is deterministic: a case-insensitive, whole-word match of an alias. A brand whose name is also a common word needs a more specific alias. ## Tracked share of voice, defined precisely **Tracked share of voice** is your brand's mentions divided by the mentions of all tracked brands in the same answers. Each brand counts once per answer. It describes your share among the brands you chose to track. It is not market share, and adding or removing a competitor changes it. Visibility answers a different question: in what share of answers does a brand appear at all? The dashboard shows visibility for every tracked brand, with its denominator. Tracked share of voice comes from the [analytics API and MCP tools](https://aeometer.com/features/mcp). Example data: 43 captured answers | Brand | Answers mentioning it | Visibility | Tracked share of voice | | --- | --- | --- | --- | | Mixpanel | 25 | 58% | 37% | | Acme Analytics (you) | 18 | 42% | 27% | | PostHog | 15 | 35% | 22% | | Amplitude | 9 | 21% | 13% | Acme's share of voice is 18 ÷ (25 + 18 + 15 + 9) = 18 ÷ 67, about 27%. Its visibility is 18 ÷ 43, about 42%. ## Answers where a competitor appears and you don't "Where you're missing" lists the prompts where at least one competitor was mentioned and your brand was not, which competitors appeared, and on how many engines it happened. Each gap is backed by answers you can open and read. Use it as a to-do list. A prompt where Mixpanel and PostHog both show up on three of four engines and you never do usually points at a comparison page or a review site you are missing from. ## Export the comparison as CSV Export one row per brand per engine for the period and filters you are looking at. Paste it into a spreadsheet or a board update. ```csv "Brand","Engine","Answers","Answers mentioning brand","Answers citing brand","Period (days)","Engine filter","Tag filter","Exported at" "Acme Analytics","Perplexity","11","5","3","30","all","all","2026-10-16T09:12:00.000Z" "Mixpanel","Perplexity","11","7","2","30","all","all","2026-10-16T09:12:00.000Z" ``` ## Questions ### Is tracked share of voice the same as market share? No. It only covers the brands you track, in the answers AEOmeter captured for your prompts. Add a competitor and every brand's share changes. Read it as "of the tracked brands AI named, how much of that was us". The [AI share of voice guide](https://aeometer.com/guides/ai-share-of-voice) goes deeper. ### How many competitors should I add? The three to five your buyers actually compare you with. More brands make the comparison noisier without telling you much more. ### Do failed checks count against me? No. Failed and blocked checks, and searches where Google showed no AI Overview, are excluded. They never count as misses. ### Do you score sentiment or ranking position? No. AEOmeter records whether each brand is mentioned and whether its site is cited, with the answer text to read. It does not score sentiment or assign positions. --- - Web page: https://aeometer.com/features/competitor-tracking - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Ask Claude or Cursor about your AI visibility > Connect AEOmeter's remote MCP server to Claude, Cursor or any MCP client with OAuth, or use the REST API and OpenAPI spec to work with your AI visibility data. AEOmeter runs a remote MCP server at `https://aeometer.com/mcp`. Connect it to Claude, Cursor or another MCP client, sign in with OAuth, and ask about your prompts, answers and competitors in plain language. The same operations are available as a REST API with an OpenAPI spec. ## Questions you can ask your assistant The assistant calls AEOmeter tools for you and answers from your data. A few examples: - "Which prompts mention PostHog but not us this week?" - "Show me what Perplexity said for 'best product analytics tool for a seed-stage startup' and which sources it cited." - "Which domains get cited most for my pricing prompts this month?" - "Import these 20 prompts and tag them alternatives." - "Is my browser ready to run checks?" - "Set my daily schedule to 9:00 in Europe/Berlin." ## Connect it in a minute Add a remote MCP server with the URL `https://aeometer.com/mcp` and sign in when your client asks. The OAuth scope is `mcp:tools`. Most clients that support remote servers accept a config like this: ```json { "mcpServers": { "aeometer": { "url": "https://aeometer.com/mcp" } } } ``` In Claude Code: ```bash claude mcp add --transport http aeometer https://aeometer.com/mcp ``` In Claude on the web or desktop, add it as a custom connector with the same URL. Config formats vary by client, so check your client's docs if a field name differs. If your client caches tool lists, reconnect after AEOmeter releases. ## What the tools cover The MCP tools use the same handlers and permissions as the web app. Every project tool checks that you own the project. | Workflow | Tools | | --- | --- | | Projects and brands | `list_projects`, `get_project`, `create_project`, `update_project` | | Prompts and tags | `list_prompts`, `create_prompt`, `import_prompts`, `add_starter_prompts`, `update_prompt`, `bulk_update_prompts` | | Browser monitoring | `get_browser_status`, `get_schedule`, `update_schedule`, `list_runs`, `create_run`, `cancel_job`, `cancel_queued_jobs` | | Results and evidence | `list_observations`, `get_analytics`, `get_answer`, `reanalyze_project` | ## What to know before you connect Starting a run from your assistant queues checks. They are collected when your Chrome is connected, and at that point real prompts go to the AI engines. Browser status reports what your extension advertises. It does not verify that you are logged in to each engine. Editing prompts or brands through MCP starts a new measurement baseline, exactly as in the web app. Credential management and the extension's collector endpoints are not exposed over MCP. Whatever the tools return goes to the AI client you connected, under that client's terms. ## REST API and OpenAPI spec Everything in the dashboard is also available over HTTP under `/api/v1`. The OpenAPI spec lives at [/openapi.json](https://aeometer.com/openapi.json) and the browsable reference at [/docs](https://aeometer.com/docs). Use it for scripts, internal dashboards or your own exports. ## Questions ### Which MCP clients work? Any client that supports remote MCP servers over HTTP with OAuth sign-in. That includes Claude, Claude Code and Cursor. ### Can the assistant start checks? It can queue a run, optionally for selected prompts. The checks run in your Chrome when the extension is connected and the engines are signed in. Nothing runs on a server on your behalf. ### Do I need an API key for MCP? No. Your client opens an AEOmeter sign-in page and receives an OAuth token with the `mcp:tools` scope. ### Is the MCP data the same as the dashboard? Yes. MCP and the web app share the same schemas and handlers, so the numbers match what you see in the app. --- - Web page: https://aeometer.com/features/mcp - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Four AI answer engines, checked the way your buyers use them > Track your brand's visibility in AI search engines: how AEOmeter checks ChatGPT, Claude, Perplexity and Google AI Overviews in your own Chrome, and what counts as a citation on each. AEOmeter tracks your brand in four AI search engines: ChatGPT, Claude, Perplexity and Google AI Overviews. Each one is checked in its real consumer interface, in your Chrome, signed in with your own accounts. Answers come from the same place your buyers get them. ## How each engine is checked The four engines answer and cite differently, so each has its own adapter. Here is what AEOmeter does on each one. | Engine | Private mode | What counts as a citation | Main caveat | | --- | --- | --- | --- | | [ChatGPT](https://aeometer.com/engines/chatgpt) | Temporary chat, Unpersonalized when offered | Source pills when ChatGPT searches the web | Grouped sources can hide some links; flagged as incomplete | | [Claude](https://aeometer.com/engines/claude) | Incognito chat | Citation links when Claude uses web search | Many answers are prose with no links, so mentions carry the signal | | [Perplexity](https://aeometer.com/engines/perplexity) | Incognito, switched on before the prompt is typed | Numbered sources, read from Perplexity's thread data | Research-heavy answers can take minutes; the wait limit is 10 | | [Google AI Overviews](https://aeometer.com/engines/google-ai-overviews) | Search without personalization | Source cards attached to the overview | "No AI Overview" is its own result and never counts as a miss | ## What stays the same on every engine Every check runs in your own Chrome through the AEOmeter extension, in a background tab inside an "AEOmeter" tab group. The extension uses Chrome's debugger API to type and click like a person, so Chrome shows a "started debugging this browser" banner while it works. - Each prompt goes to all its selected engines at once. The next prompt starts when every check in the batch has finished. - The extension turns on each engine's private mode first. This is best effort and the visible evidence is recorded. - Prompts are entered into the engine's own composer, and a prompt is never sent twice, even after an interruption. - A login wall, verification challenge or rate limit pauses the check and shows a red "!" badge on the extension. AEOmeter never tries to solve a CAPTCHA. - An answer that has not finished after 10 minutes is marked failed. Failed checks never count as misses. - Only the latest answer and its sources are saved. The extension never reads provider cookies or account tokens. ## Adapters are experimental AI engines change their pages without notice. Each adapter tries several ways to find a control and records which one worked, which helps it keep working through small layout changes. A check can still fail after an engine redesign, and some sources can be missed. When that happens you see a failed check or an "incomplete sources" note instead of a wrong number. ## Engines AEOmeter does not track AEOmeter tracks exactly these four engines. Gemini, Google AI Mode, Microsoft Copilot and Grok are not tracked. ## Questions ### Do I need an account on every engine? You need to be signed in to the chat apps you want checked, in the Chrome profile where the extension runs. Google AI Overviews are read from a normal Google search. ### Why not use the engines' APIs? API responses can differ from what buyers see in the apps, and Google AI Overviews have no equivalent API. Checking the consumer interfaces measures the experience your buyers actually have. ### Can I pick engines per prompt? You choose which engines each run or schedule covers. Every prompt in that run goes to all of them at once. ### Does the AEOmeter extension work in other browsers? AEOmeter is built for Chrome, and Chrome has to be open for checks to run. --- - Web page: https://aeometer.com/engines - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Track how ChatGPT talks about your brand > ChatGPT brand visibility tracking: see how often ChatGPT mentions and cites your brand for buyer prompts, checked in temporary chats in your own Chrome. People look for a ChatGPT rank tracker, but ChatGPT keeps no ranking. Ask the same question twice and you can get a different answer, a different order of brands and different sources. AEOmeter tracks what holds up across many checks: how often ChatGPT mentions your brand for the questions buyers ask, which sources it cites, and which competitors it names instead. ## How ChatGPT answers and cites ChatGPT decides per question whether to search the web. When it searches, sources appear as small pills next to the text and in a sources list. When it answers from the model alone, there may be no sources at all, and the only question is whether your name comes up. For software questions, cited sources tend to be vendor sites, comparison articles, review directories and community threads. The mix shifts by question. The [Sources view](https://aeometer.com/features/citation-tracking) shows the real list for your prompts. ## How AEOmeter checks ChatGPT Each ChatGPT check follows the same steps, in a background tab of your Chrome, signed in with your account. 1. **Open a temporary chat** A new chat opens with ChatGPT's temporary chat mode on. Temporary chats stay out of your chat history. 2. **Switch to Unpersonalized** If ChatGPT offers Personalized and Unpersonalized, AEOmeter picks Unpersonalized. Private mode is best effort and the visible evidence is saved with the check. 3. **Enter and send the prompt once** The prompt is filled in, checked word for word, then sent. If the check is interrupted after sending, it is never sent again. 4. **Wait for a finished answer** AEOmeter waits until the stop control is gone, a finished control such as Copy or "Response complete" appears, and the text has stopped changing. 5. **Save the answer and sources** The last answer is saved as Markdown. Source pills become citations. Links inside the text are saved separately as inline links. ## ChatGPT-specific caveats - ChatGPT sometimes groups several sources behind one pill. When AEOmeter cannot expand every group, the answer is marked "Some sources may be missing". - The answer comes from whichever model your account uses. AEOmeter records the visible model when it can. - A login prompt, verification check or usage limit pauses the check with a red "!" badge. You resolve it in the tab, then resume. - ChatGPT changes its interface often. The adapter is experimental and may need updates after a redesign. ## Example prompts to track in ChatGPT ChatGPT gets conversational, detailed questions. Write prompts the way a buyer would type them: - "What's the best product analytics tool for a seed-stage SaaS?" - "I'm outgrowing Google Analytics for my app. What should I switch to?" - "Mixpanel alternatives with a generous free tier" - "Compare Mixpanel, PostHog and Amplitude for a five-person team" ## What to look for in ChatGPT results - **Mentioned but never cited.** ChatGPT knows your name but links elsewhere. Check which pages it cites instead. - **Competitor pages in the sources.** A competitor's comparison page cited for a neutral prompt is worth reading closely. - **Prompts where you're missing.** Gaps that repeat across many checks matter more than a single miss. - **Movement over weeks.** A single answer is one sample. Look at visibility over a few weeks of checks before drawing conclusions. ## Questions ### Can AEOmeter show my rank in ChatGPT? No, because ChatGPT has no stable rank to show. The order of brands in an answer changes from one chat to the next. AEOmeter measures mention rate (shown as Visibility), owned citation rate and tracked share of voice, each with its denominator. See the [methodology](https://aeometer.com/methodology). ### Does AEOmeter use the OpenAI API? No. It uses chatgpt.com in your Chrome, signed in with your account, so the answers match what you would see yourself. ### Will these checks fill up my ChatGPT history? AEOmeter uses temporary chats, which stay out of your chat history. Private mode is best effort: if ChatGPT changes how temporary chats work, the evidence recorded with each check shows what was visible. ### How often does ChatGPT get checked? As often as your schedule says, when Chrome is open and connected. A daily schedule is a good default. More checks give you more samples to average. ### How is this different from tracking brand mentions by hand? The method is the same: ask, read, note who is mentioned. AEOmeter does it for every prompt and keeps the history. Our guide on [tracking brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt) covers the manual way too. --- - Web page: https://aeometer.com/engines/chatgpt - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # See whether Claude recommends your product > Track Claude brand visibility: how often Claude mentions your product for buyer questions, how it describes you, and which competitors it names, from incognito chats. Claude often answers from what it already knows, in plain prose, with no links. That makes Claude brand visibility mostly a question of mentions: does Claude bring up your product when a buyer asks for a recommendation, and how does it describe you? AEOmeter asks Claude your buyers' questions in incognito chats and keeps every answer. ## How Claude answers and cites For many questions Claude writes a careful answer from its own knowledge and shows no sources. When it uses web search, it adds citation links to the pages it read. Whether that happens depends on the question and on the web search setting in your Claude account. Because links are rarer, the answer text matters more on Claude. Read how it describes you next to your competitors. A product described in one vague line differs a lot from one listed with its use case. ## How AEOmeter checks Claude Each Claude check runs in a background tab of your Chrome, signed in with your Claude account. 1. **Open a new incognito chat** A new chat opens on claude.ai in incognito mode. If incognito is not on yet, AEOmeter clicks "Use incognito". Incognito chats stay out of your chat history. 2. **Enter and send the prompt once** The prompt is filled in and checked word for word before sending. A prompt that was already sent is never sent again. 3. **Wait for a finished answer** AEOmeter waits until Claude stops generating, the Copy control appears and the text has stopped changing. 4. **Save the answer and citations** The final response is saved as Markdown, without the thinking summary Claude may show above it. Citation links are saved as citations and other links as inline links. ## Claude-specific caveats - Many Claude answers have no sources. Expect a low owned citation rate on Claude even when you are mentioned often. - Claude can group citations. When AEOmeter cannot read every one, the answer is marked "Some sources may be missing". - Claude has usage limits. Hitting one pauses the check with a red "!" badge until you resume. - The adapter is experimental. Claude's interface changes and the adapter may need an update after a redesign. ## Example prompts to track in Claude People tend to give Claude more context. Your prompts can too: - "I run a two-person SaaS and need product analytics. What would you pick and why?" - "Explain the trade-offs between Mixpanel and PostHog for an early-stage startup" - "Which analytics tools are easiest to set up without a data engineer?" - "What should I look for in a product analytics tool before Series A?" ## What to look for in Claude results - **Whether you are named at all.** On Claude this is the main signal. Visibility per prompt shows where you are left out. - **How you are described.** Open the answers and read the sentence about you. Outdated or vague descriptions are worth fixing at the source. - **Who appears instead.** Gaps show prompts where competitors are named and you are not. - **Answers with web search.** When Claude does cite pages, the Sources view shows which ones. ## Questions ### Does Claude rank brands? No. Claude writes an answer, and the order of names in it can change between chats. AEOmeter records whether each tracked brand is mentioned and whether your site is cited. ### Does AEOmeter use the Anthropic API? No. It uses claude.ai in your Chrome, signed in with your account, in incognito chats. ### Why does Claude show fewer citations than Perplexity? Perplexity searches for almost every question. Claude answers many questions from its own knowledge and only cites sources when it searches the web. ### Will the checks show up in my Claude history? AEOmeter uses incognito chats, which stay out of your history. Private mode is best effort, and each check records the evidence that was visible. --- - Web page: https://aeometer.com/engines/claude - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Track your brand in Perplexity answers and sources > A Perplexity rank tracker for brands: see how often Perplexity mentions you and which of its sources link to your site, read from thread data in incognito mode. If you searched for a Perplexity rank tracker, you probably want to know where your brand lands in Perplexity's answers. Perplexity has no fixed ranking to track, but it is the most source-heavy of the four engines. AEOmeter records whether each answer mentions you and which of its many sources link to your site. ## How Perplexity answers and cites Perplexity searches the web for almost every question and numbers its sources inline. A single answer often draws on a few dozen sources in our own test runs. The numbers show the order of sources in that answer. They are not a ranking of brands. That makes Perplexity the best place to study citations. If your site is never cited here, the Sources view shows which sites are. ## How AEOmeter checks Perplexity Each Perplexity check runs in a background tab of your Chrome, signed in with your Perplexity account. 1. **Turn on incognito first** AEOmeter opens perplexity.ai and clicks "Use incognito" before anything else. Perplexity's own link prefill would send the question before incognito is on, so the prompt is typed instead. 2. **Type and send the prompt once** The prompt is typed into Perplexity's composer, checked word for word, then sent. A prompt that was already sent is never sent again. 3. **Wait for the thread to complete** AEOmeter waits until Perplexity's own thread data reports the answer as completed. The page is used as a fallback. 4. **Save the answer and sources** The answer text and its source list are read from the same thread data, which is more complete than scraping the page. Sources are saved as citations. ## Perplexity-specific caveats - Research-heavy answers can take several minutes. AEOmeter waits up to 10 minutes before marking a check failed. - Verification challenges and rate limits pause the check with a red "!" badge. AEOmeter never tries to solve them. - Perplexity's thread data and page layout can change. The adapter is experimental and falls back to the page when the data is not available. ## Example prompts to track in Perplexity People use Perplexity like a research assistant. Comparison and "best for" questions work well: - "Best product analytics tools for startups in 2026" - "Mixpanel vs PostHog pricing for 1 million events a month" - "Open-source alternatives to Amplitude" - "Which product analytics tools have a free plan?" ## What to look for in Perplexity results - **Owned citation rate.** With so many sources per answer, Perplexity is where your own pages have the most chances to be cited. - **The domains it relies on.** Group the Sources view by site to see which review sites and forums Perplexity reads for your category. - **Competitor pages.** Pages from competitor sites that appear for neutral prompts show what Perplexity treats as a reference. - **Mentions without citations.** You can be named from someone else's list. The answer view shows which source the list came from. ## Questions ### Can AEOmeter show my rank in Perplexity? Perplexity has no stable rank. Brand order and sources change between answers. AEOmeter measures how often you are mentioned and cited, with the answers behind every number. ### Does AEOmeter use the Perplexity API? No. It uses perplexity.ai in your Chrome, signed in with your account, in incognito mode. ### Why read sources from thread data? Perplexity's page shows sources in collapsible panels that are easy to miss. The thread data behind the page lists them all, so the source list is more complete. ### Do source numbers mean anything? They show where a source sits in that one answer. AEOmeter does not treat them as a ranking. --- - Web page: https://aeometer.com/engines/perplexity - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Track your brand in Google AI Overviews > Google AI Overview tracker: see when an overview appears for your prompts, whether it mentions you and which source cards it shows. No overview is never a miss. Google shows an AI Overview above the results for some searches and not for others. AEOmeter works as a Google AI Overview tracker: it runs your prompts as real Google searches in your Chrome, saves the overview when one appears with its source cards, and records "No AI Overview shown" when it doesn't. ## How AI Overviews answer and cite An AI Overview is a short generated summary at the top of a Google results page, with source cards and links next to it. Google decides per search whether to show one, and that can change with the query, the day and where you search from. Overviews tend to answer shorter, search-style queries. Source cards often point to pages that also rank in the regular results below. ## How AEOmeter checks Google AI Overviews Google has no composer to type into, so this check is a search. 1. **Run the search without personalization** A background tab opens a Google search for the exact prompt with personalization turned off, using Google's own "without personalization" option, and an English interface. 2. **Wait for the overview** If an AI Overview appears, AEOmeter clicks "Show more" and waits for the text to stop changing. 3. **Save the overview and source cards** The overview text is saved as Markdown and its source cards as citations. Regular search results below it are not saved. 4. **Or record "No AI Overview shown"** If no overview appears within 20 seconds, the check ends with that result. ## "No AI Overview" is not a miss When Google shows no overview, there was no AI answer that could have mentioned you. Counting that as a miss would drag your visibility down for reasons that have nothing to do with your brand. AEOmeter keeps it as a separate result, shows it as a dash in the Prompts view, and leaves it out of every rate. Example data: one prompt, 12 searches | Searches | Overviews shown | Overviews mentioning you | Visibility | | --- | --- | --- | --- | | 12 | 7 | 3 | 3 of 7, about 43% | ## Google-specific caveats - Google picks results partly by location. Searches run from your own network, and AEOmeter records the browser language, not a verified location. - Personalization is turned off on a best-effort basis through Google's own option. - A Google "unusual traffic" check pauses the run with a red "!" badge. AEOmeter never tries to solve it. - AEOmeter reads AI Overviews only. Google AI Mode and regular rankings are not tracked. - The adapter is experimental. Google changes its results layout often. ## Example prompts to track in Google AI Overviews Write these like searches, shorter than chat prompts: - "best product analytics for startups" - "mixpanel vs posthog" - "product analytics free plan" - "how to track feature adoption in a saas app" ## What to look for in AI Overview results - **How often an overview appears.** A prompt that rarely triggers one matters less on Google than on the chat apps. - **Source cards.** Which pages Google attaches to the overview, and whether one of them is yours. - **Mentions inside the overview.** Whether the summary names you or only your competitors. ## Questions ### Does AEOmeter track Google AI Mode? No. It reads AI Overviews on the regular results page only. ### Does it track my organic Google rankings? No. Regular search results are not saved. Use an SEO rank tracker for those. ### Why did a prompt show "No AI Overview"? Google did not show an overview for that search at that time. It is recorded as its own result and excluded from your rates. ### Do I need to be signed in to Google? Searches run in your normal Chrome profile with personalization turned off. AEOmeter does not need anything beyond what a normal Google search needs. --- - Web page: https://aeometer.com/engines/google-ai-overviews - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Free AI visibility checker > Check if ChatGPT, Claude, Perplexity and Google AI Overviews mention your brand. Get buyer prompts, ask each engine and score your mention rate. Free, no signup. Find out if ChatGPT, Claude, Perplexity and Google AI Overviews name your brand when buyers ask for a recommendation. Describe what you sell and get the questions your buyers type into AI. Open each one on each engine in one click, note what you see, and get your mention rate. Runs in your browser. What you type and mark stays in this browser and is never sent to AEOmeter. Use it at https://aeometer.com/tools/ai-visibility-checker. ## How the checker works 1. **Describe your brand.** Brand name, what you sell, who buys it and up to five competitors. The checker writes 8 to 12 buyer prompts across five intents. 2. **Ask each engine.** Open each prompt in ChatGPT, Claude, Perplexity and Google, each in a new tab, and read what comes back. 3. **Mark what you see.** Mentioned, Not mentioned, Competitor only or No answer. Your mention rate and per-engine breakdown update as you go. ## Prompt templates The checker fills these templates from your inputs. They follow the same patterns AEOmeter uses for the starter prompts it suggests when you add a brand. None of them names your brand, because a question that names you mentions you by definition. Tokens: `{category}` is what you sell (for example "product analytics tool"), `{audience}` is who buys it (defaults to "a small team"), `{competitor1}` and `{competitor2}` are the first two competitors you list, `{problem}` is an optional problem phrased as a verb ("find out why users churn"). Templates are used in this order, up to 12 prompts. | Intent | Template | Used when | | --- | --- | --- | | Category discovery | What is the best {category}? | always | | Category discovery | Which {category} would you recommend for {audience}? | always | | Category discovery | What are the top {category} options right now? | always | | Comparison | Compare the most popular {category} options for {audience}. | always | | Comparison | {competitor1} vs {competitor2}: which is better for {audience}? | 2+ competitors | | Comparison | Is {competitor1} the best {category} for {audience}, or is there something better? | exactly 1 competitor | | Alternatives | What are the best alternatives to {competitor1}? | 1+ competitor | | Alternatives | What are the best alternatives to {competitor2}? | 2+ competitors | | Alternatives | What are some lesser-known {category} options worth trying? | no competitors given | | Problem-led | How can {audience} {problem}? | a problem is given | | Problem-led | What tools help {audience} {problem}? | a problem is given | | Use case | Which {category} is easiest to set up for {audience}? | always | | Use case | What is a good {category} for {audience} on a tight budget? | always | | Use case | Which {category} should {audience} switch to as they grow? | always | Example for a product analytics tool sold to seed-stage startups, with Mixpanel and PostHog as competitors: "What is the best product analytics tool?", "Mixpanel vs PostHog: which is better for seed-stage startups?", "What are the best alternatives to Mixpanel?". ## Asking each engine Ask each prompt in a fresh chat on every engine, so earlier conversations don't shape the answer. - **ChatGPT:** Opens a new chat in your ChatGPT account. - **Perplexity:** Runs the search right away. - **Claude:** Needs a Claude login. Start a new chat for each prompt. - **Google AI Overviews:** Look for the AI Overview above the results. Mark No answer if there is none. ## Marking each answer - **Mentioned:** The answer names your brand. - **Not mentioned:** You are not named, and neither is any competitor you listed. - **Competitor only:** A competitor you listed is named and you are not. - **No answer:** Error, refusal, login wall, or no AI Overview. Left out of the score. Mention rate = Mentioned ÷ (Mentioned + Not mentioned + Competitor only). No answer is left out of the denominator. With zero usable answers the rate is a dash, not 0%. Calculate it overall and per engine. ## How to read your results The checker uses the same mention rate AEOmeter calculates for tracked brands, with you doing the reading instead of the software. **Mention rate:** Answers that name your brand ÷ answers that came back. If ChatGPT names you in 3 of 8 answers, your ChatGPT mention rate is 38%. An engine with no usable answers shows a dash, not 0%. See the [methodology](https://aeometer.com/methodology) for how AEOmeter counts it. **No answer:** Errors, refusals, login walls and searches where Google showed no AI Overview. They say nothing about your visibility, so they are left out of the denominator. **Competitor only:** The most useful mark on the page. The AI recommended someone in your category and it wasn't you. Open that answer again and note which pages it cites. That is your reading list for this question. **Per-engine breakdown:** Each engine builds its answers differently, so a strong ChatGPT number and a weak Perplexity number is common. Start with the engine your buyers use most. ## What this score can't tell you - **It is one sample.** Ask the same prompt twice and you can get a different list of brands. A single check is a snapshot of one moment. - **Answers are personalized.** Signed-in accounts can use memory, past chats and location, so what you see can differ from what a buyer sees. - **The sample is small.** 12 prompts on 4 engines is 48 answers at most. A 10-point gap between two engines can be noise. - **You decide what counts.** AEOmeter counts a mention when an answer contains your brand name or one of its aliases as a whole word. Use the same rule here so your numbers compare. ## Track these prompts automatically Checking by hand works once. To see whether you are gaining or losing ground, run the same prompts on a schedule and keep every answer. AEOmeter does that from your own Chrome with your own AI accounts, saves the original answer and its sources, and tracks your competitors in the same answers. Copy your prompts as CSV and import them when you set up your brand. Start at https://aeometer.com/sign-up. Needs Chrome open while checks run. CSV format for import: a `prompt` column and a `tags` column with semicolon-separated tags. ```csv prompt,tags "What is the best product analytics tool?","discovery" "What are the best alternatives to Mixpanel?","alternatives" ``` ## FAQ ### Is the AI visibility checker free? Yes. It runs in your browser with no account, no API key and no server calls. Your inputs and marks are saved in this browser's local storage so you can come back to them. ### Which AI engines does it cover? ChatGPT, Claude, Perplexity and Google AI Overviews. These are the same four engines AEOmeter tracks. ### Why don't the prompts include my brand name? A prompt like "Is Acme any good?" will mention Acme every time. That tells you the AI knows you exist. It says nothing about whether buyers find you when they ask for a recommendation, which is what the mention rate measures. ### Why is No answer left out of the score? A failed or missing answer is not a miss. Counting it as one would push your score down for reasons that have nothing to do with your brand. AEOmeter treats failed, blocked and no-AI-Overview checks the same way. ### How is this different from AEOmeter? The checker is a manual snapshot. AEOmeter runs your prompts on a daily schedule in your own Chrome, keeps every answer with its cited sources, and shows your mention rate, citations and share of voice against competitors over time. Your Chrome has to be open for checks to run. ### Can I import these prompts into AEOmeter? Yes. Click Copy prompts as CSV and paste the result into the prompt import in AEOmeter. Each prompt keeps its intent as a tag. ## Related - [Track brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt): Turn this one-off check into a weekly habit, by hand or with software. - [How to choose prompts](https://aeometer.com/guides/choosing-prompts): Write prompts that match how your buyers ask, and avoid the ones that flatter you. - [Methodology](https://aeometer.com/methodology): How AEOmeter calculates mention rate, citations and share of voice. --- - Web page: https://aeometer.com/tools/ai-visibility-checker - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AI visibility tool alternatives > A fair, dated look at Peec AI, Profound, Otterly.AI, Ahrefs Brand Radar and Semrush, and when AEOmeter is or is not the right pick for a founder. Last updated: 2026-10-16 There are dozens of AI visibility tools now. Most of them answer the same question: when someone asks ChatGPT, Perplexity or Google for a recommendation in your category, does your brand come up? They differ in who they are built for, where the answers come from and what they add on top. This page gives a short, fair map of the tools founders usually compare, and says when AEOmeter is the wrong choice. > **Checked October 16, 2026** Facts about other tools come from their own public websites on that date. We did not sign in to any of them. Plans change, so confirm on their sites before you buy. ## The tools at a glance Checked October 16, 2026. Engine lists are what each vendor names publicly; availability can depend on the plan. | Tool | Built for | AI engines named publicly | Where checks run | | --- | --- | --- | --- | | [Peec AI](https://peec.ai/) | Marketing teams, SEO teams, agencies | ChatGPT, Perplexity, Gemini, Google AI Mode, Copilot ([docs](https://docs.peec.ai/intro-to-peec-ai)) | Hosted, daily | | [Profound](https://www.tryprofound.com/) | Enterprise AEO, content and PR teams, agencies | Up to 9 on Enterprise, including ChatGPT, Claude, Gemini, Copilot, AI Mode ([pricing](https://www.tryprofound.com/pricing)) | Hosted, daily on Enterprise | | [Otterly.AI](https://otterly.ai/) | Marketers, SEO teams, agencies, solo marketers | ChatGPT, Google AI Overviews, Perplexity, Copilot. Add-ons: AI Mode, Gemini, Claude ([pricing](https://otterly.ai/pricing)) | Hosted, automatic | | [Ahrefs Brand Radar](https://ahrefs.com/brand-radar) | SEO teams already using Ahrefs | AI Overviews, AI Mode, Perplexity, Gemini, Copilot, ChatGPT ([Brand Radar](https://ahrefs.com/brand-radar)) | Hosted, plus a large pre-built index | | [Semrush AI Visibility](https://www.semrush.com/solutions/ai-visibility/) | SEO teams already using Semrush | ChatGPT, Gemini, Perplexity, SearchGPT, Google AI Mode, AI Overviews ([Semrush](https://www.semrush.com/solutions/ai-visibility/)) | Hosted | | AEOmeter | Founders tracking 1 to 3 own brands | ChatGPT, Claude, Perplexity, Google AI Overviews | Your own Chrome, with your signed-in accounts, while it is open | ## Dedicated AI visibility tools ### Peec AI Peec AI is a hosted AI search analytics platform for marketing teams and agencies. Its docs say it reads answers through each engine's web interface every day. It adds sentiment, position, prompt suggestions, action recommendations, an API and an MCP server. A good fit when a team needs daily data across several engines and projects. [Read AEOmeter vs Peec AI](https://aeometer.com/alternatives/peec-ai). ### Profound Profound is the broadest platform here. Beyond answer monitoring it offers prompt demand data, crawler analytics, content agents, ChatGPT Shopping tracking and enterprise controls such as SSO and SOC 2. A good fit for large marketing teams that want one platform for AI search. [Read AEOmeter vs Profound](https://aeometer.com/alternatives/profound). ### Otterly.AI Otterly.AI is a self-serve monitoring tool with prices on its website and an entry plan for solo marketers. It includes four engines, sells others as add-ons, and adds GEO audits, sentiment and crawler analytics. A good fit if you want hosted monitoring with a known monthly price. [Read AEOmeter vs Otterly.AI](https://aeometer.com/alternatives/otterly-ai). ## AI visibility inside SEO suites If you already pay for an SEO suite, check what it includes before buying another tool. Ahrefs Brand Radar pairs a large pre-built index of AI answers with custom prompt tracking, and Ahrefs says custom prompts are included with paid plans from Lite up ([source](https://ahrefs.com/brand-radar)). Semrush offers AI visibility reporting with prompt tracking limits that depend on the plan ([source](https://www.semrush.com/solutions/ai-visibility/)). Both make most sense when AI visibility is one more report next to your rankings and backlinks. ## When AEOmeter is the right pick - You are a founder or indie hacker tracking your own product and two or three competitors. - You want answers from the consumer apps your buyers use, collected in your own Chrome under your own signed-in accounts. - You want to click any percentage and read the answers behind it, with failed checks shown separately instead of counted as misses. - You want to ask your AI assistant about your visibility through the [AEOmeter MCP server](https://aeometer.com/features/mcp). ## When AEOmeter is the wrong pick - You need checks to run while your computer is off. AEOmeter only collects while your Chrome is open and connected. - You need Gemini, Google AI Mode, Copilot or other engines beyond ChatGPT, Claude, Perplexity and Google AI Overviews. - You manage many brands or clients and need agency workspaces, white-label reports or alerts. - You need sentiment scores, ranking positions, prompt search volume or content generation. - You need a finished, stable product. AEOmeter is in beta and its adapters for each AI app are experimental. ## What to check in any AI visibility tool 1. **Can you see the answer?** A percentage you cannot open is hard to trust. Ask to see the actual answer text and sources behind a number. 2. **What is the denominator?** Find out whether failed checks or missing AI Overviews count as "not mentioned". Our [methodology](https://aeometer.com/methodology) explains how AEOmeter handles them. 3. **Which engines, on which plan?** Many tools name engines on the homepage that are add-ons or enterprise-only. 4. **Where do answers come from?** API responses, hosted browsers and your own signed-in browser can give different answers to the same prompt. 5. **What happens when you edit a prompt?** Trends only mean something when they compare the same prompt over time. ## FAQ ### What is the best AI visibility tool for a founder? It depends on what you need. If you want hosted daily data across many engines, Peec AI or Otterly.AI are worth a look. If you want to read answers collected from your own accounts and keep the tool simple, try AEOmeter. ### Do other tools also read answers from the real ChatGPT interface? Yes. Peec AI says it uses browser automation through web interfaces ([source](https://docs.peec.ai/intro-to-peec-ai)), and Profound says its data reflects the consumer experience rather than API outputs ([source](https://www.tryprofound.com/features)). AEOmeter differs in that collection runs in your own Chrome under your own accounts. ### Which AI engines does AEOmeter track? ChatGPT, Claude, Perplexity and Google AI Overviews. See [how it works](https://aeometer.com/how-it-works) for how each check runs. ### How often is this page updated? We checked every competitor fact on October 16, 2026. We recheck before changing any comparison. --- - Web page: https://aeometer.com/alternatives - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AEOmeter vs Peec AI: a Peec AI alternative for founders > Comparing Peec AI alternatives? See who Peec AI suits, where AEOmeter differs and a dated, sourced comparison, including AEOmeter's limits. Last updated: 2026-10-16 **Short version.** [Peec AI](https://peec.ai/) is an AI search analytics platform for marketing teams and agencies. It runs your prompts daily on its own infrastructure across several AI engines and adds sentiment, position and recommendations. AEOmeter is narrower. It runs checks in your own Chrome, with your own AI accounts, across four engines, and shows the exact answers behind every number. Pick Peec if you want a hosted team tool. Pick AEOmeter if you are a founder who wants to read the evidence. > **Checked October 16, 2026** Everything this page says about Peec AI comes from Peec AI's own public pages: [peec.ai](https://peec.ai/), [pricing](https://peec.ai/pricing), [docs](https://docs.peec.ai/intro-to-peec-ai) and the [MCP page](https://peec.ai/mcp). We did not sign in to Peec AI or test its paid product. Plans and features change, so confirm anything that matters on their site. If something here is out of date, tell us and we will fix it. ## Who Peec AI is best for Peec calls itself "AI search analytics for marketing teams", and its site has separate tracks for brands and agencies. Its plans are aimed at SEO and content managers, SEO teams and marketing teams running several projects. - **It runs without you.** Peec's docs say it runs your prompts daily and reads answers through each engine's web interface, "exactly as real users do". Your laptop can be closed. - **It covers more engines.** Peec's site and docs name ChatGPT, Perplexity, Gemini, Google AI Mode and Copilot. - **It is built for teams.** Projects sit under a company account, with company-wide and per-project access levels, including read-only. - **It goes past measurement.** Peec shows sentiment and position, suggests prompts and offers action recommendations such as building review profiles. - **It plugs into reporting.** CSV exports, an API, a Looker Studio connector and an MCP server at `api.peec.ai/mcp`. If you run AI visibility for several brands, report to a marketing team or need data while nobody is at a computer, Peec is a strong choice. ## What AEOmeter does differently 1. **Answers from your own signed-in accounts.** The AEOmeter Chrome extension types each prompt into ChatGPT, Claude, Perplexity and Google in background tabs of your Chrome, signed in as you. It turns on each app's private mode where it can, waits for the answer to finish and saves the answer with its sources. Other tools also read answers from the consumer apps. The difference is that these run on your machine, under your accounts. 2. **Every number opens the answers behind it.** Mention rate is answers that mention your brand divided by answers we actually captured. Failed checks, blocked checks and searches with no AI Overview are left out of the denominator and shown separately. Click any rate and you get the list of answers, with the original text and cited sources. The [methodology](https://aeometer.com/methodology) defines each metric. 3. **Made for a founder tracking one brand.** Add your brand, its aliases and domains, and two or three competitors. AEOmeter suggests starter prompts like "best product analytics tool for a seed-stage startup" and queues a first ChatGPT run. Your first answers arrive once the extension is connected. 4. **Ask your AI assistant about your AI visibility.** AEOmeter has a remote [MCP server](https://aeometer.com/features/mcp) at `https://aeometer.com/mcp` and a REST API. Connect Claude or another MCP client and ask "which prompts mention PostHog but not us?" The tools use the same handlers as the web app. ## AEOmeter vs Peec AI at a glance Checked October 16, 2026. Peec AI facts link to the page they come from. "Not verified" means we could not confirm it on Peec's public pages. | | Peec AI | AEOmeter | | --- | --- | --- | | Built for | Marketing teams, SEO teams and agencies ([peec.ai](https://peec.ai/), [pricing](https://peec.ai/pricing)) | Founders tracking 1 to 3 of their own brands plus competitors | | How answers are collected | Browser automation through each engine's web interface, run by Peec ([docs](https://docs.peec.ai/intro-to-peec-ai)) | The AEOmeter extension in your own Chrome, signed in to your own AI accounts | | When checks run | Daily, according to Peec's docs ([docs](https://docs.peec.ai/intro-to-peec-ai)) | On your daily schedule, while your Chrome is open and connected | | AI engines | ChatGPT, Perplexity, Gemini, Google AI Mode and Copilot are named on its site and docs. Per-plan availability not verified ([peec.ai](https://peec.ai/), [docs](https://docs.peec.ai/intro-to-peec-ai)) | ChatGPT, Claude, Perplexity, Google AI Overviews | | Metrics | Visibility, position and sentiment ([peec.ai](https://peec.ai/)) | Mention rate, owned citation rate, tracked share of voice. No sentiment or position | | Answer-level evidence | Individual responses ("Chats") behind metrics ([docs](https://docs.peec.ai/understanding-chats)) | Every rate links to the original answers and their cited sources | | Source analysis | Cited domains and URLs ([docs](https://docs.peec.ai/understanding-sources)) | Sources by domain or page. Engine citations kept apart from links in the answer text | | Team access | Company and per-project roles, including read-only ([pricing](https://peec.ai/pricing)) | Organizations with members and invitations. No agency features | | API and MCP | API and MCP server ([MCP page](https://peec.ai/mcp)) | REST API and remote MCP server with OAuth | | Exports | CSV and a Looker Studio connector ([pricing](https://peec.ai/pricing), [project docs](https://docs.peec.ai/manage-your-project)) | CSV export of comparisons | | Price | Not verified. The plan prices did not appear in the page text we could read ([pricing](https://peec.ai/pricing)) | See [pricing](https://aeometer.com/pricing) | ## Choose Peec AI or AEOmeter ### Choose Peec AI if - You need checks to run every day with no computer of yours involved. - You need Gemini, Google AI Mode or Copilot. - You manage several brands or clients, or report to a team that wants roles and dashboards. - You want sentiment, position and suggested actions in the same tool. ### Choose AEOmeter if - You are a founder tracking your own product against two or three competitors. - You want answers from the same signed-in apps your buyers use, collected under your own accounts. - You want to open the exact answer behind every percentage, and see which checks failed instead of having them counted as misses. - You want to ask Claude or another MCP client about your AI visibility. - Your Chrome is open most of the working day anyway. ## Where AEOmeter falls short Read this part before you sign up. These are the trade-offs of how AEOmeter works today. - **Your Chrome has to be open.** Checks run in your own browser. A daily schedule queues checks, and they run when Chrome is open, the extension is connected and you are signed in to each AI app. Close the laptop and collection waits. [How it works](https://aeometer.com/how-it-works) explains the details. - **Four AI engines.** ChatGPT, Claude, Perplexity and Google AI Overviews. Gemini, Google AI Mode and Copilot are not tracked. - **It is in beta.** The extension, "AEOmeter BETA", is unlisted while Chrome Web Store review is pending. AI apps change their pages, and our adapters for each one are experimental. When a page changes, a check can fail until we update the adapter. - **Built for a few brands.** You track your own brand and its competitors. There are no agency features: no client workspaces, white-label reports or alerts. - **Fewer metrics.** You get mention rate, owned citation rate and tracked share of voice. There is no sentiment score, ranking position or search-volume estimate. - **Logins and CAPTCHAs pause checks.** If an AI app asks you to sign in or verify, the check pauses with a red "!" badge until you deal with it. AEOmeter never bypasses a CAPTCHA. ## Moving from Peec AI to AEOmeter Peec lets you export chats as CSV, and most teams keep their prompt list in a spreadsheet. AEOmeter imports prompts from CSV or plain text with a preview, so you can bring the same questions over and keep tags. History does not transfer. Your AEOmeter trend lines start from the first answers you collect. ## FAQ ### Is AEOmeter a Peec AI alternative? For founders, yes. Both track how AI answers mention your brand and competitors and which sources they cite. Peec is a hosted platform for marketing teams and agencies. AEOmeter runs in your own Chrome and is built for one person tracking a few brands. ### Does Peec AI use the real ChatGPT interface too? Yes. Peec's docs say it uses browser automation through the engines' web interfaces rather than APIs ([source](https://docs.peec.ai/intro-to-peec-ai)). Reading consumer apps is not unique to AEOmeter. What differs is where it happens: AEOmeter uses your Chrome and your signed-in accounts. ### Does AEOmeter track Gemini or Google AI Mode? No. AEOmeter tracks ChatGPT, Claude, Perplexity and Google AI Overviews. If you need Gemini or AI Mode, Peec names both on its site. ### Will AEOmeter collect answers while my laptop is closed? No. Checks run only while your Chrome is open, the extension is connected and you are signed in to each AI app. Scheduled checks wait until then. AEOmeter shows the next scheduled time and the last actual capture separately so you can see the gap. ### Can I keep using Peec and AEOmeter together? Yes. Some people keep a hosted tool for daily coverage and use AEOmeter to read individual answers from their own accounts. The numbers will not match exactly because each tool samples differently. ### How much does AEOmeter cost? See the [pricing page](https://aeometer.com/pricing). This page makes no price claims about AEOmeter. --- - Web page: https://aeometer.com/alternatives/peec-ai - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AEOmeter vs Profound: a Profound alternative for founders > Looking at Profound alternatives? An honest, dated comparison of Profound and AEOmeter, who each one suits, and where AEOmeter falls short. Last updated: 2026-10-16 **Short version.** [Profound](https://www.tryprofound.com/) is a broad AI marketing platform for enterprise and agency teams. It covers up to nine answer engines on its Enterprise plan, plus prompt demand data, crawler analytics and content agents. AEOmeter does one job for founders: it runs your prompts in your own Chrome across four AI apps and shows the answers behind every number. Pick Profound if you are a large marketing team. Pick AEOmeter if you want a focused tool you can set up alone this afternoon. > **Checked October 16, 2026** Everything this page says about Profound comes from Profound's own public pages: [tryprofound.com](https://www.tryprofound.com/), [features](https://www.tryprofound.com/features), [pricing](https://www.tryprofound.com/pricing) and the [changelog](https://product.tryprofound.com/changelog). We did not sign in to Profound or test its paid product. Plans and features change, so confirm anything that matters on their site. If something here is out of date, tell us and we will fix it. ## Who Profound is best for Profound sells to AEO, content and PR teams, with programs for agencies and pages for travel, retail and pharmaceutical companies. It says its enterprise capabilities are "trusted by a third of the Fortune 100". It is a serious platform with a lot of surface area. - **Wide engine coverage.** The Enterprise plan lists up to nine answer engines: ChatGPT, Perplexity, Google AI Mode, Gemini, Microsoft Copilot, DeepSeek, Claude, Google AI Overviews and Exa Search. - **Data from the consumer experience.** Profound says its insights come from what consumers see, "not API outputs". - **More than monitoring.** Prompt Volumes shows what people ask AI. Agent Analytics tracks AI crawlers on your site. Agents and an AI Marketer help produce content. Shopping tracks products in ChatGPT Shopping. - **Enterprise requirements.** SSO/SAML and SOC 2 on Enterprise, unlimited seats, dedicated support options and an API. - **Shipping fast.** The public changelog has weekly entries. An October 2, 2026 entry introduced the Profound MCP. If you have a marketing team, several markets and a budget conversation with sales, Profound can cover far more ground than AEOmeter. ## What AEOmeter does differently 1. **Answers from your own signed-in accounts.** The AEOmeter Chrome extension types each prompt into ChatGPT, Claude, Perplexity and Google in background tabs of your Chrome, signed in as you. It turns on each app's private mode where it can, waits for the answer to finish and saves the answer with its sources. Other tools also read answers from the consumer apps. The difference is that these run on your machine, under your accounts. 2. **Every number opens the answers behind it.** Mention rate is answers that mention your brand divided by answers we actually captured. Failed checks, blocked checks and searches with no AI Overview are left out of the denominator and shown separately. Click any rate and you get the list of answers, with the original text and cited sources. The [methodology](https://aeometer.com/methodology) defines each metric. 3. **Made for a founder tracking one brand.** Add your brand, its aliases and domains, and two or three competitors. AEOmeter suggests starter prompts like "best product analytics tool for a seed-stage startup" and queues a first ChatGPT run. Your first answers arrive once the extension is connected. 4. **Ask your AI assistant about your AI visibility.** AEOmeter has a remote [MCP server](https://aeometer.com/features/mcp) at `https://aeometer.com/mcp` and a REST API. Connect Claude or another MCP client and ask "which prompts mention PostHog but not us?" The tools use the same handlers as the web app. ## AEOmeter vs Profound at a glance Checked October 16, 2026. Profound facts link to the page they come from. "Not verified" means we could not confirm it on Profound's public pages. | | Profound | AEOmeter | | --- | --- | --- | | Built for | AEO, content and PR teams, agencies and enterprises ([tryprofound.com](https://www.tryprofound.com/)) | Founders tracking 1 to 3 of their own brands plus competitors | | How answers are collected | From the consumer experience, "not API outputs". Exact method not verified ([features](https://www.tryprofound.com/features)) | The AEOmeter extension in your own Chrome, signed in to your own AI accounts | | When checks run | Daily on Enterprise. The trial runs 50 prompts daily for 7 days ([pricing](https://www.tryprofound.com/pricing)) | On your daily schedule, while your Chrome is open and connected | | AI engines | Trial: ChatGPT, Gemini, Google AI Overviews. Enterprise: up to 9 engines including Claude, Copilot and AI Mode ([pricing](https://www.tryprofound.com/pricing)) | ChatGPT, Claude, Perplexity, Google AI Overviews | | Metrics | Visibility, sentiment, citations and competitive benchmarking ([features](https://www.tryprofound.com/features)) | Mention rate, owned citation rate, tracked share of voice. No sentiment | | Prompt demand data | Prompt Volumes ([features](https://www.tryprofound.com/features)) | None. You choose the prompts | | Crawler analytics | Agent Analytics ([features](https://www.tryprofound.com/features)) | None | | Content creation | Agents and AI Marketer ([features](https://www.tryprofound.com/features)) | None | | Security and seats | SSO/SAML and SOC 2 on Enterprise. Unlimited seats ([pricing](https://www.tryprofound.com/pricing)) | Organizations with members and invitations. No SAML SSO | | API and MCP | API on Enterprise ([pricing](https://www.tryprofound.com/pricing)). Profound MCP announced October 2, 2026 ([changelog](https://product.tryprofound.com/changelog)) | REST API and remote MCP server with OAuth | | Price | Free trial. Enterprise priced on request. A self-serve Agency Growth plan is listed without a price ([pricing](https://www.tryprofound.com/pricing)) | See [pricing](https://aeometer.com/pricing) | ## Choose Profound or AEOmeter ### Choose Profound if - You need many answer engines, regions and languages under one contract. - You want prompt demand data, crawler analytics and content workflows in the same platform. - Your security review requires SSO, SAML or SOC 2. - You run AI visibility for many clients or a large brand portfolio. ### Choose AEOmeter if - You are a founder or a small team and you mainly want to know whether ChatGPT, Claude, Perplexity and Google mention you. - You want to set it up yourself in a few minutes, without a sales call. - You want answers collected under your own signed-in accounts, on your machine. - You want every rate to open the answers behind it, with failed checks shown instead of counted as zero. - You want to query your data from Claude or another MCP client. ## Where AEOmeter falls short Read this part before you sign up. These are the trade-offs of how AEOmeter works today. - **Your Chrome has to be open.** Checks run in your own browser. A daily schedule queues checks, and they run when Chrome is open, the extension is connected and you are signed in to each AI app. Close the laptop and collection waits. [How it works](https://aeometer.com/how-it-works) explains the details. - **Four AI engines.** ChatGPT, Claude, Perplexity and Google AI Overviews. Gemini, Google AI Mode and Copilot are not tracked. - **It is in beta.** The extension, "AEOmeter BETA", is unlisted while Chrome Web Store review is pending. AI apps change their pages, and our adapters for each one are experimental. When a page changes, a check can fail until we update the adapter. - **Built for a few brands.** You track your own brand and its competitors. There are no agency features: no client workspaces, white-label reports or alerts. - **Fewer metrics.** You get mention rate, owned citation rate and tracked share of voice. There is no sentiment score, ranking position or search-volume estimate. - **Logins and CAPTCHAs pause checks.** If an AI app asks you to sign in or verify, the check pauses with a red "!" badge until you deal with it. AEOmeter never bypasses a CAPTCHA. ## FAQ ### Is AEOmeter a Profound alternative? For founders and small teams who mainly need answer monitoring, yes. AEOmeter does not replace Profound's demand data, crawler analytics, content agents or enterprise controls. ### Does Profound track Claude? Profound's pricing page lists Claude among the up to nine engines on its Enterprise plan. The trial lists ChatGPT, Gemini and Google AI Overviews ([source](https://www.tryprofound.com/pricing)). AEOmeter tracks Claude as one of its four engines. ### Does AEOmeter have prompt volume or search demand data? No. AEOmeter measures the prompts you choose. It does not estimate how often people ask them. Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) covers how to pick them. ### Does AEOmeter run checks in the cloud? No. Checks run in your own Chrome while it is open and connected. That keeps collection cheap and uses your own accounts, but scheduled checks wait whenever your browser is closed. ### How much does AEOmeter cost? See the [pricing page](https://aeometer.com/pricing). This page makes no price claims about AEOmeter. --- - Web page: https://aeometer.com/alternatives/profound - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AEOmeter vs Otterly.AI: an Otterly.AI alternative for founders > Comparing Otterly AI alternatives? A sourced, dated comparison of Otterly.AI and AEOmeter, with who each suits and AEOmeter's honest limits. Last updated: 2026-10-16 **Short version.** [Otterly.AI](https://otterly.ai/) is a self-serve AI search monitoring tool for marketers, SEO teams and agencies, with published prices and plans starting at a solo tier. It runs your prompts automatically and adds sentiment, GEO audits and crawler analytics. AEOmeter runs checks in your own Chrome, with your own signed-in AI accounts, and links every number to the answers behind it. Pick Otterly.AI for hosted monitoring with a lot of features. Pick AEOmeter if you want to read the evidence from your own accounts. > **Checked October 16, 2026** Everything this page says about Otterly.AI comes from Otterly.AI's own public pages: [otterly.ai](https://otterly.ai/), [pricing](https://otterly.ai/pricing) and the [help center](https://help.otterly.ai/search-prompt-monitoring). We did not sign in to Otterly.AI or test its paid product. Plans and features change, so confirm anything that matters on their site. If something here is out of date, tell us and we will fix it. ## Who Otterly.AI is best for Otterly.AI targets marketing professionals, SEO and content teams and agencies. Its entry plan is described as "for solo marketers, and small teams". It is one of the few tools in this category with prices on its website. - **Clear, public pricing.** Plans and prompt counts are listed on the pricing page, with a free trial for new users. - **Hosted monitoring.** Otterly.AI sends your prompts to AI engines automatically. Your computer does not need to be on. - **Engine choice.** Every plan includes ChatGPT, Google AI Overviews, Perplexity and Microsoft Copilot. Google AI Mode, Gemini and Claude are paid add-ons. - **Extras around monitoring.** Sentiment, a GEO content audit, Agent Analytics for crawler activity, a Looker Studio connector, and API and MCP access on Standard plans and above. - **Unlimited team members** on every plan, according to the pricing page. If you want a hosted tool with a known monthly price, or you run monitoring for clients, Otterly.AI is a sensible pick. ## What AEOmeter does differently 1. **Answers from your own signed-in accounts.** The AEOmeter Chrome extension types each prompt into ChatGPT, Claude, Perplexity and Google in background tabs of your Chrome, signed in as you. It turns on each app's private mode where it can, waits for the answer to finish and saves the answer with its sources. Other tools also read answers from the consumer apps. The difference is that these run on your machine, under your accounts. 2. **Every number opens the answers behind it.** Mention rate is answers that mention your brand divided by answers we actually captured. Failed checks, blocked checks and searches with no AI Overview are left out of the denominator and shown separately. Click any rate and you get the list of answers, with the original text and cited sources. The [methodology](https://aeometer.com/methodology) defines each metric. 3. **Made for a founder tracking one brand.** Add your brand, its aliases and domains, and two or three competitors. AEOmeter suggests starter prompts like "best product analytics tool for a seed-stage startup" and queues a first ChatGPT run. Your first answers arrive once the extension is connected. 4. **Ask your AI assistant about your AI visibility.** AEOmeter has a remote [MCP server](https://aeometer.com/features/mcp) at `https://aeometer.com/mcp` and a REST API. Connect Claude or another MCP client and ask "which prompts mention PostHog but not us?" The tools use the same handlers as the web app. ## AEOmeter vs Otterly.AI at a glance Checked October 16, 2026. Otterly.AI facts link to the page they come from. "Not verified" means we could not confirm it on Otterly.AI's public pages. | | Otterly.AI | AEOmeter | | --- | --- | --- | | Built for | Marketers, SEO and content teams, agencies. Entry plan "for solo marketers, and small teams" ([otterly.ai](https://otterly.ai/), [pricing](https://otterly.ai/pricing)) | Founders tracking 1 to 3 of their own brands plus competitors | | How answers are collected | Prompts sent to AI engines automatically by Otterly.AI. Exact method not verified ([otterly.ai](https://otterly.ai/)) | The AEOmeter extension in your own Chrome, signed in to your own AI accounts | | When checks run | Automatically, on Otterly.AI's side. Exact frequency not verified ([otterly.ai](https://otterly.ai/)) | On your daily schedule, while your Chrome is open and connected | | AI engines | Included: ChatGPT, Google AI Overviews, Perplexity, Microsoft Copilot. Add-ons: Google AI Mode, Gemini, Claude ([pricing](https://otterly.ai/pricing)) | ChatGPT, Claude, Perplexity, Google AI Overviews | | Metrics | Brand coverage, sentiment, citations, competitor benchmarking ([help](https://help.otterly.ai/search-prompt-monitoring)) | Mention rate, owned citation rate, tracked share of voice. No sentiment | | Answer-level evidence | Prompt detail drills into individual responses ([help](https://help.otterly.ai/search-prompt-monitoring)) | Every rate links to the original answers and their cited sources | | Editing a prompt | Changing wording means adding a new prompt; history does not carry over ([help](https://help.otterly.ai/relevant-prompts)) | Editing starts a new baseline so trends compare like with like | | Extras | GEO audits, Agent Analytics, Looker Studio connector ([pricing](https://otterly.ai/pricing)) | None of these | | API and MCP | API and MCP requests on Standard plans and above ([pricing](https://otterly.ai/pricing)) | REST API and remote MCP server with OAuth | | Price | Lite plan listed at $29/month for 15 search prompts. Free trial offered ([pricing](https://otterly.ai/pricing)) | See [pricing](https://aeometer.com/pricing) | ## Choose Otterly.AI or AEOmeter ### Choose Otterly.AI if - You want checks to run without your computer. - You want Microsoft Copilot, or the option to add Gemini or Google AI Mode. - You want sentiment, GEO audits or crawler analytics alongside monitoring. - You run monitoring for several clients or workspaces. ### Choose AEOmeter if - You are a founder tracking your own brand and a few competitors. - You want Claude tracked as one of the core engines. - You want answers collected in your own Chrome, under your own signed-in accounts. - You want every rate to show its denominator and open the answers behind it. - You want to ask Claude or another MCP client about your AI visibility. ## Where AEOmeter falls short Read this part before you sign up. These are the trade-offs of how AEOmeter works today. - **Your Chrome has to be open.** Checks run in your own browser. A daily schedule queues checks, and they run when Chrome is open, the extension is connected and you are signed in to each AI app. Close the laptop and collection waits. [How it works](https://aeometer.com/how-it-works) explains the details. - **Four AI engines.** ChatGPT, Claude, Perplexity and Google AI Overviews. Gemini, Google AI Mode and Copilot are not tracked. - **It is in beta.** The extension, "AEOmeter BETA", is unlisted while Chrome Web Store review is pending. AI apps change their pages, and our adapters for each one are experimental. When a page changes, a check can fail until we update the adapter. - **Built for a few brands.** You track your own brand and its competitors. There are no agency features: no client workspaces, white-label reports or alerts. - **Fewer metrics.** You get mention rate, owned citation rate and tracked share of voice. There is no sentiment score, ranking position or search-volume estimate. - **Logins and CAPTCHAs pause checks.** If an AI app asks you to sign in or verify, the check pauses with a red "!" badge until you deal with it. AEOmeter never bypasses a CAPTCHA. ## Moving from Otterly.AI to AEOmeter Copy your search prompts into a CSV or a plain-text list and import them into AEOmeter. You get a preview before anything is saved, and you can tag prompts on the way in. History does not transfer, so your AEOmeter trends start from your first captured answers. ## FAQ ### Is AEOmeter an Otterly.AI alternative? For founders who want to read answers from their own accounts, yes. Both track brand mentions and cited sources in AI answers. Otterly.AI is hosted and adds audits and crawler analytics. AEOmeter runs in your Chrome and stays focused on answers and evidence. ### How much does Otterly.AI cost? On October 16, 2026, Otterly.AI's pricing page listed the Lite plan at $29/month for 15 search prompts, with higher tiers and paid engine add-ons ([source](https://otterly.ai/pricing)). Check the page for current prices. ### Does AEOmeter track Microsoft Copilot or Gemini? No. AEOmeter tracks ChatGPT, Claude, Perplexity and Google AI Overviews. ### Does AEOmeter measure sentiment? No. AEOmeter reports whether you were mentioned and cited, and how often against tracked competitors. You can read each answer to judge the tone yourself. ### How much does AEOmeter cost? See the [pricing page](https://aeometer.com/pricing). This page makes no price claims about AEOmeter. --- - Web page: https://aeometer.com/alternatives/otterly-ai - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AI visibility guides > Practical guides to AI visibility and answer engine optimization: what to measure, how to show up in ChatGPT and other assistants, and how to track it. ## All guides - [What is AI visibility?](https://aeometer.com/guides/ai-visibility): AI visibility is how often AI assistants mention and cite your brand for the questions your buyers ask. What it means, what you can measure, and what you can’t. - [Answer engine optimization](https://aeometer.com/guides/answer-engine-optimization): What answer engine optimization is, how AEO differs from SEO and GEO, the steps a founder can take this month, and how to tell whether they are working. - [How to rank in ChatGPT](https://aeometer.com/guides/how-to-rank-in-chatgpt): There is no fixed rank in ChatGPT. Here is what influences whether it mentions your brand, a practical checklist for founders, and how to tell if it is working. - [Track brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt): A step-by-step method to track how often ChatGPT mentions your brand and competitors: a free spreadsheet approach, a sample tracking table, and how to automate it. - [AI share of voice](https://aeometer.com/guides/ai-share-of-voice): What AI share of voice means, the formula with a worked example, the pitfalls that make it misleading, and how AEOmeter calculates tracked share of voice. - [Choosing prompts](https://aeometer.com/guides/choosing-prompts): How to pick the buyer prompts you track in ChatGPT, Claude, Perplexity and AI Overviews: prompt types, how many, keeping them stable, tags, and an example set. ## Where to start | If you want to… | Read | | --- | --- | | Understand what AI visibility is and what can be measured | [What is AI visibility?](https://aeometer.com/guides/ai-visibility) | | Know what to actually do to show up in AI answers | [Answer engine optimization](https://aeometer.com/guides/answer-engine-optimization) | | Get your brand mentioned by ChatGPT | [How to rank in ChatGPT](https://aeometer.com/guides/how-to-rank-in-chatgpt) | | Start tracking this week with a spreadsheet | [Track brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt) | | Compare yourself with competitors | [AI share of voice](https://aeometer.com/guides/ai-share-of-voice) | | Pick the questions you track | [Choosing prompts](https://aeometer.com/guides/choosing-prompts) | ## How these guides are written They are written for founders who do their own marketing. They describe what can be measured and what can’t, use fictional example data that is labelled as such, and avoid promises about rankings or results. Metric definitions match the [methodology](https://aeometer.com/methodology) page, and terms are defined in the [glossary](https://aeometer.com/glossary). --- - Web page: https://aeometer.com/guides - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # What is AI visibility? > AI visibility is how often AI assistants mention and cite your brand for the questions your buyers ask. What it means, what you can measure, and what you can’t. Last updated: 2026-10-16 Ask ChatGPT "what’s the best product analytics tool for a seed-stage startup?" and you get a short answer that names three or four products. **AI visibility** is whether your product is one of them, how often that happens across the questions your buyers ask, and which sources the assistant points to when it answers. This guide defines the term, explains why it matters for a small company, and draws a clear line between what you can measure and what nobody can. It ends with two ways to measure it: by hand this week, or with a tool. ## AI visibility, defined **AI visibility is the share of relevant AI answers that mention or cite your brand.** "Relevant" means answers to questions a potential buyer would ask, such as "best CRM for a two-person agency" or "Notion alternatives for engineering docs". "AI answers" means the responses people read in assistants like ChatGPT, Claude and Perplexity, and in Google’s AI Overviews above search results. In practice it has three parts: - **Presence.** Does the answer name your brand at all? Measured as a [mention rate](https://aeometer.com/glossary#mention-rate). - **Evidence.** Does the answer link to your own site as a source? Measured as an [owned citation rate](https://aeometer.com/glossary#owned-citation-rate), plus the list of other domains the assistant cites. - **Position against competitors.** When the assistant talks about your category, how much of that talk is about you compared with the brands you compete with? Measured as [tracked share of voice](https://aeometer.com/guides/ai-share-of-voice). > **Short version** AI visibility = for the questions my buyers ask, how often do AI assistants name me, link to me, and name me instead of my competitors. ## Why AI visibility matters for founders A growing share of software research starts in a chat box. A buyer types a question, reads a shortlist and opens one or two of the names. If you are not on the list, you never learn that buyer existed. Nothing shows up in your analytics, and Google Search Console has no report for it. The answer is also a summary of what the web says about your category. If review sites, Reddit threads and comparison posts describe you as "an enterprise analytics suite" while you sell to seed-stage teams, the assistant repeats that description to the exact people you want to reach. You can only fix a description you have seen. Three things make this worth tracking even for a two-person company: - Shortlists form before anyone visits your site, so your landing page never gets a chance to make the case. - Competitors who get named keep getting named, because the pages that mention them get cited and linked more often. - You can act on it. The sources an assistant cites for a prompt are a concrete list of places to earn a mention, fix a description or publish a better page. ## What you can measure Everything measurable in AI visibility comes from sampling: you ask a fixed set of questions, save the answers and count what is in them. These are the numbers that hold up. | Metric | Question it answers | How it is counted | | --- | --- | --- | | Mention rate | How often does the assistant name us? | Answers that mention any name or alias of your brand ÷ answers successfully captured | | Owned citation rate | How often does it link to our site? | Answers that cite at least one URL on your domain ÷ answers successfully captured | | Cited domains | Which sites shape answers in our category? | Sources attributed by the assistant, grouped by domain or page | | Tracked share of voice | How much of the category talk is about us? | Your mentions ÷ mentions of every brand you track, in the same answers | | Gaps | Where do competitors appear without us? | Answers where a tracked competitor is mentioned and you are not | The denominator matters as much as the number. An answer that failed to load, hit a login wall or returned no AI Overview tells you nothing about your brand, so it should be left out instead of counted as a miss. Our [methodology](https://aeometer.com/methodology) page spells out exactly how AEOmeter counts each one. ## What you can’t measure Plenty of dashboards show numbers that look like SEO metrics. Be skeptical of these: - **Impressions.** No AI provider publishes how many people asked a question or how many times your name appeared in their answers. Nobody outside OpenAI, Anthropic, Perplexity or Google can see other people’s chats. An "AI impressions" figure is an estimate stacked on assumptions. - **A global rank.** Answers change with the exact wording, the engine, the run, the account’s memory and settings, and location. There is no fixed "position 1 in ChatGPT" that every user sees. Our guide on [how to rank in ChatGPT](https://aeometer.com/guides/how-to-rank-in-chatgpt) goes deeper on this. - **Prompt search volume.** Providers don’t publish what people type. Keyword volume from SEO tools is a useful proxy for demand, nothing more. - **Cause and effect.** If an answer mentions you and cites a review site, that does not prove the review site caused the mention. Treat cited sources as strong hints about where to look. What remains is honest and useful: a consistent sample of real answers, repeated over time, with every number traceable to the answer it came from. ## How AI visibility differs from search rankings | | Classic search | AI answers | | --- | --- | --- | | What the user sees | A ranked list of links | A written answer naming a few brands, sometimes with sources | | Stability | Positions move slowly | The same prompt can produce a different answer on the next run | | First-party data | Search Console shows impressions and clicks | None from the providers | | What you track | Position per keyword | Mentions, citations and share across a prompt set | | Main lever | Your own pages ranking | What many sources say about you, including your own pages | The two overlap. Google says its AI features are rooted in its core Search ranking systems, and ChatGPT, Claude and Perplexity can all search the web before answering. Pages that rank well and get referenced elsewhere are more likely to be read when an answer is put together, so good SEO work still pays off here. ## How to measure AI visibility by hand You can get a first baseline in an afternoon with a spreadsheet. 1. **Write 10 to 20 buyer prompts.** Mix category questions ("best product analytics tool for a seed-stage startup"), comparisons ("Mixpanel vs PostHog for a B2B SaaS") and alternatives ("Amplitude alternatives for small teams"). Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) has a full method. 2. **List your brand and three competitors.** Include every name the assistant might use: product name, company name, common misspellings and your domain. 3. **Use a clean session.** In ChatGPT, turn on temporary chat so memory and past chats don’t color the answer. In other assistants, use their private or incognito mode where it exists, or a logged-out window. 4. **Ask each prompt in each assistant.** Paste the exact same wording every time. Save the answer text and every source link it shows. 5. **Record what you see.** One row per prompt and assistant: mentioned yes or no, which competitors were named, which URLs were cited and whether any of them are on your domain. 6. **Repeat on a schedule.** Run the same set weekly with identical wording. Change a prompt and that row starts a new baseline. Twenty prompts across four assistants is 80 answers per round. That is enough to see patterns and small enough to do by hand a few times. The [free AI visibility checker](https://aeometer.com/tools/ai-visibility-checker) gives you prompts and a scoring sheet for exactly this, and our guide to [tracking brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt) shows the full spreadsheet. ## How to measure it with a tool A tracking tool runs the same loop for you on a schedule. When you compare tools, ask four questions: 1. **Where do the answers come from?** Some tools query model APIs. API output can differ from what people see in the consumer apps, which add web search, memory and their own interface. 2. **Can you read the original answer?** Every number should open the answer text and sources it was counted from. 3. **How are failures counted?** A blocked or timed-out check should be excluded and shown, never silently counted as "not mentioned". 4. **What happens when you edit a prompt?** Trends should only compare answers to the same wording. [AEOmeter](https://aeometer.com/product) runs your prompts in the consumer apps of ChatGPT, Claude, Perplexity and Google AI Overviews through a Chrome extension, signed in with your own accounts, using each engine’s private mode where it can. Every metric links back to the answer and its cited sources. The trade-off is that your Chrome has to be open for scheduled checks to run. [How it works](https://aeometer.com/how-it-works) covers the details. ## What to do with the numbers - **Low mention rate on category prompts:** open the answers and list the domains cited. Those review sites, directories and comparison posts are where your category gets described. - **A competitor appears and you don’t:** read that answer’s sources. If a "best tools" list includes them and not you, that is a specific page to get onto. - **You are mentioned but described wrong:** find the page that should state the fact plainly, usually your homepage, pricing page or docs, and fix it there first. - **Mentioned but never cited:** publish pages an assistant can quote, like clear comparisons and use-case pages. The [answer engine optimization guide](https://aeometer.com/guides/answer-engine-optimization) walks through this. ## FAQ ### Is AI visibility the same as AEO? They are related. [Answer engine optimization](https://aeometer.com/guides/answer-engine-optimization) is the work you do to appear in AI answers. AI visibility is the result you measure to see whether that work is landing. ### Can I see how many people saw my brand in ChatGPT? No. OpenAI and the other providers do not share impression or query data with brands. You can measure how often your brand appears in a sample of answers, which is a different and more honest number. ### Why does the same prompt give different answers? AI assistants generate text with some randomness, may run a fresh web search each time and adjust to account settings. That is why you track rates over many answers instead of one screenshot. ### Which AI assistants should I track? The ones your buyers use. For most software founders that means ChatGPT, plus Google AI Overviews because they appear in ordinary search. Perplexity and Claude are worth adding if your buyers are technical or research-heavy. See our [engine pages](https://aeometer.com/engines) for each one. ### How often should I measure? Weekly is enough by hand. With a tool, a daily schedule gives you more answers to average over. Either way, keep prompt wording fixed so you compare like with like. --- - Web page: https://aeometer.com/guides/ai-visibility - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Answer engine optimization (AEO): a practical guide > What answer engine optimization is, how AEO differs from SEO and GEO, the steps a founder can take this month, and how to tell whether they are working. Last updated: 2026-10-16 **Answer engine optimization (AEO)** is the work of making your brand show up, accurately, in the answers AI assistants give. The answer engines that matter today are ChatGPT, Claude, Perplexity and Google’s AI Overviews. Each one reads the web, writes a short answer and sometimes lists its sources. AEO is about being in that answer and being described the way you would describe yourself. This guide is for founders who sell software and do their own marketing. It covers what AEO is, how it relates to SEO and GEO, a practical set of steps, and how to measure the result without fooling yourself. It makes no promises about outcomes, because nobody can honestly make them. ## What is answer engine optimization? An answer engine returns an answer instead of a list of links. The term existed before chatbots: SEOs used it for featured snippets and voice assistants. It now mostly refers to AI assistants that generate answers, often after running a web search of their own. AEO has two goals that are easy to blur: - **Get mentioned.** When someone asks "best uptime monitoring tool for a small SaaS", your product is one of the names in the answer. - **Get cited.** The assistant links to a page on your site, or to a page that describes you well, as a source for what it says. A mention puts you on the shortlist. A citation gives the reader a path to you and tells you which page the assistant trusted. Track them separately, because you can have one without the other. ## AEO vs SEO vs GEO You will see several names for overlapping ideas. Here is how they differ in practice. | | SEO | AEO | GEO | | --- | --- | --- | --- | | Stands for | Search engine optimization | Answer engine optimization | Generative engine optimization | | Goal | Rank pages in search results | Appear and be described correctly in AI answers | Same as AEO; the term comes from academic research | | Unit of success | A position for a keyword | A mention or citation in an answer | A mention or citation in an answer | | Main inputs | Your pages, links, technical health | Your pages plus what the rest of the web says about you | Same as AEO | | How you measure | Search Console, rank trackers | Sampling answers to a fixed prompt set | Same as AEO | GEO and AEO describe the same work, and LLMO (large language model optimization) is a third name for it. Pick whichever term your team uses. SEO is the foundation under all of them: Google says its AI features rely on the same ranking systems as Search, and the other assistants search the web when they need current information. A page that cannot be crawled or found will not be cited. ## How answer engines pick what to say You can’t see inside the models, but the broad mechanics are public and stable enough to plan around. - **Training data.** Models learn from a large snapshot of the web. If your category has been written about for years and you are rarely mentioned in it, the model starts with a weak picture of you. - **Live web search (grounding).** For questions about products and recommendations, assistants often search first and then write from the results. Pages that are findable and say clearly what a product does have a better chance of being used. - **The consensus of sources.** Answers tend to reflect what many sources agree on. One page on your own site calling you "the best" carries less weight than ten independent pages that describe you the same way. - **The prompt itself.** "Best analytics tool" and "best analytics tool for a seed-stage B2B startup" pull different sources and different names. ## Practical AEO steps for founders These are ordered by how much they tend to matter for a small company. None of them is a trick. Most are good marketing that also happens to feed AI answers. ### 1. Find out where you stand Pick 10 to 20 questions your buyers ask and check them in each assistant. Note who gets named and which sources are cited. Without this baseline you can’t tell later whether anything changed. Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) shows how to build the set, and the [free AI visibility checker](https://aeometer.com/tools/ai-visibility-checker) gives you a template. ### 2. Get onto the sources that are already cited For each prompt, list the domains the assistant cites. In software categories these are often review and directory sites, "best X tools" roundups, comparison posts, Reddit and community threads, and documentation of tools you integrate with. Each one is a specific opportunity: - Ask the author of a roundup to review your product. Offer a trial account and a short, factual summary. - Create or complete your profiles on the directories and review sites that show up in citations. - Answer real questions in communities where you are a genuine participant. Disclose that you are the founder. - Get listed in the integration directories of tools you connect with. Citations show you which pages were used for an answer. They don’t prove those pages caused a mention, so treat this list as a strong hint about where to work. ### 3. Publish clear positioning and comparison pages Assistants answer comparison questions constantly. Give them something accurate to read: - A homepage that says in the first two sentences what the product is, who it is for and what it replaces. - Honest "you vs competitor" and "competitor alternatives" pages with a comparison table and a section on when the competitor is the better choice. - Use-case pages that match how buyers phrase their situation ("analytics for a two-person B2B SaaS", "analytics without a data engineer"). - A pricing page with plain numbers and limits, if you publish pricing. Write each key fact as a short, self-contained paragraph that still makes sense if it is quoted alone. That helps human readers too. ### 4. Keep your entity information consistent Use the same product name, one-line description, category and founder names everywhere: your site, social profiles, directories, app marketplaces and press. If your product was renamed, say so on your site ("Acme Analytics, formerly Acme Metrics"). Inconsistent descriptions lead to answers that mix you up with another company or describe an old version of you. ### 5. Make your docs and site easy for machines to read - Serve content as HTML that loads without client-side JavaScript, with descriptive headings. - Check that your robots.txt does not block the crawlers used for AI search. OpenAI, for example, documents a crawler called OAI-SearchBot for ChatGPT search. Blocking a training crawler and blocking a search crawler are separate decisions. - Keep docs public and current, with one page per topic and the product’s limits written down. - Add an `llms.txt` file if it is cheap for you to do. > **Be realistic about llms.txt** `llms.txt` is a proposed convention from 2024: a Markdown file at your site root that lists your important pages for language models. It is useful for coding agents and tools that look for it. None of the major AI assistants has documented that it uses `llms.txt` to decide which brands to mention or cite, and Google says AI Overviews need no special files. Add one if it takes an hour. Don’t expect it to move your numbers. ### 6. Fix wrong descriptions at the source When an assistant gets something wrong about you, like an old price, a missing feature or the wrong audience, find the page that should state the correct fact. Usually that is your own homepage, pricing or docs page, so fix it there first. Then look at the cited third-party pages and ask for corrections where they are out of date. ## What to skip - Pages written only for AI, stuffed with question-and-answer blocks nobody would read. Google treats mass-produced pages as spam, and readers bounce. - Hidden text or prompt-injection tricks aimed at AI crawlers. They are easy to detect and can get a site excluded. - Anyone who promises a position in ChatGPT. There is no fixed position to promise. ## How to measure AEO AEO work is slow to show up, because assistants re-read the web on their own schedule and answers vary from run to run. Measure it the way you would measure anything noisy: the same questions, the same conditions, many answers, compared over time. | Metric | What it tells you | What it does not tell you | | --- | --- | --- | | [Mention rate](https://aeometer.com/glossary#mention-rate) | How often you are named in the answers you sampled | How many real people saw those answers | | [Owned citation rate](https://aeometer.com/glossary#owned-citation-rate) | How often your own site is a cited source | Whether the citation caused the mention | | [Tracked share of voice](https://aeometer.com/guides/ai-share-of-voice) | Your share of mentions among the brands you track | Your market share | | [Gaps](https://aeometer.com/glossary#gap) | Specific answers where a competitor appears and you don’t | Why the assistant chose them | Exclude failed or blocked checks from the denominator, and start a new baseline whenever you reword a prompt. Our [methodology](https://aeometer.com/methodology) page explains how AEOmeter handles both. To run the loop automatically across ChatGPT, Claude, Perplexity and Google AI Overviews, see [the product tour](https://aeometer.com/product). AEOmeter runs the checks in your own Chrome, so your browser needs to be open while they run. ## FAQ ### Is AEO replacing SEO? No. AI answers lean on search results and on pages that rank, so SEO remains the foundation. AEO adds a new place to measure and a stronger focus on what third-party sources say about you. ### How long does AEO take to work? It varies by engine and category, and nobody can give you a reliable timeline. Changes that depend on live web search can show up sooner than changes that depend on model training. Track weekly and look at trends over a couple of months, not single answers. ### Do I need an llms.txt file? It is optional. It costs little and helps some agent tools, but no major assistant has documented using it to choose which brands to mention. ### What are AEO tools? Mostly trackers: they send a set of prompts to AI assistants on a schedule and report mentions, citations and competitor share. Some also suggest content. Check whether a tool reads the consumer apps or model APIs, and whether you can open the original answers. See [AEOmeter’s product tour](https://aeometer.com/product) for how we do it. ### Can I do AEO without a tool? Yes. A spreadsheet and an hour a week is enough to start. Our guide on [tracking brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt) shows the manual method step by step. --- - Web page: https://aeometer.com/guides/answer-engine-optimization - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # How to rank in ChatGPT > There is no fixed rank in ChatGPT. Here is what influences whether it mentions your brand, a practical checklist for founders, and how to tell if it is working. Last updated: 2026-10-16 People search for "how to rank in ChatGPT" because they want what SEO gave them: a position they can climb. ChatGPT doesn’t work that way. It writes a new answer each time, and the brands it names depend on the question, the sources it reads and some randomness. What you can do is make it more likely that your brand is named, and measure how often that happens. This guide explains what influences ChatGPT’s answers, gives you a checklist you can work through, and shows how to track whether any of it is moving the needle. ## Is there a ranking in ChatGPT? **No. ChatGPT has no fixed ranking of brands.** Ask "best email marketing tool for a Shopify store" twice and you may get the same four names in a different order, or a different fourth name. Two people asking the same question can get different answers because of memory, custom instructions, location and the exact words they use. So "ranking in ChatGPT" really means being mentioned in a high share of answers to the questions your buyers ask. That is a rate, measured across many answers, and it is the only version of a ChatGPT rank that holds up. Example data for illustration. | Question | SEO answer | ChatGPT answer | | --- | --- | --- | | Where do I appear? | Position 4 for "email marketing tool" | Named in 7 of 20 answers to my prompt set | | Does it stay put? | Usually for days or weeks | Can change on the next run | | Who sees the same thing? | Most searchers in one location | Varies by account, memory and wording | | Official data | Search Console | None from OpenAI | ## What influences whether ChatGPT mentions you ### Web search grounding For questions about products, prices and recommendations, ChatGPT often searches the web before answering and then shows the pages it used as sources. When that happens, the answer is built from whatever pages the search turned up. If those pages are roundups that don’t include you, you are unlikely to be named. You can see this directly: open the sources on any answer about your category. ### The sources it cites For software questions, the cited sources tend to be "best tools" lists, review and directory sites, comparison articles, community threads and vendor pages. Each of those pages is a place where you are either present or absent. This is the most actionable thing you can learn from ChatGPT: a list of URLs that shape your category’s answers. ### How your category is described across the web The model has also learned from a large snapshot of the web. If your category’s articles, threads and reviews keep naming the same three competitors, ChatGPT has a strong default answer before it searches at all. A brand mentioned consistently across many independent sources, with the same description, is easier to recommend than one mentioned once on its own site. ### The wording of the question Broad prompts ("best CRM") favor the biggest names. Specific prompts ("best CRM for a two-person agency that lives in Gmail") give smaller products a real chance, because fewer sources match and fit matters more than size. This is good news for founders. Your buyers ask specific questions too. ### The user’s own context Memory, custom instructions and earlier messages in a conversation shape answers. You can’t control this for your buyers. When you check your own visibility, use temporary chat so your history doesn’t skew the result. ChatGPT has seen you research your own product many times. ## A practical checklist Work through these roughly in order. Each item is something a founder can do without an agency. 1. **Write down 15 buyer questions.** Category, comparison, alternatives and problem-led prompts in the words buyers use. See [choosing prompts](https://aeometer.com/guides/choosing-prompts). 2. **Check them in a temporary chat.** Record who is named and which sources are cited. This is your baseline. 3. **List the cited domains.** Sort by how often each appears. The top ten are your target list. 4. **Get onto the lists that leave you out.** Contact roundup authors with a short, factual pitch and a free account. Complete your profiles on review and directory sites that ChatGPT cites. 5. **Say what you are, plainly.** The first two sentences on your homepage should state the product, the audience and the category in words a buyer would search for. 6. **Publish comparison and alternatives pages.** Fair "you vs competitor" pages with a table and an honest "choose them if" section. These match exactly the questions ChatGPT gets asked. 7. **Keep your description consistent.** Same name, same one-liner and same category on your site, social profiles, directories and marketplaces. 8. **Let ChatGPT’s search crawler in.** Make sure robots.txt and your firewall don’t block OAI-SearchBot, which OpenAI documents as the crawler for ChatGPT search. Serve key pages as HTML that works without JavaScript. 9. **Fix wrong facts on your own pages first.** If ChatGPT gets your pricing or audience wrong, update the page that should state it clearly, then ask third-party sites to correct theirs. 10. **Re-check on a schedule.** Same prompts, same wording, same conditions. Compare rates, not single answers. > **What not to spend time on** Pages written only for AI, hidden instructions aimed at crawlers and anyone who guarantees a ChatGPT ranking. The first two can hurt you in Google, and the third is selling something that does not exist. ## How to track whether it’s working Changes take time to show, and a single answer proves nothing. Track three numbers over your fixed prompt set: - **Mention rate:** answers that name you ÷ answers you captured. This is the closest thing to "rank" that exists. - **Owned citation rate:** answers that link to your site ÷ answers you captured. - **Gaps:** answers where a competitor is named and you are not. Each one is a specific answer you can open and read. Example data for illustration. Acme Analytics, 20 prompts in ChatGPT. | Week | Answers captured | Mentioned | Mention rate | Owned citations | | --- | --- | --- | --- | --- | | Week 1 (baseline) | 20 | 4 | 20% | 1 | | Week 4 | 20 | 6 | 30% | 2 | | Week 8 | 19 (1 failed, excluded) | 8 | 42% | 4 | Note the third row: a check that failed is left out of the denominator instead of counted as a miss. Do the same in your own sheet. Small samples swing a lot, so look for a trend across several weeks before you credit or blame any one change. You can do this by hand with the method in [how to track brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt). To automate it, [AEOmeter’s ChatGPT tracking](https://aeometer.com/engines/chatgpt) sends your prompts through ChatGPT in your own Chrome, using temporary chat, and saves each answer with its sources. It needs your Chrome to be open while checks run. ## FAQ ### Can I pay to rank in ChatGPT? You can’t pay to be written into the answer itself. Be wary of any service that promises a position, since there is no fixed position to deliver. ### Does ranking on Google help with ChatGPT? Often, yes. When ChatGPT searches the web, pages that are easy to find and clearly relevant are more likely to be used. Good SEO is a solid base, though ChatGPT also cites pages that are not on the first page of Google. ### Why does ChatGPT mention my competitor and not me? Open the answer and read its sources. Usually the competitor is on pages you are missing from, or is described more clearly for that use case. That gives you a specific page to work on. ### How long until changes show up? There is no reliable timeline. Answers that rely on live web search can pick up new pages sooner than the model’s built-in knowledge changes. Measure weekly and judge over a couple of months. ### Should I check ChatGPT while logged in? Use temporary chat, which does not use or update memory. Signed-in and logged-out sessions can behave differently, so pick one setup and keep it the same every time you check. --- - Web page: https://aeometer.com/guides/how-to-rank-in-chatgpt - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # How to track brand mentions in ChatGPT > A step-by-step method to track how often ChatGPT mentions your brand and competitors: a free spreadsheet approach, a sample tracking table, and how to automate it. Last updated: 2026-10-16 You can’t search ChatGPT the way you search Twitter. There is no feed of everyone’s conversations, and OpenAI doesn’t tell brands when they are mentioned. Tracking brand mentions in ChatGPT means asking it the questions your buyers ask, the same way every time, and recording what it says. This guide gives you two ways to do that. The manual method costs nothing and takes about an hour a week for a small prompt set. The automated method does the same thing on a schedule. Both use the same logic, so you can start by hand and switch later without losing comparability. ## What you are tracking For each answer, you want four facts: - **Mentioned:** does the answer name your brand, under any of its names? - **Competitors:** which of the competitors you track are named? - **Citations:** which sources does ChatGPT attribute the answer to, and is any of them on your domain? - **Description:** how does the answer describe you? Old pricing, wrong audience and missing features all show up here. Counted over many answers, these become a [mention rate](https://aeometer.com/glossary#mention-rate), an [owned citation rate](https://aeometer.com/glossary#owned-citation-rate) and a [tracked share of voice](https://aeometer.com/guides/ai-share-of-voice). ## The manual method, step by step 1. **Pick 10 to 20 prompts.** Use questions a buyer would type before they know your name: "best product analytics tool for a seed-stage startup", "Mixpanel alternatives for a small team", "PostHog vs Amplitude for B2B SaaS". Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) explains how to pick them. Write them in a sheet and never edit the wording casually. 2. **List your brand names and competitors.** For your brand, write every alias: product name, company name, domain and common spellings. Add three to five competitors with their aliases too. "Acme", "Acme Analytics" and "acme.io" should all count as one brand. 3. **Open a temporary chat.** In ChatGPT, start a new chat and turn on temporary chat. Temporary chats don’t use or create memories and don’t appear in your history, so your past research doesn’t color the answer. Use a new temporary chat for every prompt. 4. **Paste the prompt exactly.** Copy it from your sheet. Small wording changes produce different answers, so consistency matters more than phrasing. 5. **Wait for the full answer and open the sources.** If ChatGPT searched the web, it shows source links. Open the sources list and copy each URL. 6. **Record one row per answer.** Date, prompt, mentioned yes or no, competitors named, cited URLs, whether any cited URL is on your domain, and a short note on how you were described. 7. **Mark failures separately.** If ChatGPT errored, asked you to log in or hit a limit, write "failed" and leave that row out of your rates. A failed check is not a missed mention. 8. **Repeat weekly.** Same prompts, same day of the week if you can, same setup. If you reword a prompt, treat it as a new prompt with a new baseline. ## A sample tracking table Here is what one week looks like for a fictional product, Acme Analytics, tracked against Mixpanel, PostHog and Amplitude. Example data for illustration. | Prompt | Acme mentioned | Competitors named | Cited sources | Acme domain cited | | --- | --- | --- | --- | --- | | best product analytics tool for a seed-stage startup | Yes | Mixpanel, PostHog | a "best tools" roundup, posthog.com, acme.io | Yes | | Mixpanel alternatives for a small team | Yes | PostHog, Amplitude | a review site, a Reddit thread | No | | PostHog vs Amplitude for B2B SaaS | No | PostHog, Amplitude | posthog.com, amplitude.com, a comparison blog | No | | how do I track feature adoption without a data engineer | No | PostHog, Mixpanel | a how-to article, posthog.com docs | No | | simple analytics for a two-person SaaS | Yes | PostHog | acme.io, a directory listing | Yes | | Amplitude pricing for startups | (failed: rate limit) | | | | From this week: 5 answers captured (the failed one is excluded), Acme mentioned in 3, so the **mention rate is 3 ÷ 5 = 60%**. Acme’s own domain was cited in 2 of 5, an **owned citation rate of 40%**. The two answers without Acme are [gaps](https://aeometer.com/glossary#gap): PostHog appears in both, and their cited sources show where to look next. > **Keep the answer text** Paste the full answer into a second tab or a doc, keyed by date and prompt. When the numbers move, you will want to reread what ChatGPT actually said. ## Reading the results - **Look at rates over weeks, not single answers.** One answer can change on the next run. Five weeks of the same 20 prompts tells you something. - **Group by prompt type.** You might be named in every "simple analytics" prompt and none of the comparison prompts. That points to the pages you are missing. - **Count cited domains.** Across all answers, which domains come up most? Those sources describe your category. If you are absent from the top five, start there. - **Watch the description.** A wrong fact repeated across answers usually traces back to one outdated page. ## Where the manual method breaks down The spreadsheet works well for a first baseline. It gets hard when: - You want more than one assistant. Twenty prompts across ChatGPT, Claude, Perplexity and Google AI Overviews is 80 answers per round. - You want daily data. More answers smooth out run-to-run variation. - You lose track of which wording was used when, and trends start comparing different prompts. - Copying sources by hand becomes the slowest part of the job. ## The automated method with AEOmeter AEOmeter runs the same loop for you. You set it up once: 1. **Add your brand and competitors.** Your brand, its aliases and owned domains, then the competitors you want to compare against. 2. **Add prompts.** Start from the suggested prompts, paste your own, or import a CSV or text list with a preview before saving. Tag them by type. See [prompt tracking](https://aeometer.com/features/prompt-tracking). 3. **Connect the Chrome extension.** It runs checks in background tabs of your own Chrome, signed into your own ChatGPT account, with temporary chat turned on where it can. 4. **Set a daily schedule.** Checks are queued at the time you choose and run when your Chrome is open and connected. Each answer is saved with its cited sources. Mentions are counted with exact, case-insensitive alias matching, so the counts are repeatable. Failed and blocked checks are shown and kept out of the rates. Editing a prompt starts a new baseline, so trends compare like with like. The [ChatGPT engine page](https://aeometer.com/engines/chatgpt) covers ChatGPT specifics, and [citation tracking](https://aeometer.com/features/citation-tracking) shows how sources are grouped by domain and page. One trade-off to know up front: collection happens in your browser, so your Chrome has to be open for scheduled checks to run. | | Manual spreadsheet | AEOmeter | | --- | --- | --- | | Setup | A sheet and a temporary chat | Brand, prompts and the Chrome extension | | Time per round | About a minute or two per answer | Runs in background tabs | | Assistants | Whatever you have time for | ChatGPT, Claude, Perplexity, Google AI Overviews | | Sources | Copied by hand | Saved with every answer | | Requirement | Your time | Chrome open during checks | ## FAQ ### Can I get alerts when ChatGPT mentions my brand? No tool can see other people’s ChatGPT conversations, so true mention alerts are not possible. What you can track is how often ChatGPT mentions you in answers to a prompt set you choose. ### Why use temporary chat? Temporary chat does not use or create memories, so ChatGPT doesn’t tailor the answer to everything you have asked about your own product before. That gets you closer to what a new buyer would see. ### How many prompts should I track? Ten to twenty is enough to start by hand. With a tool, 20 to 50 prompts across your main buyer questions gives a steadier picture. See [choosing prompts](https://aeometer.com/guides/choosing-prompts). ### Does a mention in ChatGPT count if it is negative? In this method a mention is a mention. Read the answer text to judge how you were described. AEOmeter also counts mentions without scoring sentiment, and keeps the original answer for you to read. ### Should I track my brand name as a prompt? A prompt like "what is Acme Analytics?" is useful for checking how you are described, but it will almost always mention you. Keep it separate from buyer prompts so it doesn’t inflate your mention rate. --- - Web page: https://aeometer.com/guides/track-brand-mentions-in-chatgpt - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AI share of voice > What AI share of voice means, the formula with a worked example, the pitfalls that make it misleading, and how AEOmeter calculates tracked share of voice. Last updated: 2026-10-16 **AI share of voice is the portion of brand mentions in AI answers that belong to you, compared with the competitors you track.** If ChatGPT, Claude, Perplexity and Google AI Overviews mention brands in your category 100 times across your prompts and 20 of those mentions are you, your share of voice is 20%. It is a useful competitive number and an easy one to misread. This guide gives the formula, works through an example, lists the traps and explains exactly how AEOmeter computes it. ## Share of voice vs mention rate These two numbers answer different questions, and you need both. | Metric | Question | Formula | | --- | --- | --- | | [Mention rate](https://aeometer.com/glossary#mention-rate) | How often am I named at all? | Answers mentioning you ÷ answers captured | | Tracked share of voice | When brands are named, how much of that is me? | Your mentions ÷ mentions of all tracked brands | Mention rate measures presence. Share of voice measures your slice of the conversation. You can grow your mention rate while losing share of voice if competitors grow faster, and the reverse can happen too. ## The formula ``` tracked share of voice = (answers that mention your brand) ÷ (sum, over every tracked brand, of answers that mention that brand) ``` Two rules keep it honest: - **A brand counts once per answer.** An answer that says "Mixpanel" five times counts as one Mixpanel mention. Otherwise a single long answer about one tool would dominate the number. - **Only answers that were captured successfully count.** Failed checks, blocked pages and queries with no AI Overview contain no mentions of anyone and are left out. ## A worked example Acme Analytics tracks three competitors: Mixpanel, PostHog and Amplitude. It runs 20 buyer prompts in ChatGPT and captures 20 answers. Example data for illustration. Percentages are rounded. | Brand | Answers that mention it | Mention rate | Share of voice | | --- | --- | --- | --- | | Acme Analytics | 6 | 6 ÷ 20 = 30% | 6 ÷ 32 = 19% | | Mixpanel | 12 | 12 ÷ 20 = 60% | 12 ÷ 32 = 38% | | PostHog | 10 | 10 ÷ 20 = 50% | 10 ÷ 32 = 31% | | Amplitude | 4 | 4 ÷ 20 = 20% | 4 ÷ 32 = 13% | | **Total** | **32 mentions** | | **100%** | The mention rates add up to more than 100% because one answer can name several brands. Share of voice always adds up to 100% across the tracked brands, because it splits the same pool of 32 mentions. Now suppose Acme gets added to two "best tools" roundups and, a month later, appears in 9 of 20 answers while the others stay the same. Total mentions become 35. Acme’s mention rate rises from 30% to 45% and its share of voice rises from 19% to 9 ÷ 35 = 26%. ## Pitfalls ### It only covers the brands you track The denominator is the brands you chose. If ChatGPT keeps naming Heap and you don’t track Heap, Heap’s mentions are missing from the pool and your share looks bigger than it is. Add a competitor and every share changes, even though no answer did. That is why AEOmeter calls it **tracked** share of voice, and why it should never be read as market share. Scan the answer text for names you don’t track. If a brand keeps appearing, add it and accept that the numbers before and after are not comparable. ### Prompt selection bias Share of voice describes your prompt set, not the world. Ten prompts that all contain "for startups" will flatter a startup-focused product. A set full of "Mixpanel alternatives" prompts will inflate the brands that commonly appear as alternatives. Choose prompts the way your buyers ask, across several types, before you look at any results. Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) explains how. ### Small samples swing hard With 20 answers, one answer changing from "Mixpanel" to "Acme" moves Acme’s share by several points. That is normal noise. Look at the trend over several weeks or many daily runs before you react. Per-prompt history helps: you can see whether the change came from one prompt flipping or many prompts drifting. ### Mixing engines and periods ChatGPT and Google AI Overviews behave differently and some queries return no AI Overview at all. Compare each engine with itself over time, and only compare periods that used the same prompts with the same wording. ### Alias mistakes If your brand is "Linear" or "Notion", a naive text search will count every use of the ordinary word. Use specific aliases and check a few answers by hand. AEOmeter matches aliases on word boundaries, case-insensitively, so "Acme" does not match "Acmesoft", but it can’t tell a common word from your brand name. Pick aliases that are unambiguous. ## How AEOmeter computes tracked share of voice - For every successfully captured answer, AEOmeter checks each tracked brand’s aliases with deterministic, case-insensitive, word-boundary matching. - Each brand counts at most once per answer. - Your tracked share of voice is your mentions divided by the mentions of all tracked brands in the same answers. - Failed, blocked and "no AI Overview" checks are excluded. When there are no answers to count, it shows a dash, not 0%. - Editing a prompt starts a new baseline for that prompt, so trends compare the same wording. - Every number links to the answers behind it, so you can read the exact text that was counted. You can see it next to mention rate and competitor gaps in [competitor tracking](https://aeometer.com/features/competitor-tracking). The full definitions live on the [methodology](https://aeometer.com/methodology) page. ## Slice it before you act on it One blended share-of-voice number hides most of what is useful. Break it down by prompt tag and by engine before deciding what to do. Example data for illustration. | Prompt type | Acme | Mixpanel | PostHog | Amplitude | | --- | --- | --- | --- | --- | | Category discovery (8 answers) | 15% | 40% | 35% | 10% | | Alternatives (6 answers) | 33% | 22% | 33% | 11% | | Comparison (6 answers) | 0% | 50% | 33% | 17% | In this example Acme holds its own on alternatives prompts and is absent from comparisons. That points to a concrete job: publish fair comparison pages and get included in the comparison articles those answers cite. The same split by engine often shows one assistant naming you far more than another, which usually traces back to which sources each one cites. Tag prompts by type when you create them so this view is always available. Our guide on [choosing prompts](https://aeometer.com/guides/choosing-prompts) suggests a tag scheme. ## How to improve AI share of voice Share of voice moves when you show up in more answers or competitors show up in fewer. You can only influence the first one. Start with [gaps](https://aeometer.com/glossary#gap): answers where a competitor is named and you are not. Open each one, read the cited sources and work out what the competitor has that you lack. Often it is a listing on a roundup, a clear comparison page or a use-case page that matches the prompt. The [answer engine optimization guide](https://aeometer.com/guides/answer-engine-optimization) covers the full set of steps. ## FAQ ### What is a good AI share of voice? There is no universal benchmark, because the number depends on which competitors and prompts you track. Compare yourself with your own past numbers on the same prompt set. ### Is AI share of voice the same as market share? No. It is your share of mentions among the brands you chose to track, in answers to the prompts you chose. Change either choice and the number changes. ### Should I count mentions or citations for share of voice? Mentions are the common definition, since a mention puts you on the shortlist. Citations are worth tracking separately as owned citation rate and by domain. ### Does position in the answer matter? Being named first may matter to readers, but answer order changes between runs and has no stable meaning. AEOmeter counts whether you are mentioned, not where. --- - Web page: https://aeometer.com/guides/ai-share-of-voice - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # How to choose prompts for AI visibility tracking > How to pick the buyer prompts you track in ChatGPT, Claude, Perplexity and AI Overviews: prompt types, how many, keeping them stable, tags, and an example set. Last updated: 2026-10-16 Your AI visibility numbers are only as good as the prompts behind them. Track ten prompts that all contain your product’s niche and you will look great. Track ten generic prompts like "best analytics tool" and you will look invisible next to companies with ten years of press. Neither tells you much about your buyers. This guide shows how to choose a prompt set that reflects how buyers actually ask, how many prompts you need, why you should keep them stable, and how to tag them. It ends with a complete example set for a fictional product. ## What makes a good tracking prompt A **prompt** is a question you send to an AI assistant, in the same words, every time you check. A good one has four properties: - **A buyer would type it.** It sounds like a person with a problem, not like a keyword. "What’s the best product analytics tool for a seed-stage B2B startup?" beats "product analytics software". - **It doesn’t name you.** Prompts that contain your brand will almost always mention you and inflate your numbers. Track those separately if at all. - **Winning it would matter.** If you appeared in this answer, could it lead to a signup? A definition question like "what is product analytics?" rarely produces a buyer. - **It stays the same.** You will run it for months. Write it once, carefully. ## The five prompt types Cover several types so one kind of question doesn’t dominate your numbers. | Type | Shape | Example | Why track it | | --- | --- | --- | --- | | Category discovery | "best X for Y" | best product analytics tool for a seed-stage startup | This is where shortlists are built. The most valuable type for most founders. | | Comparison | "X vs Y" | Mixpanel vs PostHog for a small B2B SaaS | Buyers close to a decision. Shows whether you get pulled into the comparison. | | Alternatives | "X alternatives" | Amplitude alternatives for a small team | People unhappy with an incumbent. Often the easiest place for a small product to appear. | | Problem-led | "how do I…" | how do I see which features new users actually use? | Buyers who don’t know the category name yet. Tests whether you are linked to the problem. | | Use-case | "tool for [situation]" | analytics for a two-person SaaS with no data engineer | Matches your positioning. Small products can win these against big ones. | You can also add one or two **brand prompts** ("what is Acme Analytics?", "is Acme Analytics good for B2B SaaS?") to check how assistants describe you. Tag them separately and leave them out of headline numbers. ## Where to find real buyer prompts 1. **Ask customers.** On your signup form or in onboarding calls, ask how they found you and, if it was an AI assistant, what they asked. Their words are the best prompts you will get. 2. **Sales and support notes.** The questions prospects ask before buying ("does it work without a data engineer?") translate directly into problem-led prompts. 3. **Search terms.** Your Search Console queries and keyword research show demand. Rewrite them as full questions. "crm small agency" becomes "what CRM should a five-person agency use?" 4. **Communities.** Threads where people ask for recommendations in your category show the phrasing and constraints buyers mention. 5. **Competitor pages.** The alternatives and comparison pages your competitors publish tell you which matchups they think matter. ## How many prompts to track | Stage | Prompt count | Why | | --- | --- | --- | | First manual check | 10 to 20 | Enough to see patterns. Small enough to run by hand across a few assistants. | | Ongoing tracking | 20 to 50 | Covers each prompt type with several prompts, so one answer changing doesn’t swing everything. | | Several products or segments | 20 to 50 per segment | Keep segments separate with tags so their numbers don’t blur together. | More prompts is not automatically better. Each prompt runs on each engine selected for it, so 50 prompts across four engines is 200 checks per run. Start with a set you will actually read, then add prompts where you see gaps. A rough split for a 25-prompt set: 8 category discovery, 5 comparison, 4 alternatives, 4 problem-led and 4 use-case. Adjust toward the types your buyers use most. ## Keep prompts stable AI answers are sensitive to wording. "Best analytics tool for startups" and "best analytics tools for a startup" can produce different names. If you tweak prompts every week, your trend line mixes different questions and means nothing. - Write each prompt carefully once, then freeze it. - If a prompt is wrong, retire it and add a new one. Treat the new one as a new [baseline](https://aeometer.com/glossary#baseline). - Keep a short log of what changed and when. - Add new prompts freely. Adding doesn’t break the history of existing ones, though it does change any number that averages across all prompts. > **How AEOmeter handles edits** In AEOmeter, editing a prompt’s text starts a new baseline for that prompt, so trends only compare answers to the same wording. You can also pause a prompt you no longer want checked. ## Tag your prompts Tags let you slice results without keeping separate lists. Useful tags: - **Type:** discovery, comparison, alternatives, problem, use-case, brand. - **Segment:** the audience in the prompt, such as seed-stage, agency or ecommerce. - **Competitor:** for comparison and alternatives prompts, the competitor named. - **Priority:** a "core" tag for the prompts you would report on if you could only pick ten. With tags you can answer questions like "how do we do on alternatives prompts?" or "are we visible to agencies at all?" without building a new prompt set. ## An example prompt set Acme Analytics is a fictional product analytics tool for small B2B SaaS teams. Its main competitors are Mixpanel, PostHog and Amplitude. Here is a 20-prompt starting set. Example prompt set for a fictional product. | # | Prompt | Tags | | --- | --- | --- | | 1 | best product analytics tool for a seed-stage B2B startup | discovery, seed-stage, core | | 2 | what analytics tool should a small SaaS team use to track activation? | discovery, core | | 3 | best product analytics tools for startups on a small budget | discovery, seed-stage | | 4 | which product analytics tools are easiest to set up? | discovery | | 5 | best analytics tool for a product-led SaaS | discovery, core | | 6 | top product analytics software for B2B companies | discovery | | 7 | Mixpanel vs PostHog for a small B2B SaaS | comparison, mixpanel, posthog, core | | 8 | PostHog vs Amplitude for an early-stage startup | comparison, posthog, amplitude | | 9 | Mixpanel vs Amplitude: which is better for a small team? | comparison, mixpanel, amplitude | | 10 | Mixpanel alternatives for a small team | alternatives, mixpanel, core | | 11 | Amplitude alternatives that are simpler to use | alternatives, amplitude | | 12 | PostHog alternatives without self-hosting | alternatives, posthog | | 13 | how do I see which features new users actually use? | problem, core | | 14 | how can I measure activation in my SaaS without a data engineer? | problem | | 15 | how do I find where users drop off during onboarding? | problem | | 16 | analytics for a two-person SaaS with no data team | use-case, seed-stage, core | | 17 | tool to track feature adoption in a B2B app | use-case | | 18 | product analytics for a SaaS that sells to enterprises | use-case | | 19 | what is Acme Analytics? | brand | | 20 | is Acme Analytics a good choice for B2B SaaS? | brand | Notice what this set does: it never mentions Acme outside the two brand prompts, it covers all five types, it names each competitor in at least one comparison and one alternatives prompt, and seven prompts are tagged "core" for a quick headline view. ## Common mistakes - **Only tracking head terms.** "Best CRM" is dominated by the biggest brands. Add the specific situations you actually serve. - **Writing prompts after looking at results.** Choosing prompts you already win makes the numbers meaningless. Pick the set first. - **Too many near-duplicates.** Five phrasings of the same question count one intent five times. Keep one and vary the type instead. - **Changing wording silently.** Every edit is a new baseline. Log it. - **Forgetting the problem-led prompts.** Buyers who don’t know your category’s name are often the ones a small product can win. ## Putting the set to work Run the set by hand with the method in [how to track brand mentions in ChatGPT](https://aeometer.com/guides/track-brand-mentions-in-chatgpt), or try the [free AI visibility checker](https://aeometer.com/tools/ai-visibility-checker) for a structured first pass. To run it on a schedule across ChatGPT, Claude, Perplexity and Google AI Overviews, [prompt tracking in AEOmeter](https://aeometer.com/features/prompt-tracking) lets you paste or import a list (CSV or plain text, with a preview), tag prompts, pause them and see each prompt’s history. When you add a brand, AEOmeter also suggests starter prompts you can edit before saving. ## FAQ ### Should prompts be questions or keywords? Questions, written the way a person types into a chat box. Assistants handle full sentences, and the constraints in a question ("for a two-person team") change which brands get named. ### Should I include my brand name in prompts? Only in a separate, tagged set of brand prompts used to check how you are described. Buyer prompts should not contain your name. ### How often should I change my prompt set? Rarely. Add prompts when you spot new buyer questions or competitors, and retire prompts that no longer fit. Avoid editing the wording of prompts you are trending. ### Should I use the same prompts for every AI assistant? Yes. Using identical prompts lets you compare engines. Just compare each engine’s trend with itself, since they behave differently. ### What about follow-up questions? Real buyers often ask follow-ups in the same chat. Single prompts are easier to track consistently, so start there and read a few multi-turn conversations by hand when you want more depth. --- - Web page: https://aeometer.com/guides/choosing-prompts - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # AI visibility glossary > Plain definitions of the terms used in AI visibility and answer engine optimization: AEO, GEO, mention rate, citations, share of voice, grounding, llms.txt and more. Short definitions of the terms you will meet when tracking how AI assistants talk about your brand. Where a term is also an AEOmeter metric, the definition says exactly how AEOmeter counts it. The [methodology](https://aeometer.com/methodology) page has the full detail. ## The practice **AEO (answer engine optimization)**: The work of making your brand appear, and be described accurately, in answers from AI assistants such as ChatGPT, Claude, Perplexity and Google AI Overviews. The term predates chatbots and was first used for featured snippets and voice assistants. See the [AEO guide](https://aeometer.com/guides/answer-engine-optimization). **GEO (generative engine optimization)**: Another name for the same work as AEO, popularized by a 2023 research paper of that name. In practice GEO and AEO are used interchangeably. **LLMO (large language model optimization)**: A third name for optimizing how large language models mention and describe a brand. Same idea as AEO and GEO. **AI visibility**: How often, and how, AI assistants mention or cite your brand in answers to the questions your buyers ask. Measured by sampling answers to a fixed set of prompts. See [what is AI visibility?](https://aeometer.com/guides/ai-visibility). **AI Overview**: The AI-generated summary Google shows above regular search results for some queries, with links to the pages it drew on. Many queries show no AI Overview at all; AEOmeter records those as "no AI Overview" and does not count them as misses. See [Google AI Overviews tracking](https://aeometer.com/engines/google-ai-overviews). ## Metrics **Mention**: An answer that names a brand. AEOmeter detects mentions with deterministic, case-insensitive matching of the brand’s aliases on word boundaries, and counts a brand at most once per answer. **Mention rate**: Answers that mention any alias of your brand ÷ answers successfully captured. Shown as "Visibility" in AEOmeter. When there are no captured answers, it shows a dash instead of 0%. **Citation**: A source the AI engine itself attributes to an answer, such as the source links ChatGPT, Perplexity or an AI Overview show. A citation tells you which page was used. It does not prove that page caused a mention. See [citation tracking](https://aeometer.com/features/citation-tracking). **Owned citation rate**: Answers that cite at least one URL on your own domain or its subdomains ÷ answers successfully captured. **Inline link**: A URL that appears inside the answer text rather than in the engine’s list of sources. AEOmeter keeps inline links separate from citations, because the engine has not attributed the answer to them. **Share of voice**: The portion of brand mentions in a set of answers that belong to one brand. In AI visibility it only covers the brands you choose to track, so it is not market share. See [AI share of voice](https://aeometer.com/guides/ai-share-of-voice). **Tracked share of voice**: AEOmeter’s share-of-voice metric: your brand’s mentions ÷ mentions of all tracked brands in the same answers. Adding or removing a competitor changes it. **Gap**: An evidence-backed finding such as "a competitor is mentioned in this answer and you are not". Each gap links to the answer and its sources, so you can see what to work on. ## How tracking works **Prompt**: A question sent to an AI assistant with the same wording every time, such as "best product analytics tool for a seed-stage startup". See [choosing prompts](https://aeometer.com/guides/choosing-prompts). **Baseline**: The first set of results for a prompt, which later results are compared against. In AEOmeter, editing a prompt’s wording starts a new baseline, so trends only compare answers to the same text. **Check**: One prompt sent to one engine, producing one answer, or a recorded failure such as a login wall, verification page or timeout. Failed checks are shown and never counted as misses. **Run**: A batch of checks: the enabled prompts, each sent to its selected engines. A run can be started by hand or queued by a daily schedule. **Engine**: An AI product that answers prompts, also called an answer engine. AEOmeter supports four: ChatGPT, Claude, Perplexity and Google AI Overviews. See [engines](https://aeometer.com/engines). **Grounding**: When an assistant runs a web search and bases its answer on the pages it finds, often listing them as sources. Grounded answers can change as the web changes, without any change to the model. **Temporary chat**: A ChatGPT mode that does not use or create memories and keeps the conversation out of your chat history. Useful when checking visibility, so your own history doesn’t shape the answer. AEOmeter turns on each provider’s private mode where it can. ## Technical **llms.txt**: A proposed convention from 2024: a Markdown file at a site’s root that lists its important pages for language models. Some coding agents and tools read it. No major AI assistant has documented using it to choose which brands to mention or cite. AEOmeter publishes one at [/llms.txt](https://aeometer.com/llms.txt). **MCP (Model Context Protocol)**: An open protocol that lets AI assistants connect to outside tools and data. AEOmeter runs a remote MCP server, so you can ask your AI assistant about your prompts, runs and AI visibility results. See [MCP and API](https://aeometer.com/features/mcp). --- - Web page: https://aeometer.com/glossary - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # How every AEOmeter number is calculated > Exact formulas for mention rate, owned citation rate, tracked share of voice and gaps, what counts as an answer, and what browser-collected AI answers can and cannot tell you. Last updated: 2026-10-16 AEOmeter sends the prompts you choose to ChatGPT, Claude, Perplexity and Google AI Overviews from your own Chrome, saves each finished answer with its sources, and counts which brands each answer mentions and which websites it cites. This page defines every number the dashboard shows, what goes into each denominator, and what the data can and cannot tell you. > **The short version** Rates are always "answers with X ÷ answers we actually captured". Failed, blocked and "no AI Overview" checks are left out of the denominator, never counted as misses. Zero answers shows a dash instead of 0%. Every number links back to the original answers and their sources. ## What counts as an answer One **check** is one prompt sent to one engine. A **run** is a set of checks, started by you (Run now) or by a daily schedule. A check only becomes an **answer** when the engine shows its own finished state and AEOmeter saves the answer text. Every rate on this page uses answers as its unit. How each check status is treated in metrics | Check status | What happened | In the denominator? | | --- | --- | --- | | Answered | The engine finished and the answer was saved with its sources. | **Yes** | | No AI Overview shown | Google returned results without an AI Overview for that search. | No. There was nothing to measure. | | Failed | The answer timed out (10 minutes), the tab closed, the engine returned an error reply, or a sign-in was missing. | No. A failure is not a brand miss. | | Needs attention | A sign-in wall, verification page or usage limit stopped the check. | No, unless it later finishes. | | Cancelled / Skipped (newer run) | You cancelled the check, or a newer run replaced it before it started. | No. | | Queued / Running / Waiting for answer | The check has not finished yet. | Not yet. | When a period has no answers, the dashboard shows a dash (–) instead of 0%. "No answers" and "answers that never mention you" are different facts, and the dashboard keeps them apart. ## Mention rate (shown as Visibility) Mention rate is the share of captured answers that mention your brand by name or by one of its aliases. ``` mention rate = answers mentioning any alias of your brand ÷ answers captured successfully ``` - A brand counts **at most once per answer**. An answer that names you five times counts the same as one that names you once. - The denominator is the same set of answers for every brand in the comparison, so competitor mention rates are directly comparable to yours. - Filters (period, engine, tag, prompt) narrow numerator and denominator together. ## Alias matching rules Mentions are found with deterministic text matching. There is no model deciding whether a brand "was really meant", so the same answer always produces the same result and you can check every match yourself in the highlighted answer. - Each brand matches on its **name plus every alias** you add in **Settings → Brand & competitors** (for example "Acme Analytics", "Acme", "acme.io"). - Matching is **case-insensitive**: "ACME", "acme" and "Acme" all match. - Matches need **word boundaries**. The characters on either side of the match must not be letters, numbers or underscores, in any language. "Acme" matches "Acme's pricing" and "acme.io" but not "Acmeify". - When two aliases overlap in the same spot, the longest one wins, so "Acme Analytics" is one occurrence and not two. - Matching runs on the saved answer text. It does not read sentiment, recommendation order or whether the mention was positive. Alias "Acme" against sample answer text | Answer text | Counts as a mention? | Why | | --- | --- | --- | | "Acme is a good fit for early teams" | Yes | Whole word. | | "…compared with ACME's free plan" | Yes | Case-insensitive; the apostrophe is a boundary. | | "see acme.io for details" | Yes | The dot is a boundary. | | "Acmeify automates onboarding" | No | Letters continue after the alias. | | "a pinnacle of analytics" | No | The alias is not a standalone word. | > **Common-word brand names need care** If your brand or a competitor is an everyday word ("Linear", "Notion", "Amplitude"), plain-text matching will also count ordinary uses of that word. Open a few answers and check the highlights. Adding a more specific alias does not remove matches on the name itself, so pick the name you track with that in mind. ## Owned citation rate (shown as Cited) Owned citation rate is the share of captured answers that cite at least one page on a domain you own. ``` owned citation rate = answers citing ≥ 1 URL on your domains or their subdomains ÷ answers captured successfully ``` - Your domains are the websites you list for your own brand. `www.` is ignored, and **subdomains count**: `docs.acme.io` and `blog.acme.io` count for `acme.io`. - Only **citations** count (the sources the engine attributes to its answer). A link that appears inline in the answer text does not make an answer "cited". - An answer citing three of your pages counts once. - If you have not added a website for your brand, the dashboard asks you to add one instead of showing a rate. ## Citations vs inline links AI answers contain two kinds of links, and they mean different things. AEOmeter keeps them separate. | Kind | Where it comes from | Used for | | --- | --- | --- | | **Citation** | The engine's own source pills, numbered footnotes, source cards or source list attached to the answer. Perplexity sources come from its thread data. | Owned citation rate, the Sources view, "sources citing competitors, not you". | | **Inline link** | A link written into the answer text, or a plain URL typed in the prose. | Kept with the answer for inspection. Never counted as a citation. | The Sources view groups citations by domain or by page and labels each one: **owned** (your domains), **competitor** (a tracked competitor's domains), **community** (Reddit, YouTube, LinkedIn, X, Facebook) or **other**. A source counts once per answer. A citation shows that the engine attached that page to its answer. It does not prove the page caused the answer, and AEOmeter never claims it did. ### Incomplete source lists Engines hide sources in different ways. ChatGPT can group several sources behind one pill, and Claude often answers in prose with no links at all. When AEOmeter cannot confirm it read the full source list, the answer keeps the citations it did find and is marked as having incomplete source coverage. It never invents the missing URLs. ## Tracked share of voice Tracked share of voice compares your brand with the competitors you chose to track, in the same answers. ``` tracked share of voice (brand) = answers mentioning that brand ÷ sum of answers mentioning each tracked brand ``` - Several brands can count in one answer, so the shares across all tracked brands add up to 100%. - It is a share **among the brands you track**. Adding or removing a competitor changes everyone's share. It says nothing about brands you do not track, and it is not market share. - The **Competitors** page shows each brand's mention rate side by side. Tracked share of voice is returned by the analytics API and the MCP tools. ## Gaps A gap is an evidence-backed finding that a competitor appears where you do not. Each one links to the answers behind it. AEOmeter reports two kinds: **Prompts where competitors appear and you don't**: For each prompt and engine, AEOmeter takes the **latest** answer in the period. If that answer mentions at least one tracked competitor and does not mention you, it is a miss for that engine. The prompt is listed with the competitors found and "missed on N of M engines", where M is the number of engines that returned an answer. An answer that mentions nobody is not a gap. **Sources citing competitors, not you**: Domains cited in answers that mention a competitor but not you. Your own domains are excluded and each domain counts once per answer. These are the pages AI leaned on when it recommended someone else. Gaps are findings, not instructions. They tell you where to look. Whether to publish a page, pitch a list, or ignore the prompt is your call. ## Worked example The following is **example data** for a fictional brand, Acme Analytics, tracked against Mixpanel, PostHog and Amplitude. Three prompts ran once on four engines: 12 checks. Example data: one run of 3 prompts × 4 engines | # | Prompt | Engine | Status | Acme mentioned | Competitors mentioned | Acme cited | | --- | --- | --- | --- | --- | --- | --- | | 1 | best product analytics tool for a seed-stage startup | ChatGPT | Answered | Yes | Mixpanel, PostHog, Amplitude | Yes | | 2 | best product analytics tool for a seed-stage startup | Claude | Answered | No | Mixpanel, PostHog | No | | 3 | best product analytics tool for a seed-stage startup | Perplexity | Answered | Yes | Mixpanel, PostHog | Yes | | 4 | best product analytics tool for a seed-stage startup | Google AI Overviews | No AI Overview shown | – | – | – | | 5 | product analytics that works without a data engineer | ChatGPT | Answered | Yes | PostHog | No | | 6 | product analytics that works without a data engineer | Claude | Failed (timed out) | – | – | – | | 7 | product analytics that works without a data engineer | Perplexity | Answered | No | Mixpanel, PostHog, Amplitude | No | | 8 | product analytics that works without a data engineer | Google AI Overviews | Answered | No | Mixpanel | No | | 9 | how to track feature adoption in a B2B SaaS app | ChatGPT | Answered | No | Mixpanel, Amplitude | No | | 10 | how to track feature adoption in a B2B SaaS app | Claude | Answered | No | None | No | | 11 | how to track feature adoption in a B2B SaaS app | Perplexity | Answered | Yes | PostHog | No | | 12 | how to track feature adoption in a B2B SaaS app | Google AI Overviews | Answered | No | PostHog | No | Checks 4 and 6 produced no answer, so **10 answers** form the denominator. Example data: the resulting metrics | Metric | Calculation | Result | | --- | --- | --- | | Mention rate (Acme) | Answers 1, 3, 5, 11 mention Acme: 4 ÷ 10 | **40%** | | Owned citation rate (Acme) | Answers 1 and 3 cite acme.io: 2 ÷ 10 | **20%** | | Mention rate (PostHog) | 7 ÷ 10 | 70% | | Mention rate (Mixpanel) | 6 ÷ 10 | 60% | | Mention rate (Amplitude) | 3 ÷ 10 | 30% | | Tracked share of voice (Acme) | 4 ÷ (4 + 6 + 7 + 3) = 4 ÷ 20 | **20%** | | Gap: seed-stage prompt | Missed on Claude | Missed on 1 of 3 engines | | Gap: data engineer prompt | Missed on Perplexity and Google AI Overviews | Missed on 2 of 3 engines | | Gap: feature adoption prompt | Missed on ChatGPT and Google AI Overviews (Claude named nobody) | Missed on 2 of 4 engines | If the failed check and the missing AI Overview were counted as misses, Acme's mention rate would read 4 ÷ 12 = 33%. That number would describe the tooling, not the answers, which is why AEOmeter leaves them out. ## Editing a prompt starts a new baseline Changing one word of a prompt can change the answers completely, so AEOmeter treats an edited prompt as a new measurement. By default the dashboard shows answers for the **current** wording of each prompt and the current brand setup (names, aliases, domains, competitors). Answers to older wordings stay available under "All saved definitions" and are labelled "Older version of this prompt". The API returns all definitions unless you ask for the current ones, and it reports how many answers were left out and whether definitions were mixed. ## How trends compare like with like Home compares the selected period (for example the last 30 days) with the previous period of the same length and shows the change in percentage points. Daily trend lines are stricter. Each day, AEOmeter takes the latest answer for each combination of prompt (ID and exact text), engine, recorded model, locale, collection context, collection method, extension version and private mode. Only combinations that appear on **every** observed day are plotted. Extra repetitions and inputs that do not appear every day are excluded and counted, so you can see how much was left out. Days with no answers are gaps in the line, never zeros. When nothing matches across all days, the dashboard says so instead of drawing a misleading chart. Matching recorded context reduces noise from changes in setup. It does not control personalization on the engine side and does not prove that a change is statistically significant. ## Sample size: why small samples are noisy AI answers vary from one ask to the next, even for the same prompt on the same day. With few answers, one answer moves the rate a lot. Every rate in AEOmeter is shown next to its sample ("4 of 10 answers mention Acme Analytics") so you can judge this yourself. How much one answer moves a rate, and a rough range of uncertainty | Answers in the sample | One answer moves the rate by | Rough uncertainty (worst case) | | --- | --- | --- | | 10 | 10 points | ± about 30 points | | 30 | 3.3 points | ± about 18 points | | 100 | 1 point | ± about 10 points | | 400 | 0.25 points | ± about 5 points | The last column uses the textbook approximation of ±1/√n for a proportion near 50%. Treat it as a reminder of scale, not a precise interval: repeated answers to the same prompt are not independent random samples. AEOmeter does not calculate statistical significance. - More prompts beat more repetitions of one prompt. Ten different buyer questions tell you more than one question asked ten times. - A daily schedule builds a sample over weeks. Read 30-day views before reacting to a single day. - Compare engines only when each has a reasonable number of answers. "1 of 2" on Claude is an anecdote. ## Collection context: what browser data represents AEOmeter collects from the consumer apps your buyers use, signed in with your own accounts, in your Chrome. This is a different measurement from calling a model API, and it comes with its own context. | Context | What AEOmeter does | What it means for the data | | --- | --- | --- | | Account | Uses the account you are signed in with on each engine. | Your plan and model access affect answers. A free account and a paid one can see different models. | | Private mode | Turns on each engine's private mode where it exists: ChatGPT temporary chat (unpersonalized), Claude and Perplexity incognito, Google's "without personalisation" search. Best effort, recorded with each answer. | Reduces memory and history effects. It does not make the session anonymous. Claude incognito can still apply profile preferences. | | Locale and location | Records the browser language with each answer. Google searches use English. | Location, language and device still affect results, especially Google. Locale is not a verified location. | | Model | Records the model the engine shows, when it shows one. | Engines change default models without notice. Trends match on the recorded model. | | Fresh conversation | Every check starts a new conversation with the exact prompt. No follow-ups. | Answers reflect a first question, not a multi-turn chat. | Browser data is evidence of what these engines answered for your account, at that time, in that context. It **does not** represent what every user sees, how many people asked, how often an answer was shown, or how many clicks it produced. AEOmeter does not report impressions, search volume, ranking position or sentiment. ## Known limitations - **Your Chrome must be open.** Schedules queue checks. Collection only happens while Chrome is running and connected, with each engine signed in. A day with Chrome closed has no answers, and the dashboard shows the last real capture separately from the next scheduled time. - **Engines change their pages.** The engine integrations are experimental. A layout change can break a check until AEOmeter is updated. Broken checks fail visibly instead of producing empty answers. - **Four engines.** ChatGPT, Claude, Perplexity and Google AI Overviews. Other assistants are not measured. - **Sign-ins and verification need you.** AEOmeter stops at login walls, verification pages and usage limits. It never bypasses a CAPTCHA. - **Source lists can be incomplete.** Grouped or hidden sources are flagged as incomplete coverage, not filled in. - **Text matching is literal.** Misspellings you did not add as aliases are missed, and common-word names over-match. - **No sentiment or ranking.** A mention in a list of ten and a mention as the top pick count the same. ## Questions ### Why does a failed check not count as a miss? A failed check has no answer, so there is nothing to search for your brand in. Counting it as a miss would lower your mention rate every time an engine timed out or asked you to sign in. ### Is Visibility the same as mention rate? Yes. The dashboard labels mention rate "Visibility" and owned citation rate "Cited". The formulas are identical to the ones on this page. ### Does a link to my site inside the answer count as a citation? No. Only sources the engine attributes to its answer count as citations. Inline links are saved with the answer so you can see them, but they do not change the owned citation rate. ### Can I check the numbers myself? Yes. Every rate links to the answers behind it, with brand mentions highlighted and sources listed. You can export comparisons as CSV, or pull the raw answers through the [API and MCP server](https://aeometer.com/features/mcp). --- - Web page: https://aeometer.com/methodology - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Getting started with AEOmeter > Set up AEOmeter step by step: add your brand and competitors, review starter prompts, connect the Chrome extension, run your first checks and schedule daily ones. This guide takes you from a new account to your first AI answers and a daily schedule. Most of the time goes into signing in to the AI engines in Chrome. > **What you need** Desktop Chrome, and your own accounts on the AI engines you want to check (ChatGPT, Claude, Perplexity). Google AI Overviews run as an ordinary Google search. Checks run in your browser, so Chrome has to be open for them to run. ## 1. Create your account 1. **Sign up** Go to [sign up](https://aeometer.com/sign-up) and use your email and a password, or Continue with Google. 2. **Verify your email** Email sign-ups get a verification link. Open it to activate the workspace. Google sign-ups are verified already and open the workspace right away. ## 2. Add your brand and competitors AEOmeter is organized around one brand of yours plus the competitors you want to compare with. 1. **Your website** Enter your domain, for example `acme.io`. It is used to count answers that cite your site. Subdomains such as `docs.acme.io` count automatically. 2. **Brand name** The name buyers would see in an answer, for example "Acme Analytics". 3. **What it does (optional)** One line such as "product analytics for early-stage SaaS". It helps AEOmeter suggest better starter prompts. 4. **Competitors** Add each competitor's name and website. Three to five is plenty to start. After the brand is created, open **Settings → Brand & competitors** to add **aliases**: other names that should count as a mention, such as "Acme" or a product name. Matching is case-insensitive and needs whole words. The [methodology](https://aeometer.com/methodology#alias-matching-rules) has the exact rules. Click the sparkle next to Other names to get suggested spellings. Other names also count on past answers; other changes apply to new checks. ## 3. Review your starter prompts When you add a brand, AEOmeter writes a set of starter prompts and shows them to you before anything runs. Untick the ones that do not fit and add your own. Later, open **Prompts** to make them sound like your buyers. - Keep prompts brand-neutral. "best product analytics tool for a seed-stage startup" tells you more than "is Acme Analytics good". - Edit, pause or tag prompts. Tags let you filter results later (for example "comparison" or "use case"). - Import a list as CSV or plain text. You see a preview first, and exact duplicates are skipped. - Editing a prompt starts a new baseline for it. Older answers stay available. Need ideas? The guide to [choosing prompts](https://aeometer.com/guides/choosing-prompts) covers which questions to track. ## 4. Install and connect the Chrome extension The extension is what asks the AI engines your prompts. It is currently a beta Chrome extension, and the app walks you through installing it. 1. **Add to Chrome** The last setup step, **Connect Chrome**, has an **Add to Chrome** button. Install from there; the step updates by itself when the extension connects. 2. **Connect** After install, the extension opens its Connect page. Click **Connect** and sign in with your AEOmeter account. The sidebar status changes to "Extension ready". 3. **Expect the debugging banner** While checks run, Chrome shows a banner saying AEOmeter started debugging this browser. That is how the extension types into the AI engines in background tabs. Clicking **Cancel** on the banner pauses the run. ## 5. Sign in to each AI engine in that Chrome profile Checks use your own sessions. In the same Chrome profile where the extension is installed, open and sign in to: - **ChatGPT** at chatgpt.com - **Claude** at claude.ai - **Perplexity** at perplexity.ai AEOmeter never sees your passwords, cookies or account tokens. It turns on each engine's private mode where one exists (ChatGPT temporary chat, Claude and Perplexity incognito, Google without personalisation) so your checks stay out of your chat history. ## 6. Run your first checks New brands check every prompt on ChatGPT, Claude, Perplexity and Google AI Overviews, once a day. The setup asks which engines you use and queues the first check at the end. 1. **Change engines** Open **Settings → Engines & schedule**. The page shows how many checks one run makes (prompts × engines). The same choice is used for Run now, retries and daily checks. 2. **Run now** Click **Run now** on Home. To test a single prompt, open it and use **Run all now**, or **Run** on one engine. 3. **Watch progress** Open **Activity** to see the queue and each engine's status. The extension opens background tabs grouped as "AEOmeter". Each prompt goes to all its engines at once, and the next prompt starts when they finish. The toolbar icon shows progress, finished tabs close themselves, and you can keep using Chrome. ## 7. Read your results | Where | What you see | | --- | --- | | **Home** | Anything that needs you (a signed-out engine, the extension offline), Visibility, Cited and Position with "X of Y answers", **What changed** since each prompt's previous answer, prompts where competitors appear and you don't, and sources citing competitors but not you. | | **Competitors** | Share of answers, average position, sentiment, recommendations and traits for you and each competitor. | | **Prompts** | Each prompt's latest result on every engine. Add, import, edit and turn prompts off on the same page; open one for its full history. | | **Answers** | The original answer text with brand mentions highlighted, the cited sources, the analysis (position, sentiment, traits, contradicted facts) and the collection context. | | **Sources** | Sites and pages AI cites, grouped by domain or page, labelled owned, competitor, community or other. | Failed checks and searches with no AI Overview are not counted as misses. The [methodology](https://aeometer.com/methodology) explains every number. ## 8. Adjust the daily schedule 1. **Daily checks are on** New brands run daily at 09:00. Change it in **Settings → Engines & schedule**. 2. **Pick a time and time zone** Choose a time when your computer is usually on with Chrome open, for example 09:00 in your time zone. 3. **Leave Chrome open** The schedule queues the checks. They run when Chrome is open and connected, with each engine signed in. If Chrome was closed, they wait in the queue. The dashboard shows the next scheduled time and the last real capture separately. Something not working? See [troubleshooting](https://aeometer.com/help/troubleshooting). ## Questions ### Does my computer need to be on for checks to run? Yes. AEOmeter collects answers in your own Chrome, so checks run only while Chrome is open and connected. Scheduled checks wait in the queue until then. ### Which AI engines can I check? ChatGPT, Claude, Perplexity and Google AI Overviews. ### Can I keep using Chrome while checks run? Yes. Checks run in background tabs in an "AEOmeter" tab group, and finished tabs close on their own. Chrome shows its debugging banner while they run. ### Does AEOmeter see my AI account passwords? No. You sign in to each engine yourself. The extension never captures provider cookies or account tokens. It saves the latest answer and its sources, not the page. --- - Web page: https://aeometer.com/help/getting-started - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Troubleshooting > Fix paused or failed checks: the red "!" badge, sign-in and verification pauses, the Chrome debugging banner, missing AI Overviews and sources, and offline browsers. Most problems come down to three things: Chrome is closed, an engine wants you to sign in or verify, or an engine changed its page. Find your symptom below. ## The AEOmeter icon shows a red "!" A check stopped and needs you. While things run normally, the icon shows progress such as `3/8` instead. 1. **Click the AEOmeter icon** The popup shows the engine, the prompt and the reason, for example a sign-in page or a verification check. 2. **Click Show tab** Fix it yourself in that tab: sign in, complete the verification, or wait out a usage limit. 3. **Click Resume** The check continues where it stopped. If you would rather drop it, click **Skip check**. In the dashboard, the same check shows as **Needs attention** with "Resume it from the AEOmeter icon in Chrome, or cancel." ## Sign-in walls, verification and usage limits AEOmeter stops at any sign-in wall, CAPTCHA, verification page or usage limit. It never tries to get around them and never handles your credentials. - If a check failed with **"Sign in to Claude in Chrome, then run the check again"** (or another engine), sign in to that engine in the Chrome profile where the extension is installed, then start another run. Failing that one check keeps the other engines moving. - Repeated verification pages usually mean an engine is suspicious of the session. Sign in fresh in a normal tab, use the engine once by hand, then run again. - Usage limits come from your plan on that engine. Run fewer prompts per day or wait for the limit to reset. ## Chrome says AEOmeter "started debugging this browser" This is expected. The extension uses Chrome's debugger to type into the AI engines in background tabs, and Chrome shows this banner whenever an extension does that. It appears while checks run. Clicking **Cancel** on the banner pauses the run. Nothing is lost. Open the AEOmeter popup and click **Resume** to continue. ## A check failed A failed check is never counted as a brand miss. It is left out of every rate. Common reasons: | Reason | What to do | | --- | --- | | No finished answer within 10 minutes | Run it again. If it keeps happening on one engine, the engine may have changed its page. | | The engine replied with an error instead of an answer | Open the engine yourself to see whether it is having problems, then run again later. | | Sign-in required | Sign in to that engine in the same Chrome profile, then run again. | | The tab was closed after the prompt was sent | Run it again. A check closed before sending reopens by itself. | | Chrome or the extension restarted mid-answer | Run it again. A prompt that was already sent is never sent a second time. | ### How to retry AEOmeter does not resend failed checks on its own. When an engine needs you (signed out, a verification page), Home and the extension icon say which one. Fix it in Chrome, then click **Retry** there, in the extension popup, or **Retry failed** on **Activity**. Each retry opens a fresh tab and a fresh conversation. ## "No AI Overview shown" Google ran the search and showed results without an AI Overview. This is a normal outcome. Google shows AI Overviews for some searches and not others, and it can vary by day, location and wording. Before deciding, the check waits about 20 seconds and scrolls through the results, because some overviews load lower on the page. These checks are excluded from rates, never counted as misses. If a prompt never gets an AI Overview, try wording it the way people search, for example "best crm for freelancers" rather than a long conversational question. ## Sources are missing or incomplete - **Claude** often answers in prose with no sources. An answer without links is a real result. - **ChatGPT** can group several sources behind one pill. When AEOmeter cannot read every grouped source, the answer is marked as having incomplete source coverage and keeps the sources it found. - **Inline links** inside the answer text are saved with the answer but are not counted as citations. See [citations vs inline links](https://aeometer.com/methodology#citations-vs-inline-links). - If your own site never shows as cited, check that its domain is listed for your brand in **Settings → Brand & competitors**. ## Extension status: not installed, offline, paused | Status in the dashboard | Meaning | Fix | | --- | --- | --- | | Extension not installed | No browser has connected to this organization. | Use **Add to Chrome** on Home, then click **Connect**. | | Extension offline | A connected browser has not checked in recently, usually because Chrome is closed. | Open Chrome. Queued checks start once it reconnects. | | Extension paused | Chrome is open but collection is paused. | Click the AEOmeter icon in Chrome and choose **Resume**. | | Extension ready | Chrome is open, connected and collecting. | Nothing. Checks run while Chrome stays open. | If the popup says **"N answers waiting to upload"**, answers were saved in Chrome but not sent yet, often because you were offline. Click **Retry** in the popup. ## Daily checks did not run - Scheduled checks only run while Chrome is open and connected. If your computer was asleep at the scheduled time, the checks wait in the queue and run once Chrome is back. - A check marked **Skipped (newer run)** was replaced by a newer run before it started, so you do not get duplicate answers. - The dashboard shows the next scheduled time and the last actual capture separately. If the last capture is old, Chrome has not been available. - Pick a schedule time when your computer is usually on. ## Pause, Resume and Skip prompt | Button in the popup | What it does | | --- | --- | | **Pause** | Stops new checks. An answer already generating on the engine is not cancelled. | | **Resume** | Continues the paused check, or turns collection back on. | | **Skip prompt** | Gives up on the paused prompt on every engine and turns collection back on, so the queue moves on. Skipped checks show as cancelled. | | **Open / Retry** (red box) | An engine needs you: open it to sign in or finish a verification, then **Retry** its failed checks. | | **Show tab** | Brings the running check's tab to the front so you can see or fix it. | You can also cancel queued checks from **Activity** in the dashboard, one at a time or all at once. Checks that already started are not cancelled this way. If the popup says a check from another account is unfinished, someone else used this browser with a different AEOmeter account. Click **Cancel check** to continue with yours. ## Still stuck? Email [hello@aeometer.com](mailto:hello@aeometer.com) with the engine, the prompt, roughly when it happened and what the popup said. A screenshot helps. Never send passwords or account tokens. ## Questions ### Will AEOmeter solve CAPTCHAs for me? No. It stops and waits for you. Verification pages and CAPTCHAs are never bypassed. ### Can a prompt be sent twice by accident? No. A check interrupted before sending starts over in a fresh tab. One interrupted after sending fails instead of sending again. ### Do failed checks lower my Visibility? No. Failed, blocked and "no AI Overview" checks are excluded from the denominator. See the [methodology](https://aeometer.com/methodology). --- - Web page: https://aeometer.com/help/troubleshooting - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # API and MCP developer guide > Connect to the AEOmeter MCP server and REST API: endpoints, OAuth authentication, rate limits, the error format, and the versioning and deprecation policy. AEOmeter has a remote MCP server for AI assistants and a REST API for scripts. Both use the same handlers as the web app, so anything you see in the dashboard you can read or change programmatically. ## Endpoints | Surface | URL | Description | | --- | --- | --- | | MCP server | `https://aeometer.com/mcp` | Streamable HTTP. Server card at [/.well-known/mcp/server-card.json](https://aeometer.com/.well-known/mcp/server-card.json). | | REST API | `https://aeometer.com/api/v1` | OpenAPI spec at [/openapi.json](https://aeometer.com/openapi.json), reference at [/docs](https://aeometer.com/docs). | | Agent skill | [/skill.md](https://aeometer.com/skill.md) | Instructions an agent can load before using AEOmeter for a user. | | API catalog | [/.well-known/api-catalog](https://aeometer.com/.well-known/api-catalog) | RFC 9727 index linking the specs, docs and health endpoint. | | Health | [/health](https://aeometer.com/health) | Returns `{"status":"ok"}` when the service is up. | ## Authentication The MCP server uses OAuth 2.1 with client ID metadata documents. An unauthenticated request gets a 401 whose `WWW-Authenticate` header points to the [protected resource metadata](https://aeometer.com/.well-known/oauth-protected-resource/mcp); MCP clients follow it to the [authorization server metadata](https://aeometer.com/.well-known/oauth-authorization-server), open the sign-in page, and ask the user to approve the `mcp:tools` scope. Access tokens last five minutes and refresh automatically. The REST API accepts a signed-in website session or a bearer token. Accounts are free to create at [sign up](https://aeometer.com/sign-up); the free plan needs no card. ## Rate limits The REST API and the MCP endpoint allow **120 requests per minute** per credential, or per client address when there is none. Every response carries `RateLimit-Limit`, `RateLimit-Remaining` and `RateLimit-Reset` (seconds until the window resets). Over the limit you get HTTP 429 with a `Retry-After` header; wait that many seconds before retrying. ## Errors Every 4xx and 5xx response from the API has the same JSON body. Branch on `code`, show `message` to people, and read `data` for validation details. ```json { "defined": false, "code": "NOT_FOUND", "status": 404, "message": "Not found." } ``` Common codes: `BAD_REQUEST` (400, invalid input), `UNAUTHORIZED` (401), `FORBIDDEN` (403, including plan limits), `NOT_FOUND` (404), `TOO_MANY_REQUESTS` (429) and `INTERNAL_SERVER_ERROR` (500). MCP tool failures return a tool result with `isError: true` and the same message. ## Versioning and deprecation - The API version is in the path: `/api/v1`. The MCP server reports its version in `serverInfo`. - Within a version, changes are additive only: new endpoints, new optional inputs, new response fields. Clients should ignore fields they do not recognize. - Breaking changes ship under a new path version. The previous version keeps working for at least 90 days after the new one is announced. - Deprecated endpoints respond with `Deprecation` and `Sunset` headers and a `Link` to migration notes, and every deprecation is listed in the [changelog](https://aeometer.com/changelog). ## When to use AEOmeter from an agent Use it when a user asks whether ChatGPT, Claude, Perplexity or Google AI Overviews mention their brand, how they compare with competitors, or which pages AI answers cite. Start with `list_organizations` and `list_projects`, read results with `list_observations`, and queue fresh checks with `create_run`. Checks run in the user's Chrome through the AEOmeter extension, so a new run can take minutes and needs that browser open. The [MCP feature page](https://aeometer.com/features/mcp) shows example questions. --- - Web page: https://aeometer.com/help/api - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Changelog > Every AEOmeter release in plain language, newest first: engine support, extension changes and new dashboard features, with dates. What changed in AEOmeter, newest first. Version numbers match the Chrome extension. The web app updates continuously, and changes are listed under the release they shipped with. ## 0.7.2: Run one prompt now, fewer extension permissions Released October 16, 2026. - Run a single prompt right away from its page, on every selected engine with **Run all now** or on one engine with **Run**. - The extension no longer needs permission to read and change each AI site. It reaches their pages only through Chrome's debugger, so it installs with fewer permissions. - Switch organizations from the account menu and brands from the card at the top of the sidebar. Switching keeps you on the same section. - Press Cmd+K to jump to any brand or organization. New brands and organizations open in a dialog, and organization settings have their own Brands page. ## 0.7.1: Google AI Overviews lower on the page, cancel queued checks Released October 15, 2026. - Google AI Overviews that appear below the first search result are now captured, with their sources. Before, these checks could wrongly end as "No AI Overview shown". - Queued checks are listed on the monitoring page, and you can cancel them one at a time or all at once. Checks that already started are left alone. The same action is available to AI assistants through MCP. ## 0.7.0: Gemini removed Released October 15, 2026. - Gemini is no longer supported. It never completed a reliable test run: every attempt ended in a service-error reply. Stored Gemini answers were deleted and Gemini was removed from schedules. - AEOmeter now checks ChatGPT, Claude, Perplexity and Google AI Overviews. ## 0.6.0: Parallel checks and a new way to read answers Released October 14, 2026. - Each prompt now goes to all its engines at the same time, in background tabs grouped as "AEOmeter". The next prompt starts when every check of the current one has finished. - Checks type and click like a person through Chrome's debugger, so Chrome shows its "started debugging this browser" banner while they run. Clicking **Cancel** on the banner pauses the run. - Answers are read from the page as Markdown once the engine shows it has finished. Perplexity answers and their full source list come from Perplexity's own thread data. - Google AI Overviews that appear directly in the search page are captured. - ChatGPT's private-mode menu works again. - A check interrupted before the prompt was sent starts over in a fresh tab. One interrupted after sending fails, so a prompt is never sent twice. ## Earlier versions (0.1.0 to 0.5.1) AEOmeter went from a local prototype to a working product in its first week. These versions shipped before named releases. ### October 13, 2026 (0.5.0 and 0.5.1) - A shorter path from sign-up to the first answer: new brands get starter prompts and a queued ChatGPT check, the extension opens a Connect page on install, and checks start as soon as it connects. - The toolbar icon turns red when a check needs you. A check blocked by a missing sign-in fails on its own instead of holding up the other engines. - Organization settings with members. - Fixes for checks that got stuck after a tab closed, ChatGPT source pills saved as citations, trend charts that skipped days, and engine choices that did not survive a reload. ### October 8, 2026 - Organizations with invitations. Connected browsers are shared across the organization. - A prompt-by-engine matrix and a detail page for every prompt. - A new answer viewer that opens beside the list, with Answer, Sources and Details tabs. - Light, dark and system themes. Website analytics load only after you consent. ### October 7, 2026 - The web workspace: brand-first setup with competitors, aliases and domains, and Overview, Prompts, Answers, Sources, Monitoring and Settings views. - Daily schedules with a time zone. Answers upload from the extension to your dashboard. - Prompt tags, pause and bulk actions, plus CSV and plain-text import with a preview. - Competitor comparisons with explicit denominators, gaps, and CSV export. Editing a prompt starts a new baseline. - Continue with Google sign-in, a public privacy policy, and an MCP server so AI assistants can read your results. ### October 6, 2026 (0.1.0 to 0.4.4) - The first Chrome extension: add brands and prompts, run them on the AI engines in background tabs, and save each finished answer with its sources and brand mentions. - Private modes are turned on before a prompt is typed: ChatGPT temporary chat, Claude and Perplexity incognito, Google without personalisation. - Finished tabs close automatically, and pending checks explain what they are waiting for. Want to see how checks work today? Read [how it works](https://aeometer.com/how-it-works) or the [methodology](https://aeometer.com/methodology). --- - Web page: https://aeometer.com/changelog - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # About AEOmeter > Why AEOmeter exists, who it is for, and the principles behind it: evidence first, honest denominators, your data in your browser, and an open MCP server and API. ## Why AEOmeter exists Your buyers now ask ChatGPT, Claude, Perplexity and Google questions like "best product analytics tool for a seed-stage startup" before they ever reach your site. The answer names three or four products. You are either one of them or you are not, and none of it shows up in your analytics. We wanted a simple way to see those answers: the real ones, from the consumer apps buyers use, signed in like a real person. Then count them honestly, over time, against the competitors that matter. AEOmeter is that tool. ## Who it is for Founders and indie hackers who track one to three of their own brands plus a handful of competitors. You want to know whether AI recommends you, which competitors it picks instead, and which pages it cites when it does. You do not want a dashboard built for a 40-person marketing department. ## Principles **Evidence first**: Every number links to the original answer and the sources the engine cited. If you cannot click through to the answer behind a metric, we should not show the metric. **Honest denominators**: Rates divide by answers actually captured. Failed checks, verification pauses and searches with no AI Overview are left out instead of counted as misses, and an empty period shows a dash instead of 0%. The [methodology](https://aeometer.com/methodology) spells out every formula. **Your data, your browser**: Checks run in your Chrome with your own AI accounts. The extension never captures provider cookies or account tokens, and saves only the latest answer and its sources, not the page. Website analytics load only after you consent. Read the [privacy policy](https://aeometer.com/privacy). **Built in the open**: A remote [MCP server](https://aeometer.com/features/mcp) lets Claude, Cursor and other assistants read your results, and a REST API with an OpenAPI spec at [/openapi.json](https://aeometer.com/openapi.json) uses the same handlers as the web app ([API reference](https://aeometer.com/docs)). The [changelog](https://aeometer.com/changelog) lists what shipped and when. ## What we will not claim AEOmeter measures what AI answered for your account, in your browser, at a point in time. It does not estimate impressions, search volume, sentiment or ranking position, and it does not promise you will "rank #1" in ChatGPT. The engine integrations are experimental because the engines change their pages, and your Chrome has to be open for checks to run. We would rather say that up front than surprise you later. ## Where things stand AEOmeter is in beta. It checks ChatGPT, Claude, Perplexity and Google AI Overviews through a beta Chrome extension. The app guides you through installing it after you sign up. See [how it works](https://aeometer.com/how-it-works) for the details. ## Who makes it AEOmeter is made by Andrei Canta, its founder. ## Contact Questions, bug reports, data deletion requests or a prompt that gives strange results: email [hello@aeometer.com](mailto:hello@aeometer.com). --- - Web page: https://aeometer.com/about - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Contact AEOmeter > Email hello@aeometer.com for help with checks, bug reports, billing, data deletion requests, security reports and API or MCP questions. The fastest way to reach AEOmeter is email: [hello@aeometer.com](mailto:hello@aeometer.com). It goes straight to Andrei Canta, who builds and runs AEOmeter, so there is no ticket queue or sales team in between. ## What to write about **Help with a check**: Tell us the brand, the engine and roughly when the check ran. If the extension shows a red "!" badge, say what the popup says. The [troubleshooting guide](https://aeometer.com/help/troubleshooting) covers the common pauses and failures. **Bug reports**: Describe what you expected and what happened instead, plus your Chrome version and the extension version from the popup. A screenshot of the answer page helps when an answer looks wrong. **Billing and plans**: Questions about plans, invoices or cancelling. You can also manage your subscription yourself under Organization settings, Billing. See [pricing](https://aeometer.com/pricing) for what each plan includes. **Privacy and data deletion**: Ask for a copy of your data or for your account and its data to be deleted. The [privacy policy](https://aeometer.com/privacy) explains what we store. **Security**: If you find a vulnerability, email the details privately before sharing them anywhere else, with steps to reproduce. **API, MCP and integrations**: Questions about connecting an assistant or a script. Start with the [API and MCP developer guide](https://aeometer.com/help/api). ## Postal address Bulevardul Bucureștii Noi 136, parter, ap. 5, Sector 1, București, Romania. Email is much faster; use post only for letters that have to be sent on paper. ## Before you write Most setup questions are answered in [getting started](https://aeometer.com/help/getting-started), and every metric is defined in the [methodology](https://aeometer.com/methodology). The [changelog](https://aeometer.com/changelog) lists recent fixes, so check it if something changed overnight. --- - Web page: https://aeometer.com/contact - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Privacy policy > What the AEOmeter website and Chrome extension store, what they never capture, and how to ask for your data to be deleted. Last updated: October 16, 2026 AEOmeter is a workspace for monitoring how brands appear in AI answers. This page explains the information used by the website and Chrome extension. ## Account information When you create an account, we store your name, email address, verification status, and session information. Email accounts use a password hash. Google sign-in supplies your name, email address, and profile image. Authentication records can include Google account identifiers and tokens. We request only basic identity scopes: openid, email, and profile. ## Workspace and browser data We store the projects, brands, prompts, schedules, and browser connections you configure, together with answers, citations, and run diagnostics uploaded by connected browsers. The extension stores its job queue and pending results locally. Provider cookies and account tokens from AI websites are not captured or uploaded by the extension. ## How information is used and shared We use this information to authenticate you, operate your workspace, collect requested answers, and display results. Selected prompts are submitted to the AI providers you choose. Google processes Google sign-in. Hosting infrastructure stores workspace data, and the email service processes verification and password-reset messages. Google identity data is used for account access and is not sold or used for advertising. ## Cookies and storage The website uses authentication cookies to keep you signed in. The extension uses local browser storage for its connection, queue, and pending uploads. Signing out ends your website session; you can revoke browser connections from the workspace. ## Website analytics If you allow analytics, Google Analytics uses cookies to measure website visits and page usage. We load the Google tag only after you agree. Page URLs sent to Google exclude query strings, fragments, and workspace project identifiers. We do not send account names, email addresses, prompts, or captured AI answers to Analytics. Advertising storage and personalization are disabled. Your choice is saved in this browser. Learn [how Google uses information from sites that use its services](https://policies.google.com/technologies/partner-sites). You can withdraw analytics consent below; this clears Analytics cookies and reloads the page. Change analytics cookie preferences ## Your information Account and workspace information is retained while needed to operate your workspace. You can request access, correction, or deletion by contacting [hello@aeometer.com](mailto:hello@aeometer.com). You can also remove AEOmeter’s Google authorization from your Google Account settings. [Back to AEOmeter](https://aeometer.com/) --- - Web page: https://aeometer.com/privacy - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt # Terms of service > The rules for using AEOmeter: your account, your data, your AI accounts, the beta, and the limits of what we promise. Last updated: 2026-10-16 These terms apply when you use the AEOmeter website, web app, API, MCP server or Chrome extension (together, "AEOmeter"). By creating an account or using AEOmeter, you agree to them. If you use AEOmeter for a company, you agree on its behalf. ## The service AEOmeter sends the prompts you choose to AI engines through your own Chrome browser, saves the answers and their sources, and reports on how brands appear in them. AEOmeter is in beta. Features, limits and supported engines can change, and checks can fail when an engine changes its pages. ## Your account - Give accurate account information and keep your password safe. You are responsible for what happens under your account. - You must be old enough to form a binding contract where you live. - Members you invite to an organization can see and change its brands, prompts and results. ## Your AI accounts Checks run in your browser, using the AI accounts you are signed in to. You are responsible for following the terms of ChatGPT, Claude, Perplexity, Google and any other engine you check, including their limits on automated use. AEOmeter never captures your AI account passwords, cookies or account tokens, and never bypasses verification or CAPTCHA. ## Your data You own the brands, prompts and other content you add, and the answers collected for you. You give us permission to store and process them only to run AEOmeter for you. We don't sell your data. The [privacy policy](https://aeometer.com/privacy) explains what we store and how to ask for deletion. ## Acceptable use Don't use AEOmeter to: - break the law or anyone else's rights; - send prompts meant to harm, harass or deceive people, or to get illegal content from AI engines; - attack, overload, reverse engineer or get unauthorized access to AEOmeter or the engines it checks; - resell or share access to AEOmeter without our written permission. We may suspend accounts that do these things. ## API and MCP You can use the API and MCP server to read and manage your own account's data. Keep your access credentials private. We may limit request rates to keep the service working for everyone. ## Fees The Free plan costs nothing. Paid plans and their limits are listed on the [pricing page](https://aeometer.com/pricing) and are billed in advance, monthly or yearly, through our payment provider Polar. You can cancel at any time from the Billing page; your plan stays active until the end of the period you paid for. We will tell you before we change the price of a plan you are subscribed to. ## Results are observations AI answers vary by account, location, time and model. AEOmeter records what it observed in your browser and shows how each number was calculated. It does not promise that any brand will be mentioned, cited or ranked, and it is not a guarantee of business results. Check results before making important decisions on them. ## No warranty AEOmeter is provided "as is" and "as available". To the extent the law allows, we make no warranties, express or implied, including fitness for a particular purpose, accuracy or uninterrupted availability. ## Limitation of liability To the extent the law allows, AEOmeter is not liable for indirect, incidental, special or consequential damages, or for lost profits, revenue or data. Our total liability for any claim about AEOmeter is limited to the amount you paid us in the 12 months before the claim, or 100 USD if that is more. ## Ending your use You can delete your account by emailing us. We may suspend or close accounts that break these terms, or end the service with reasonable notice. Sections about your data, liability and warranties continue after your account ends. ## Changes to these terms We may update these terms. The date at the top shows the latest version. If a change matters, we will tell you by email or in the app before it applies. Using AEOmeter after a change means you accept it. ## Contact Questions about these terms: [hello@aeometer.com](mailto:hello@aeometer.com). --- - Web page: https://aeometer.com/terms - Start tracking your brand: https://aeometer.com/sign-up - All pages for AI agents: https://aeometer.com/llms.txt