# 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