Frequently Asked Questions
Everything you need to know about AI visibility and meraGEO
What is AI visibility?
AI visibility measures how often and how prominently AI language models like ChatGPT, Claude, and Gemini mention your brand when users ask relevant questions. It's the AI equivalent of search engine rankings. As more people use AI assistants to research products, being recommended by AI models directly impacts discovery and sales.
What does GEO stand for?
GEO stands for Generative Engine Optimization, the practice of improving how generative AI assistants (ChatGPT, Claude, Gemini, and others) cite, describe, and recommend your brand when users ask questions.
How do AI assistants decide which brands to mention?
Generative engines draw on their training data and, increasingly, live retrieval. They favor brands that are clearly described, frequently referenced across reputable third-party sites (review platforms, communities, comparison content), and easy to parse via structured data and llms.txt.
How do I measure my GEO performance?
Run an AI visibility scan: query the major models with real-world queries in your category and measure how often, and how prominently, your brand is mentioned versus competitors. meraGEO does this automatically across ChatGPT, Claude, and Gemini.
How does meraGEO scan my brand?
We crawl your website to understand your product, category, and competitors. Then we generate a range of diverse, real-user-intent queries that users would ask AI models about your space, things like 'What's the best project management tool?' or 'How does X compare to Y?' We send these to ChatGPT, Claude, and Gemini, then analyze every response to determine if your brand was mentioned, in what context, and at what position.
How is the visibility score calculated?
The overall score is the average of per-model visibility scores. Each model's score combines mention rate (how often you're mentioned, weighted 80%) and position bonus (how early you appear in recommendations, weighted 20%). The score ranges from 0 (invisible) to 100 (excellent).
What is llms.txt?
llms.txt is a text file placed at your website's root (like robots.txt or sitemap.xml) that provides structured information about your product specifically for AI models. It helps AI crawlers understand what your product does, who it's for, and how it compares to alternatives. Adding an llms.txt file is one of the highest-impact actions you can take to improve AI visibility. How to write an llms.txt file, step by step.
Is the scan really free?
Yes. The visibility scan is completely free, with no signup. You get your overall score, per-model breakdown, missing mentions analysis, competitor comparison, and site audit at no cost. What you pay for, from $19/month, is monitoring: we rescan your domain every week and email you when what AI assistants say about you changes, and you get the fix report with llms.txt, FAQ, descriptions and citation targets tailored to your gaps.
What do I get on a paid plan?
Monitor ($19/month, one domain) gives you a weekly rescan across ChatGPT, Claude and Gemini, with an email when something actually changes -- and only one line when nothing did. It includes the fix report: a custom llms.txt file, an AI-optimized FAQ section, a competitor comparison page, improved meta descriptions, a prioritized list of citation targets and a score improvement plan, plus one AGENT-BRIEF.md file that hands all of it to a coding agent. Because we keep scanning, we also notice when your site changes (an llms.txt, structured data or a sitemap appearing) and show your mentions before and after. You can also add 5 questions of your own, which are asked every week alongside the standard ones, and see the searches an assistant would likely run for each. Optimize ($49/month, up to 5 domains) adds rewrites of your worst-scoring pages, keeping every fact, and 15 questions of your own per domain. Scale ($149/month, 25 domains) adds a weekly report of which competitors the assistants name and where one took your place, and 25 questions of your own per domain.
Can I choose the questions meraGEO asks?
Yes, on any paid plan. The score always comes from the same eight standard questions, so it stays comparable from week to week and with other brands. On top of those you can add your own: 5 on Monitor, 15 per domain on Optimize and 25 per domain on Scale. Each one is put to ChatGPT, Claude and Gemini as soon as you save it and again every week, and you see whether each assistant mentions you, where you rank and which competitors it names instead. You can edit or remove them at any time. For any of your questions you can also ask for the likely searches: the web searches an assistant would probably run before answering, predicted by our query fan-out model. They are predictions, not searches we watched happen, and they tell you which terms to be findable on.
What is the AGENT-BRIEF.md in my fix report?
One markdown file that turns your fix report into an ordered task list for a coding agent such as Claude Code, Cursor or Codex. Each task says where the file goes, includes the content, and ends with a check the agent can run. Anything that has to happen on another site, like claiming a review listing, is listed separately for a person. It is a plain file, so it works with any agent and needs no connection or account. Fix your site with a coding agent.
Can I use meraGEO from Claude Code, opencode or Codex?
Yes. meraGEO runs a remote MCP server, so a coding agent can read your latest visibility report, start a scan and check its status. You sign in to meraGEO once in your browser and approve access. Claude Code, opencode, Codex and optiq code have been tested, and the guide has the exact commands for each. Connect your coding agent to meraGEO.
What is your refund policy?
You can cancel any time from your dashboard, which stops the next renewal and leaves your access running until the end of the period you already paid for. Past billing periods are not refunded, because each fix report is bespoke digital content generated and delivered immediately. EU/UK customers: by subscribing you expressly request immediate delivery and acknowledge that you waive the statutory 14-day right of withdrawal (under EU Directive 2011/83/EU and equivalent UK regulations) once report generation begins. If a technical issue ever prevents delivery, email hello@merageo.com and we'll make it right.
How often should I scan my brand?
Weekly is the right cadence, which is why monitoring runs on that schedule. AI models get updated, retrieval indexes shift and competitors publish, so answers drift from week to week whether or not you changed anything. After you make changes to your own site, give AI models a few weeks to recrawl before reading much into a single move.
Which AI models do you scan?
We currently scan ChatGPT (OpenAI), Claude (Anthropic), and Gemini (Google). These are the three most widely used AI assistants that people rely on for product research and recommendations.
Why is my brand invisible to AI models?
Common reasons include: missing llms.txt file, lack of structured data (Schema.org markup), weak meta descriptions, no comparison pages against competitors, limited presence on review platforms (G2, Capterra), and insufficient content that directly answers the questions people ask AI models. Why AI doesn't mention your brand, and what to fix first.
How is this different from traditional SEO?
Traditional SEO optimizes for search engine crawlers and ranking algorithms. AI visibility optimizes for language models that generate conversational responses. The techniques overlap (structured data, good content) but AI models also rely on llms.txt files, FAQ sections, comparison pages, and presence on review platforms that they were trained on.