How to Know if Your Brand Appears in ChatGPT and Gemini — A Practical Guide

Most brands have no idea how they appear in AI-generated responses. They may have typed their brand name into ChatGPT once and seen a result — but that is not the same as systematic monitoring. A single manual query tells you almost nothing about actual brand visibility: it doesn't reflect how the LLM responds to the prompts your customers actually use, it doesn't compare you to competitors, and it doesn't tell you whether you are trending up or down over time. This guide walks you through a practical audit.

Step 1: Define Your Monitoring Prompts

AI visibility is prompt-dependent. Your brand may appear in ChatGPT's response to "best CRM for SaaS startups" but not "best CRM for Latin American SMBs" — and the latter may be what your actual buyers are asking. Start by listing the 5–10 questions your ideal buyers are most likely to ask an AI when looking for a solution like yours:

• Category questions: "What's the best [product category] for [use case]?"
• Comparison questions: "What are the top alternatives to [competitor]?"
• Problem-solution questions: "How do I [solve specific problem]?"
• Regional questions: "What [category] solutions are popular in [country/region]?"

These become your monitoring prompts. Run them consistently — not just once.

Step 2: Execute the Prompts and Record Results

Run each prompt manually against both ChatGPT and Gemini. For each response, record:

• Is your brand mentioned? (yes / no)
• What position? (first brand named, second, third…)
• Which competitors appear, and in what positions?
• What sources or sites does the AI cite? (these are the authority sites that shape its recommendations)
• What language does the AI use to describe your category? (this reveals what positioning it has learned)

This is tedious manually. With 5 prompts and 2 LLMs, you already have 10 responses to analyze — and you need to repeat the exercise monthly at minimum.

Step 3: Calculate a Visibility Score

A raw ranking number is hard to interpret and compare over time. A normalized score makes it actionable. One approach: convert each ranking into a score using the formula score = (1 / rank) × 100. This gives rank 1 → 100 points, rank 2 → 50 points, rank 3 → 33 points, and no mention → 0 points. Average the scores across all prompts and both LLMs to get a single Visibility Score for your brand — the number you track over time and report to clients.

Step 4: Diagnose the Gap

Once you know your current score, the next step is understanding why you are not appearing where you should be. Common causes:

1. 1
   Missing FAQ content: Your website does not explicitly answer the questions buyers are asking LLMs. Adding structured FAQ pages that mirror your monitoring prompts increases the chance of the AI synthesizing from your content.
2. 2
   Low authority site coverage: Your brand is not mentioned on the review sites, comparison platforms, or industry publications that LLMs tend to cite. A targeted PR and outreach campaign focused on these specific sites is often the highest-impact action.
3. 3
   Weak or missing llms.txt: Without a well-structured llms.txt file, AI crawlers must infer what your brand does — and may get it wrong. A clear llms.txt with your category, use cases, and key differentiators is low-effort, high-impact.
4. 4
   Competitor advantage: A competitor is appearing consistently because they have stronger signals in the training data or authority sites. Analyzing what content they have that you lack is the competitive intelligence layer.

Step 5: Monitor, Act, and Repeat

GEO optimization is not a one-time project. LLMs update their training data and retrieval indices on a rolling basis. Content published today may not be reflected in AI responses for 2–6 weeks. A systematic monitoring cadence means running your prompts every 3–7 days, detecting drops in visibility quickly, attributing changes to specific actions, and tracking the impact of new content over 30–60 days.

Doing This at Scale

Manual monitoring of even 10 prompts across 2 LLMs, for multiple client brands, becomes unmanageable quickly. Agencies need a platform that executes prompts automatically on a schedule, aggregates results across all clients, calculates Visibility Scores, identifies competitors, flags changes, and generates the specific content recommendations needed to improve each client's performance.