How to check if ChatGPT recommends your brand: a 30-minute manual audit

In today’s digital landscape, consumers are increasingly turning to AI assistants like ChatGPT, Perplexity or Gemini to quickly find answers to their queries. Rather than navigating through search results pages, they simply type in a question, such as "what's the best invoicing tool for freelancers?" and select one of the suggested options.

If your brand isn't among the names provided, you risk being overlooked entirely. The good news is that you can determine your brand's visibility within this AI-driven search landscape in just about 30 minutes with a simple spreadsheet and a few AI assistants.

To conduct this 30-minute manual audit, start by compiling a list of the questions your customers typically ask. Focus on genuine, context-rich queries rather than generic keywords. Aim for 15 to 20 questions, divided into three categories: discovery questions (e.g., "What tools help with X?"), comparison questions (e.g., "X vs Y"), and problem questions (e.g., "What pain does your product solve?").

Extract the exact wording from support inquiries, sales conversations, and community discussions, ensuring you capture the language used by customers themselves, not your marketing material.

Next, run each question through at least three AI assistants in separate, clean sessions to avoid any bias from previous interactions. For every answer, record whether your brand is mentioned, its position in the response, which competitors are named instead, the cited sources (if any), and whether the information provided is accurate.

After completing the audit for 20 real customer questions across three or more AI engines, you'll quickly notice patterns emerging. Some competitors may consistently appear in the results, indicating they are genuine competitors in AI search, while others may never mention your brand despite your clear answers to those questions, signaling potential content gaps.

Additionally, some engines may cite outdated information about your brand, often due to outdated pages, review sites, or directories that no longer accurately reflect your current offerings.

When examining the sources cited by AI assistants, it's essential to open these sources and assess their relevance. Often, you'll find that the same small set of pages consistently backs many answers, whether it's a comparison article, a Reddit thread, a review site, or a structured guide. If a cited page belongs to your brand, ensure its accuracy and upkeep to maintain its relevance.

Conversely, if a cited page doesn't represent your brand, consider why your category might be explained better by another source and whether you should be featured on that platform.

AI responses are not static; they can change after model updates, new articles, or alterations in how engines source information. Therefore, it's crucial to repeat this audit regularly. While conducting the same 20 questions across several engines each week may seem tedious, it provides a snapshot of your AI visibility rather than a final ranking.

To streamline this process, tools like CiteMe can automate the prompts and keep track of citations over time, but starting with a manual audit will yield valuable insights into your AI visibility that no existing dashboard can provide. By dedicating just 30 minutes to this manual audit, you can gain a significant understanding of your brand's presence in AI-driven searches, allowing you to address gaps, stay competitive, and continuously refine your online presence.