Googling your own name to check your reputation is a reflex. In 2026 your customers have a second reflex: they ask ChatGPT "is [your brand] any good?" or "what are the alternatives to [your brand]?" With hundreds of millions of weekly users treating ChatGPT as an answer engine, and a typical answer naming only a handful of brands, being absent from that shortlist means being invisible at the exact moment of decision.
Checking ChatGPT is trickier than a Google search. Answers are personalized, non-deterministic, and sensitive to phrasing. This guide gives you a repeatable manual audit plus the way to track your presence at scale.
Key takeaways
- ChatGPT answers are personalized and change run to run, so audit from a neutral session and across several prompts, never a single lucky query. - Test three intent layers: direct brand ("what is X"), category ("best X for Y"), and comparison ("X vs Y"), because each reveals a different visibility gap. - Look for three things in every answer: accuracy, hallucinations, and sentiment, and log which competitors get named instead of you. - A manual audit is a snapshot; Share of Model tracks your presence over time against competitors so you can see trend, not just a moment. - Fix gaps by strengthening your entity definition, earning trusted third-party citations, and correcting the sources that feed hallucinations.
Step 1: Prepare a neutral session
ChatGPT personalizes from your history and memory. If you often discuss your own company, the model is biased to mention it, which makes your audit lie to you.
Open a new chat while logged out, or turn off Memory and use Temporary Chat, and ideally an incognito window. The goal is to simulate a first-time user who has no prior context about you. Run the same prompt two or three times, since answers vary, and treat the pattern rather than any single response as your data.
Step 2: Test direct brand queries
Start with navigational intent: does ChatGPT know who you are?
- "What is [your brand]?" - "Is [your brand] legit?" - "Who is [your brand] for?"
Check the description for accuracy, watch for invented features or pricing (hallucinations), and note the tone: positive, neutral, or skeptical. If the model cannot describe you correctly, you have an entity-definition problem before you have a visibility problem.
Step 3: Test category queries
Now commercial intent, where most buying decisions happen: "best [category] for [audience]", "top [category] tools in 2026", "affordable [category] for [use case]".
Log whether you appear at all, your position in the list, and every competitor named alongside or instead of you. This is the layer that matters most for revenue, because it is where a buyer builds a shortlist. If competitors consistently make the list and you do not, that is your priority gap.
