Does your company appear when customers ask ChatGPT questions?

Whether your company appears in ChatGPT answers depends on the question, the context, the sources retrieved and the way the brand is represented on the web. An isolated test does not answer this question. It is necessary to observe a set of real journeys and distinguish three results: the brand being mentioned, being cited as a source and being recommended as an option.

Start with customer questions

List discovery, comparison, and decision queries. Instead of just testing the company name, use questions like “which software solves this problem?”, “alternatives for a certain category” and “how to choose a supplier for this scenario?”. Include segment, location, size, and restrictions that change the answer.

Also test validation questions: price, integration, security, deadline and use cases. A brand may appear in broad queries and disappear when the buyer details the problem.

Register the type of presence

Create a spreadsheet with prompt, date, platform, response, brands cited, sources, mention position, and tone. Run each query more than once. Generative responses vary; therefore, an appearance is no guarantee and an absence does not prove permanent invisibility.

OpenAI for publishers documentation informs that public websites can appear in ChatGPT search and recommends not blocking OAI-SearchBot when the objective is to allow summaries, snippets and citations. ChatGPT search traffic can also be identified by the utm_source=chatgpt.com parameter.

Check if the website can be used

Confirm HTTP response, indexing, canonical, robots, textual content and internal links. Then, evaluate whether the page directly answers the question, identifies the company and supports claims with evidence. A technically accessible website may still be of little use as a source.

Compare with competitors

When another company appears, open the sources used. Note which page was retrieved, what claim it supported, and what external signs support the brand. This analysis is more useful than trying to guess the perfect prompt.

Initial testing reveals hypotheses; monitoring shows trend. To start with a structured reading of the website and priority query, use AI visibility diagnostics.