ChatGPT is changing the way people discover brands, products and services. The question for content teams is no longer only “How do we rank on Google?” but also “What kind of content is likely to be retrieved, used and mentioned inside an AI-generated answer?”
A recent Search Engine Land analysis by André Pitì and Ben Tannenbaum, tested 90 prompts across beauty, legaltech/regtech and IT to understand when ChatGPT triggers web searches through query fan-out.
The core finding was sharp: commercial prompts triggered fan-out 78.3% of the time, while informational prompts did so only 3.1% of the time.
The experiment: fanouts as commercial supporters
In fact, query fan-out changes the content strategy equation.
When a user asks ChatGPT a question, the system may not search the web using only that exact prompt. It can expand the request into multiple related background searches, retrieve information across those branches, and synthesize the answer from the pages it finds.
Search Engine Land describes this as the new strategic question: not just how to rank, but which pages “open the fan-out door” in the first place.
The experiment suggests that fan-out often behaves less like broad educational discovery and more like assisted decision support. In the tested sample, commercial prompts led ChatGPT toward comparisons, feature-based filtering, product lists, pricing-adjacent searches and brand-specific evaluations. Even some prompts that looked informational at first were expanded into more evaluative, solution-oriented queries.
So the implication is not “stop writing informational content”, but informational content alone may not be enough if your goal is to appear in AI answers connected to vendor discovery, product selection or option narrowing.
Understanding how to be mentioned in ChatGPT is a matter of testing
To become more retrievable, content needs to include commercial bridges: comparison sections, alternative pages, shortlist logic, feature-led explainers, evaluation FAQs and recommendation-oriented passages inside broader educational articles.
In practice, the best AI-search content will not simply define a topic. It will help the model answer the next decision-making step.
For example: “Which tool is best for this use case?”, “How does this solution compare to alternatives?”, “What features matter most?”, “Who is this product best for?”, and “What tradeoffs should a buyer consider?”
That is the real kernel of the Search Engine Land article: to be mentioned in ChatGPT, brands need to stop thinking only in terms of top-of-funnel visibility and start engineering content for the commercial paths AI systems are likely to generate in the background.
The finding is directional, not universal.
The original analysis makes clear that the sample was limited, unevenly distributed by intent, and focused on observed ChatGPT fan-outs rather than a controlled platform-level test.
But as a strategic signal, it is hard to ignore: AI visibility is increasingly shaped by how well your content supports evaluation, comparison and selection, not just education.