A recent Semrush analysis with Kevin Indig highlights something marketers need to take seriously: ChatGPT visibility changes depending on how deeply the model reasons.
The study compared minimal reasoning with high reasoning across 100 prompts and 20 buyer journeys. The result was sharp. Only 25.6% of cited domains overlapped between the two modes. That means when ChatGPT moved from quicker answers to more thoughtful, research-heavy answers, nearly three out of four sources changed.
For brands, that is not a small technical detail. It means visibility in one answer does not guarantee visibility in another.
High reasoning behaved more like a layered research process. It cited more often, used more sources per answer, and ran far more searches. Minimal reasoning produced 245 web searches across the test set. High reasoning produced 1,130. That difference shows how much more investigation can happen before the final answer appears.
The source mix also shifted. Reddit and user-generated platforms lost ground. Government, academic, official documentation, and support pages gained more visibility. In other words, when the model worked harder, it leaned more toward structured, credible, and verifiable sources.
The most interesting shift appeared during comparison prompts. At that stage, high reasoning averaged 24 sub-queries per prompt, compared with 5.5 in minimal reasoning. A buyer asking about CRM software may not get an answer from one simple search. The model may break the question into pricing, integrations, security, support, documentation, and product fit before forming its recommendation.
That changes the way marketers should think about content.
The question is no longer just, “Are we showing up in ChatGPT?” A better question is, “Are we showing up across the smaller research steps ChatGPT takes before it answers?”
Kevin Indig’s analysis also shows that early visibility can matter later in the journey. In four of the 20 journeys tested, a brand cited at the problem stage still appeared at the selection stage in high reasoning. Minimal reasoning showed no such full-journey persistence.
That is important for B2B, finance, health, SaaS, and any category where buyers do not make decisions from one simple answer.
The lesson is clear. AI visibility is becoming more fragmented and more demanding. Brands need content that can survive deeper questioning. That means stronger documentation, better comparison pages, credible third-party mentions, clear support resources, and useful content that answers real buyer concerns.
Fast answers may reward surface-level recognition. Deeper answers reward trust, clarity, and evidence.
