ChatGPT’s thinking mode just reshuffled which brands get seen

Only 25.6% of the brands cited by ChatGPT overlap between its minimal reasoning mode and its high-reasoning mode. That figure comes from a fresh Semrush and Kevin Indig analysis across 100 prompts and 20 buyer journeys. In short, when ChatGPT stops to “think,” it cites a completely different set of sources. For PrestaShop store owners, this is a quiet but serious shift. A product page that gets recommended in a quick-fire answer might vanish when the AI engages deeper reasoning—and vice versa. As shoppers increasingly turn to chat interfaces for product comparisons and purchase advice, your brand’s visibility now depends on which mental gear the model is in. The old playbook of “optimize once and you’re done” no longer holds. You need to understand how your catalog appears across both response modes: what triggers a mention in a fast scan, what surfaces only after a detailed analysis, and where you’re completely absent. Check your structured data, your brand mentions on third-party review sites, and the clarity of your product descriptions. A free SEO scan on Fexa AI can quickly flag which pages are invisible to conversational AI and which ones are pulling their weight, so you can patch the gaps before your competitors do.