We asked AI the same 210 questions two days in a row. 84% of the answers changed
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
Google AI Mode, ChatGPT, Gemini and Perplexity returned a different set of brands for 84% of 210 identical queries asked one day apart, with Gemini changing answers 95% of the time, according to a two-day test tracking 5,072 brand mentions. Twenty-seven percent of brands named on day one disappeared by day two, and only 42% of cited source pages stayed the same, with the shift traced to which pages the engines happened to retrieve rather than any change in the products themselves. Top-ranked brands proved stickier, still appearing 90% of the time and holding the No. 1 spot 63% of the time, while brands ranked sixth or lower reappeared just 61% of the time; category leaders retained 73% of their placements versus 51% for smaller brands. In one example, Google AI Mode named Less Annoying CRM first for "easiest CRM for a team that has never used one" on November 17 after reading a Zapier list, then dropped it entirely the next day in favor of HubSpot after pulling a Slack guide and HubSpot's own page. The findings matter because a single check of whether an AI engine recommends a brand says little about tomorrow's answer, and they suggest visibility depends on presence across the third-party pages engines draw from. Watch for whether vendors begin treating AI retrieval sources, not just rankings, as the metric to manage.
Muhabbit's AI wrote this summary in its own words from Hacker News's report. It can get details wrong, so check the source.
Named in this story
Google AI Mode, Gemini, ChatGPT, Less Annoying CRM, HubSpot CRM