A preliminary two-day study tested 210 buying questions across 21 categories on ChatGPT, Gemini, Google AI Overviews, and Google AI Mode, producing 839 comparable answers. The set of recommended brands changed in 84% of answers overnight, and 27% of brands named on the first day disappeared the next. Gemini was most volatile, changing 95% of its answers; AI Overviews changed least, at 67%. A first-place brand was still mentioned the next day 90% of the time, but retained the top ranking only 63% of the time. Lower-ranked brands and smaller companies were less dependable: brands ranked sixth or below reappeared 61% of the time, while smaller brands kept about half their positions, compared with 73% for category leaders.
The findings link shifting recommendations to shifting sources: just 42% of cited pages appeared on both days, while the overlap among recommended brands was 61%. In an example about easy-to-use CRMs, Google AI Mode changed its leading recommendation after drawing on different pages. The analysis argues that appearing on many relevant lists and comparison pages makes a brand more resilient than appearing on just one, and that visibility should be measured across repeated questions and over time rather than by a single snapshot. The results are preliminary, based on one sample per day on November 17 and 18, 2026; variation may reflect ordinary randomness as well as real changes. The study recommends monitoring both mentions and cited pages, then assessing trends over a week or two.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.