Google AI Mode named more brands per answer in the week of November 9, 2026, rising from 0.64 to 0.72 industry brands per answer, a jump of 12.5% that Vilnux Research flags as larger than the usual week-to-week movement. In the same week, AI Mode cited 4.77 sources per answer on average across 774 measured answers. The weekly tracker, which measures the same questions on the same engines every Monday, found nothing outside the usual movement on Perplexity or ChatGPT.
Vilnux runs a fixed panel: up to nine Spanish brands, mostly services and business-to-business, asked the same buying questions in Spanish every week, on Google AI Mode through its real interface, Perplexity through its API with web search, and ChatGPT through its API without web search. A movement is flagged only when the week's jump is at least twice the usual week-to-week variation. In the week of November 9, only one metric cleared that bar: the brands named per AI Mode answer, up from 0.64 to 0.72.
Everything else stayed inside its normal range. AI Mode's 4.77 sources per answer was barely changed from the 4.70 it cited in the week of October 5, and its answer length and other measures showed no flagged movement. Perplexity showed nothing outside its usual week-to-week movement either, citing 20.25 sources per answer in its API measurement. The picture is a single, specific shift in AI Mode, not a broad change across the engines.
The report also carries a warning that matters for anyone reading Perplexity citation data from the autumn. Between October 7 and 25, 2026, the Perplexity API returned at most 10 sources per answer, which cut Vilnux's Perplexity citation metrics to about half for every brand at once. Perplexity support confirmed on October 23 that the old endpoint of its API lost support on September 27 and has since applied a default cap of 10 sources. From October 26, Vilnux requests 20 sources explicitly, so citation counts are not comparable across that date either.
That is a measurement trap with real consequences. A brand that saw its Perplexity citations halve in October may have concluded it lost visibility, when the API was simply returning fewer sources. Share of answer, a metric that was not affected, went from 12.6% to 13.6% across the same period, with overlapping confidence intervals, which is consistent with nothing real changing. Anyone comparing Perplexity citation counts from before and after October 26 needs to know the method changed, or the comparison is meaningless.
A rise in brands named per answer is good news for brands that are not yet being named, and bad news for the idea that AI Mode answers are getting narrower. At 0.72 brands per answer, most AI Mode answers still name fewer than one industry brand, so being named at all remains a distinction. The jump suggests AI Mode is surfacing a slightly wider set of brands for the same questions, which means the window for earning a mention is opening rather than closing, at least in this panel.
The deeper lesson is about how to read AI visibility data. Vilnux's panel is small and Spanish-language, and it says so plainly: it does not measure the whole market. But its discipline is the model to copy. Measure the same questions, on the same engines, every week, and flag only the moves that beat the normal wobble. One week of data proves little, but a weekly series built that way is the only honest way to tell a real change from noise, and from a broken API endpoint.
Watch whether the brands-per-answer figure holds or climbs further in the following weeks, and whether the same pattern shows up in English-language panels, since this one is Spanish. Watch the week of November 16 report for any movement tied to Gemini 4 Hydrogen, which became the AI Mode default on November 12, in the middle of this measurement window. And if you track Perplexity citations, annotate October 7 to 26 in your own data, because those three weeks are a measurement artefact, not a performance change.
Sources: Vilnux Research, week of November 9, 2026.
Related reading on this site: AI Overview Citations: What 15 Studies Agree On, and Don't, Wikipedia Leads ChatGPT Citations at 10.1% in Ahrefs\' November Ranking, Reddit Takes 29.9% of Gemini Citations in Ahrefs\' November Ranking, Perplexity's Citation Funnel: 29 Query Variants, 10 Pages Retrieved, 3 Cited, Only 2.9 Percent of AI Citations Point at the Brand\'s Own Domain, and ChatGPT Commands 74.78 Percent of AI Referral Traffic.
Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.
Start free trial