Google's November 2026 core update started rolling out this week. At the same time, the search index logged notable ranking volatility on 8th and 9th November. And then, within days of Gemini 4 Carbon's release, Google pushed a major update to Search powered by Gemini 4 Hydrogen. Three significant changes in a fortnight.
For teams running AEO programmes, this week is a useful stress test. If your AI search visibility is built on shallow signals - a handful of optimised pages and a schema checklist - this kind of compounding volatility will expose it. If it is built on genuine topical authority and consistent entity signals across the web, it will hold.
Core Updates Now Have an AI Layer
Traditional core updates affected organic rankings. That was painful enough. But Google AI Overviews and AI Mode now draw on the same underlying index and quality signals. When a core update reweights content quality, trustworthiness, or entity authority, it does not just move your blue links. It can remove your brand from AI-generated answers entirely, or introduce a competitor who was previously buried on page three.
This is a structural change to how update risk works. Before AI surfaces existed, losing a core update meant losing traffic. Now it can mean losing recommendations. Those are different business problems. Lost traffic can be partially offset by paid. Lost recommendations in AI Mode or AI Overviews are harder to compensate for, because the user has already received an answer.
The practical implication is that your monitoring needs to cover both surfaces simultaneously. Tracking organic position changes without tracking AI citation presence during a core update gives you half the picture. If your ranking holds but your AI citation drops, your effective visibility has still declined.
Rapid Model Iteration Compounds the Problem
The speed of Gemini's iteration cycle this week is worth examining separately. Gemini 4 Carbon was released, and within a short window, Gemini 4 Hydrogen followed as a major Search update. This is not unusual in the current AI development cycle, but the pace creates a specific challenge for AEO programmes: the model that decides what gets cited can change faster than a content calendar does.
Different model versions weight signals differently. One iteration may favour structured, definition-led content. The next may lean more heavily on entity authority signals from third-party sources - reviews, citations, press mentions. If your programme is optimised for a fixed model snapshot, you are always a release behind.
This is why the most resilient AEO strategies are not built around any single model's quirks. They are built around the signals that have remained stable across iterations: clear entity definitions, consistent factual claims across the web, structured content that answers specific questions, and citation presence on authoritative third-party domains. Those signals have survived multiple model updates and multiple core updates. They are the foundation worth investing in.
What to Check When Volatility Hits
When you see a core update landing alongside search ranking volatility, the diagnostic sequence for AI search is different from traditional SEO. Start by running a set of tracked queries through Google's AI Mode and checking whether your brand is cited, and in what context. Do the same in Perplexity and ChatGPT for queries where you have previously seen referral traffic. Changes across multiple platforms simultaneously suggest a content quality signal has shifted. Changes in one platform only suggest a model-specific weighting change.
Then look at your GA4 referral data. AI-referred sessions from Google AI Mode, Perplexity, and ChatGPT should be visible as separate referral sources if your GA4 implementation captures them correctly. A drop in AI referral traffic during core update volatility is a signal that citations have shifted, not that user behaviour has changed. That distinction matters for how you respond.
If you are tracking AI-referred leads through to CRM and can attribute revenue, you will also want to check whether lead quality from AI-referred sessions has changed. Sometimes citation position changes bring different-intent users. A drop in AI referral volume that is accompanied by an increase in AI referral conversion rate can indicate you are now being cited for higher-intent queries - which is a programme improvement, not a loss.
The Case for Diversifying Across AI Search Surfaces
One of the quieter lessons from a week like this is that dependence on any single AI search surface is a concentration risk. Google's AI Mode and AI Overviews are significant, but ChatGPT is now a material traffic and referral source for a growing number of B2B and considered-purchase categories. Perplexity has a smaller but high-intent user base. Each platform has its own citation logic and its own update cycle.
A programme that monitors and optimises for citation presence across multiple AI surfaces is naturally more resilient. When Google's core update causes volatility in AI Overviews, Perplexity citations may remain stable. That is not a reason to deprioritise Google - it is a reason to build a broader programme rather than a narrow one.
The content requirements are largely consistent across platforms. Well-structured answers to specific questions, clear entity signals, factual accuracy, and third-party citation presence all improve visibility across Google AI surfaces, ChatGPT, and Perplexity simultaneously. You are not building separate content strategies for each. You are building one strong foundation that serves all of them.
What Resilient AEO Programmes Do Differently
Programmes that hold up through core updates and model changes share some consistent characteristics. They treat digital PR as an ongoing citation-building activity, not a campaign. They maintain consistent entity information - company description, product definitions, key claims - across their own site, their Google Business Profile, Wikipedia where applicable, and third-party review platforms. They do not rely on a single piece of content to carry their AI search visibility.
They also measure at the right level. Tracking individual keyword rankings in AI search is not useful because AI surfaces do not work that way. What matters is citation frequency across a set of tracked queries, referral traffic trends from AI platforms, and - critically - whether AI-referred traffic converts and at what value. That last metric is the one that justifies the programme budget and survives scrutiny in a board conversation.
A period of heavy update activity is uncomfortable. It is also a reasonable forcing function to review whether your programme has the right foundations or whether it has been papering over gaps with surface-level optimisations. The brands that come out of November's volatility with stable or improved AI citation presence will be the ones that built on substance rather than signals they could game.