SEO in the AI Era: Core Updates, GEO, and Who Gets Cited

Google’s May 2026 core update finished rolling out after nearly 12 days — and the SEO community’s verdict is consistent: this one hit harder than March. At the same time, Gemini crossed 750 million monthly active users. These two events aren’t coincidental background noise. They’re the twin pressures now reshaping what search visibility actually means.

The May Core Update Wasn’t a Tremor — It Was a Recalibration

Search Engine Journal’s Matt Southern documented the volatility throughout the rollout, with practitioners describing ranking swings more pronounced than the March cycle. What’s significant isn’t the turbulence itself — Google runs multiple core updates a year — but what this pattern signals: the algorithm is actively re-weighting quality signals as AI-generated content floods the index.

The practical implication for SEO teams is uncomfortable but clarifying. Sites that leaned on topical volume strategies — publishing thin content at scale to capture long-tail queries — are the ones reporting the steepest declines. Sites with genuine depth, clear authorship signals, and demonstrable subject-matter expertise are holding or recovering faster. This isn’t a new lesson, but the May update appears to have tightened the tolerance threshold considerably. If your content strategy still relies on padding word counts or repurposing competitor outlines, the window for that approach has effectively closed.

Who Gemini Actually Cites — and Why It Should Change Your Content Architecture

Ahrefs’ Ryan Law published the most useful data point of the week: an analysis of the 50 most-cited domains in Gemini responses as of June 2026. The findings deserve more attention than they’ve received. Gemini’s citation behaviour isn’t simply a mirror of traditional PageRank. It skews heavily toward sources with structured, factual, consistently formatted content — reference sites, established publishers, and platforms with high-volume entity-rich data.

For brands trying to earn GEO (Generative Engine Optimisation) visibility, this has a specific structural implication: answer-forward content architecture matters more than persuasive narrative. AI models retrieve information that resolves queries with precision. A brand blog post that buries its core claim in paragraph four, after two paragraphs of context-setting, is functionally invisible to a retrieval system. The brands most likely to be cited are those that open with a direct, verifiable claim and substantiate it immediately — the inverse of most brand content conventions.

In Southeast Asia, this creates an interesting tension. Markets like Thailand, Indonesia, and Vietnam have strong content cultures around storytelling and relationship-building in brand communication. That works for human audiences. It actively disadvantages you with machine retrieval systems. Teams operating in multilingual environments face an additional challenge: Gemini’s citation data skews toward English-language sources, meaning localised content in Bahasa or Thai has a steeper path to AI citation visibility even when it’s high quality.

---

Ecommerce Brands Get a New Signal: The Merchant Center AI Performance Report

Semrush flagged a development that’s flown under the radar for non-ecommerce SEOs but deserves broader attention: Google is adding an AI performance report to Merchant Center. This gives ecommerce brands visibility into how their products appear across AI-powered shopping experiences — essentially a GEO dashboard for product discovery.

The strategic read here is that Google is beginning to separate the measurement infrastructure for traditional search and AI-mediated search. That’s significant. When platforms create distinct reporting surfaces, they’re signalling distinct optimisation pathways. Brands that treat AI shopping visibility as an extension of their existing PLA or Shopping campaign logic will likely misread the levers involved. Product data quality, structured attribute completeness, and review signal richness appear to be the primary inputs — not bid strategy.

For Southeast Asian ecommerce teams operating across Shopee, Lazada, and Google Shopping simultaneously, this adds a third optimisation layer that pulls in a different direction from platform-native algorithms. The brands that will navigate this cleanly are those that invest in product data infrastructure — clean feeds, complete attribute sets, verified review pipelines — rather than treating each platform as a separate campaign problem.

The Industry Is Catching Up — But the Frameworks Are Still Forming

MozCon NYC 2026’s first speaker batch, announced this week, frames its entire programme around a single question: is AI search making your brand harder to find? The fact that MozCon — historically the clearest bellwether for mainstream SEO practitioner concerns — has organised its agenda around AI citation and answer engine visibility tells you where the profession’s centre of gravity has shifted.

What’s still missing from most of the public conversation, including the MozCon framing, is a unified framework that integrates traditional SERP optimisation, AEO, and GEO into a single content decision model. Right now, most teams are treating these as parallel workstreams with separate owners. That’s operationally understandable, but strategically fragile. The brands building durable search visibility are the ones that have stopped asking “what does Google rank?” and started asking “what does a machine — human or otherwise — consider worth surfacing?”

Those two questions have more overlap than they did two years ago. But the delta between them is where the real strategic opportunity lives.

Key Takeaways

• Audit your content for answer-forward structure: if your core claim isn’t in the first 40 words, retrieval systems will likely skip it regardless of your domain authority.
• Treat Google’s new Merchant Center AI performance report as a separate optimisation signal from Shopping campaigns — the inputs are different, and early movers will have a data advantage.
• May’s core update volatility is a forcing function: depth-and-authority content strategies are no longer a best practice, they’re the baseline for sustained visibility.

The honest question the industry hasn’t fully answered yet: when AI systems increasingly mediate what audiences read, and those systems favour certain content architectures and source types, does optimising for machine retrieval make content better — or just more machine-legible? The two outcomes aren’t always the same thing.

---

At grzzly, we work with marketing and growth teams across Southeast Asia to build search strategies that hold up across both traditional SERP and AI-mediated discovery — including GEO content architecture, structured data implementation, and cross-platform visibility audits built for markets like SG, MY, TH, and ID. If the May update or Gemini’s citation behaviour has raised questions about where your brand actually stands, we’re happy to dig into it together. Let’s talk