Google’s May Core Update Lands, Search Goes Agentic on Gemini 3.5, and AI Mode Gets Conversational Ads

Estimated reading time: 26 minutes

The biggest week of 2026 for search marketing news. Google launched the May 2026 core update on 21 May (the second core update this year), redesigned its search box at I/O and made Gemini 3.5 Flash the default AI Mode model, and used Marketing Live to unveil two new AI Mode ad formats. Add 1 billion monthly active users on AI Mode, SynthID watermark verification rolling into Search, iPullRank’s finding that Gmail signals shift AI Mode brand visibility by 46 percentage points, and the launch of WordPress 7.0 with native AI, and there’s plenty of search marketing news to work through this week.

Table of Contents

SEO & Algorithm Updates

Google May 2026 Core Update Rolling Out

Google Search Console Links Report Broken

Google's llms.txt Guidance Contradicts Itself Across Products

Google I/O 2026 and the AI Search Surface

AI Mode Hits 1 Billion Monthly Users After One Year

Google's New Intelligent Search Box and Gemini 3.5 Flash Default

Google Brings SynthID Content Verification to Search

Universal Cart Lands at I/O as UCP Goes Cross-Surface

YouTube Brings Conversational Search and Gemini Omni to Creators

AI Search Research & GEO

iPullRank: Gmail Signals Boost AI Mode Brand Visibility +46pp

Microsoft Clarity Adds Grounding Queries Behind AI Citations

Victorious: 90% of Brands Have Zero AI Search Mentions

LLM Guidance Doesn't Transfer the Way SEO Guidance Did

Reddit's AI Search Influence Goes Beyond Training Data

PPC & Paid Media

AI Mode Gets Conversational Discovery Ads and Highlighted Answers

Asset Studio, Ask Advisor, and Meridian: GML 2026's AI Bundle

Google Ads Budget Misallocation Is Bigger Than You Think

Industry & Tooling

WordPress 7.0 Launches With Native AI Integration

Wrap-Up

Strategic Direction: Where Search Marketing Is Heading

Key Takeaways

Frequently Asked Questions

Conclusion

Google May 2026 Core Update Rolling Out

Google launched the May 2026 core update on 21 May. It’s the second core update of the year (March being the first), and Google said the rollout may take up to two weeks to complete. The company stuck to the standard line about helpful, people-first content and “no specific actions” for sites that lose ranking.

Early third-party tracking showed elevated volatility starting late on 20 May and intensifying through 21 and 22 May. The unusual element this time is the timing. The update lands the day after Google I/O 2026 announced a redesigned search box, Gemini 3.5 Flash as the default AI Mode model, and new Search Agents launching this summer. Whether the core update is connected to those AI-layer changes or sits separately on the classic ranking system is the question every SEO team will be asking through the rollout window.

Practical playbook for the next two weeks. Document rankings on top keywords now for comparison once the dust settles. Hold off on panic changes for at least seven to ten days; ranking shifts during core updates often partially reverse before stabilising. If a client is clearly down at day ten, the recovery thinking should focus on whichever quality signals March’s update rewarded (depth, originality, author attribution, internal link coherence), since the criteria don’t usually pivot wildly between consecutive updates.

For agency-side reporting, this is also a useful moment to set the right expectations with clients. A core update that lands in the same week as Google’s biggest AI announcements of the year will be hard to attribute cleanly to either source. We’re recommending clients hold off on definitive interpretation until at least mid-June.

Google Search Console Links Report Broken

Google Search Console’s Links report has been broken since 20 May. Multiple users are reporting empty data, partial loads, or stale figures across both internal and external link reports. Google hasn’t issued a public confirmation, but the pattern is consistent enough that this isn’t a per-account issue.

If you rely on GSC Links data for client reporting due this week, document the gap and pull comparison data from Ahrefs, Semrush or Majestic to cover the window. Expect Google to acknowledge and fix in the next several days, but don’t count on backfilled data for the affected period.

Google's llms.txt Guidance Contradicts Itself Across Products

A small but instructive contradiction in Google’s own product family this week. Google Search Central updated its position to say llms.txt isn’t a meaningful signal for Search. Lighthouse, meanwhile, started running experimental audits checking for the presence of llms.txt as a signal of AI agent readiness. Same company, opposite product guidance, same week.

The substantive question is what publishers are supposed to do with this. The pragmatic read is that llms.txt sits at the level of robots.txt-style hygiene. It’s not a ranking factor in itself, but a useful machine-readable signal of which content you want AI agents to access on your behalf. Pages that don’t ship with an llms.txt aren’t penalised. Pages that do can shape how agentic systems navigate the site for downstream AI tasks.

For most client sites, the lift is small (one file at the root, a structured table of preferred-content URLs) and the future-proofing argument is strong. We’re adding it to the next technical SEO audit checklist if not already there, not as a ranking lever, but as low-cost operational hygiene that’s likely to matter more as agentic flows mature.

AI Mode Hits 1 Billion Monthly Users After One Year

1 billion monthly active users on Google AI Mode after one year, with queries more than doubling every quarter since launch. That’s the headline number from Google’s first-year usage data, published at I/O on 19 May. The same scale milestone Bing reached after a decade in our 30 April roundup; AI Mode hit it in 12 months.

A few of the data points to examine:

• Average AI Mode queries are three times longer than traditional Google searches.
• Follow-up queries are up over 40% month on month in the US.
• More than one in six AI Mode searches use multimodal inputs (voice, image, video).
• Image-based search queries grew more than 40% month on month.
• The five most common AI Mode query starters: “what,” “how,” “I,” “is,” “can.”

Google grouped query intent into five categories: Explore, Decide, Learn, Create, and Do. Planning queries grew 80% faster than overall AI Mode growth over six months; brainstorming queries grew 30% faster. Both are upper-funnel behaviours, which matters for any client whose acquisition strategy depends on appearing in the brand-consideration phase rather than the transactional query.

For agency-side strategy, the implication is that AI Mode is a real channel with real volume, not a marketing-deck thought experiment. Brand visibility planning that ignored AI Mode through 2025 needs to catch up fast. The longer-tail conversational queries are a fundamentally different optimisation surface from short-keyword search, and Google has just confirmed the scale at which that new surface is operating.

Google's New Intelligent Search Box and Gemini 3.5 Flash Default

Google’s VP of Search called this the biggest upgrade to the search box in over 25 years. The redesigned Intelligent Search Box, unveiled at I/O on 20 May, is bigger than that claim makes it sound. Three changes ship together.

First, the box itself becomes dynamic. It expands as the user types a longer query, accepts images, files, video, and Chrome tabs as input alongside text, and surfaces AI-powered query suggestions that replace classical autocomplete.

Second, Gemini 3.5 Flash became the default AI Mode model globally on 19 May. This is a faster, more efficient model than the prior Gemini 3 Pro default, and (combined with the new search box) shifts the average AI Mode query toward lower latency.

Third, Search Agents launch this summer in two flavours:

• Information agents (for AI Pro and Ultra subscribers) monitor the web across blogs, news and social media on the user’s behalf and report back when something matches.
• Agentic booking (for all US users) places phone calls to businesses for the user, starting with home repair and pet care categories.

The generative-UI capability is the sleeper feature. Google demoed Search creating custom dashboards, visual tools, and mini-apps for specific tasks via the Antigravity platform. The free tier ships this summer; premium Antigravity mini-apps follow in subsequent months.

What this changes for SEO and PPC. Long-tail conversational query optimisation becomes a primary discipline rather than a secondary one. The Agents layer means brands now need to think about whether they’re a callable business (for agentic booking), a monitorable source (for information agents), and a generative-UI candidate (for any task-specific mini-app that might recommend their service). Three new surfaces to optimise for, three new measurement frameworks needed. We’ll be running internal experiments on agentic booking discoverability for relevant client verticals over the next two months.

Google Brings SynthID Content Verification to Search

Google is rolling SynthID watermark verification into Search this week. The feature lets users query an image with “Is this made with AI?” and get a verification result based on whether the content carries SynthID’s imperceptible watermark. It’s available in Lens, AI Mode, and Circle to Search now, with Chrome support arriving in the next few weeks and C2PA Content Credentials verification following over the coming months.

The catch is real. SynthID only detects content watermarked by SynthID itself (mostly Google’s own image generation models so far). AI content from OpenAI, Midjourney, Stable Diffusion or open-source models that doesn’t carry the watermark won’t be flagged. Coverage will widen as more providers adopt the watermarking standard, but for now SynthID-in-Search is a partial solution.

For publishers and content creators, the immediate question is how AI-generated content carrying SynthID watermarks will affect discoverability in search results. Google hasn’t said it will deprioritise watermarked AI content directly, but it’s reasonable to expect transparency labels in search experiences in the coming year. If you’re using Google’s image generation tools for marketing imagery, expect those images to start surfacing with AI labels in some search contexts.

Universal Cart Lands at I/O as UCP Goes Cross-Surface

Google introduced Universal Cart at I/O on 19 May. It’s a cross-surface shopping system that follows the user across Search, Gemini, YouTube, and Gmail, monitoring prices, inventory and deals, and surfacing alternatives intelligently. The launch retailers include Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, and all Shopify merchants. UCP-powered checkout (the underlying Universal Commerce Protocol) is expanding into Canada and Australia this summer, with the UK to follow.

The agentic features are the meaningful part. Universal Cart can flag part compatibility on complex purchases, surface loyalty perks the user might have forgotten about, recommend savings opportunities, and integrate with Google Wallet for payments. The pitch is that a buying journey that previously required visiting four retailer sites and comparing in spreadsheets can now resolve inside Google’s surfaces with shipping credentials already populated.

For retail clients, the operational question is straightforward. If you’re integrated with UCP, this expands your reachable transaction surface materially. If you’re not, the competitive gap will widen each quarter through 2026 as the cart fills out across markets. Shopify integration removes the technical barrier for most mid-market retailers; the strategic question is whether you want Google to keep more of the buying journey on its surfaces rather than your own.

YouTube Brings Conversational Search and Gemini Omni to Creators

YouTube announced two creator-facing additions at I/O. Conversational Search lets viewers ask natural-language questions about video content and get AI-generated answers grounded in the video transcript. Gemini Omni provides creators with multimodal AI remixing tools (turning a long-form video into shorts, generating alternative thumbnails, repurposing voiceovers into other languages).

For YouTube SEO and content teams, both updates change the optimisation calculus. Transcripts now matter more than ever for discoverability, since conversational search uses them as the answer-ground. Thumbnail experimentation becomes faster and cheaper with Gemini Omni doing the iteration. The shorts-from-long-form workflow is the productivity multiplier most creator-led brands have been waiting for.

We expect the conversational search behaviour to start cannibalising some YouTube watch time (because answers may arrive without the user needing to watch the video), but that trade is probably net positive for brand visibility on YouTube searches that previously delivered no impression at all.

iPullRank: Gmail Signals Boost AI Mode Brand Visibility +46pp

A 46 percentage point jump in brand visibility when the user’s Google account has Personal Intelligence connected. That’s the headline finding from a new iPullRank study testing whether Gmail and Google Photos signals influence AI Mode brand recommendations. The sample: 1,922 AI Mode responses across three account conditions (blank control, Personal Intelligence-connected blank account, mature account) and eight product categories.

The mechanism. Researchers seeded brands into the test accounts via Gmail messages and Google Photos albums, then queried AI Mode in product-purchase scenarios across categories including running shoes, hoodies, coffee machines, banks, smartphones, and SEO agencies. Seeded brand appearance went from 23.9% in the control to 66.8% in the Personal Intelligence account. Top-3 placement rose from 4.5% to 24.9%.

“Brand signals added to personal context in a Google Personal Intelligence connected account changed AI Mode recommendations: those brands were +46 percentage points more likely to appear versus control.”

Gmail was by far the stronger signal: 53.6% appearance rate when seeded via email versus 10.5% via photos. Coffee machines were an exception, where photo seeding of recognisable brands like Jura worked better, likely because visual brand recognition was the determining cue. The study also tested fictional brands (Velstride, Northpeak SEO, HarborTrust) and found that even invented brands could be seeded into the AI Mode recommendation set through email signals.

For agencies, this is meaningful. Owned audience touchpoints (newsletters, transactional email, receipts) now influence AI-mediated discovery for any user who opts into Google’s Personal Intelligence features. Brand investment in those channels has a measurable second-order return on AI visibility for opt-in users. The category effects also matter: preference-led verticals (apparel, beverages, electronics) respond more strongly than trust-heavy verticals (financial services, B2B), where the public web layer still dominates.

Microsoft Clarity Adds Grounding Queries Behind AI Citations

Microsoft Clarity now exposes the grounding queries that Microsoft’s AI surfaces (Copilot, Bing generative search) use to retrieve content for citations. Released as generally available this week, the feature shows the search strings the AI engine generates internally before composing an answer, alongside which of your pages were retrieved.

This is a useful new data layer for AI visibility teams. Until now, the prompt-to-citation chain has been a black box for non-Bing engines. Clarity opens it at least for the Microsoft ecosystem.

AI engines aren’t searching for the user’s query. They’re searching for the queries they’ve decomposed the user’s intent into.

Practical use. Pull the grounding queries Clarity reports for your top-cited pages, look for query patterns that don’t match your existing keyword strategy, and use them to refine content for AI retrieval. A page that ranks for “best email marketing software” via Google may be retrieved by an AI grounding query like “email platforms with native segmentation for ecommerce.” The distinction matters for content depth, internal linking, and on-page structure.

Limitations: this only covers Bing and Copilot. ChatGPT, Gemini, and Perplexity remain opaque on grounding queries. Treat Clarity’s data as directional intelligence rather than universal truth.

Victorious: 90% of Brands Have Zero AI Search Mentions

89.8% of brands tested have zero AI mentions across the eight major AI platforms. That’s the headline from Victorious’s study of 177 brands across healthcare, SaaS, financial services, ecommerce/retail and legal services, drawing on 107,011 AI responses cross-referenced with Semrush organic data from Q1 2026.

The vertical patterns are the interesting layer. Healthcare, SaaS and financial services brands tend to get both mentioned and cited (structured data and third-party validation feed the citation layer). Ecommerce brands get recognised by name but rarely linked back to (the citation drops). Legal services have the opposite problem: their content gets cited but the firm doesn’t get the attribution.

What this means practically. The white space is real. With nearly 90% of brands invisible in AI search results, first-mover advantage in any client vertical is meaningful for the next 12 to 18 months. Optimisation effort that gets a brand into the recognised-and-cited tier is going to look outsized compared to similar effort in classical SEO, where the competitive density is much higher.

Victorious also flagged the Personal Intelligence dynamic Google’s I/O announcements amplified. Once Personal Intelligence-style features become standard, the gap between AI-visible and AI-invisible brands compounds, because the visible ones are also the ones users encounter and (potentially) get seeded back into AI recommendations through their own accounts.

LLM Guidance Doesn't Transfer the Way SEO Guidance Did

Duane Forrester (a regular feature in these roundups now) made an argument worth bookmarking. LLM guidance doesn’t port between providers the way SEO guidance did between search engines. The historical comparison is the load-bearing point. Sitemaps, Schema.org, and robots.txt were collaborative standards across major search engines, which is why an SEO trick that worked for Google generally helped on Bing too. That collaborative layer doesn’t exist for LLMs.

Concrete examples Forrester names:

• Training data: OpenAI licenses News Corp and Reddit; Google has a separate Reddit deal; Anthropic’s licensing is undisclosed.
• Crawlers: OpenAI runs three distinct bots (GPTBot, OAI-SearchBot, ChatGPT-User); Anthropic and Perplexity have their own equivalents; Google uses different crawlers again.
• Retrieval: ChatGPT uses Bing’s index; Perplexity uses a Vespa-based search; Claude partners with Brave; Gemini uses Google.
• Alignment: OpenAI uses RLHF; Anthropic uses Constitutional AI. Same source content can produce demonstrably different output behaviours.

Only 11% of cited domains appear across multiple AI platforms, per Forrester’s framing. The practical takeaway is that “optimise for Google and trust portability” is no longer a defensible AI visibility strategy. Each provider needs to be tested separately, and divergence between providers should be the default expectation rather than the exception.

Reddit's AI Search Influence Goes Beyond Training Data

A piece from Search Engine Land this week argues Reddit’s outsized influence in AI search outputs isn’t primarily about its position in training data. It’s about what Reddit contains: authentic, contextual, often-disagreeing human conversation that polished brand content can’t replicate.

The framing is worth taking seriously. AI engines weight Reddit citations heavily because Reddit threads provide nuance, dissent, lived-experience caveats, and informal trust signals that don’t appear in conventional marketing content. A well-written corporate blog and a Reddit thread on the same topic don’t compete on the same dimensions. The thread offers something the blog structurally can’t.

For brands, the implication is uncomfortable. Investing in Reddit visibility is harder to control than investing in owned media. You can encourage employees to participate authentically in relevant subreddits, work with niche moderators, and seed expert-led answers, but you can’t write the threads yourself without risking the rule-breaking detection that costs brands their accounts. The teams that have figured this out (a small group of B2B SaaS vendors and consumer brands) are getting meaningful returns; the teams trying to repurpose blog content on Reddit aren’t.

AI Mode Gets Conversational Discovery Ads and Highlighted Answers

Google unveiled two new AI Mode ad formats at Marketing Live 2026 on 20 May. Both are powered by Gemini and represent the first dedicated ad surface inside AI Mode.

Conversational Discovery ads respond to detailed, exploratory prompts. Gemini generates tailored creative based on the conversational context, then surfaces product features relevant to the query. The example Google used: a user asking how to make their home smell like a fancy spa or a rainy forest, with low-maintenance fragrance products getting surfaced as ad recommendations alongside the AI response.

The second format places ads directly inside AI-generated recommendation lists. When AI Mode is asked to suggest language learning apps before a trip, advertiser content can appear inside the suggestion set rather than alongside or below it. Sponsored labels are visible on both formats for transparency.

Launch timing is ambiguous. Google framed both formats as expected to be tested inside AI Mode without a confirmed surfacing date. No eligibility or pricing details were shared.

For PPC teams, the strategic question is what these new formats do to the AI Mode citation economy. Conversational Discovery ads run alongside AI answers; the embedded format sits inside them. That second placement is potentially much more valuable (better attention, less ad-blindness, higher click intent) but raises uncomfortable transparency questions about how clearly users will distinguish AI-recommended brands from advertised ones. Expect compliance scrutiny on the disclosure layer through 2026.

Asset Studio, Ask Advisor, and Meridian: GML 2026's AI Bundle

Bundle three Marketing Live announcements together and a clearer shift in Google’s product strategy emerges. Asset Studio gained multimodal capabilities for ad creative generation (text, image, video synthesis from prompts inside Google Ads). Ask Advisor launched as a cross-product AI agent that ties together Ads Advisor and Analytics Advisor (so an account question that crosses both surfaces no longer requires switching tools). And Meridian, Google’s open-source marketing mix modelling tool, became available inside Google Analytics 360, with a new “Future Long-Term Conversions” prediction feature.

Each one of these is significant on its own, but together they signal a shift in Google’s product strategy for advertisers. The pitch is no longer “use our ad platform plus your own analytics plus your own measurement infrastructure.” It’s “use our ad platform plus our analytics plus our measurement modelling plus our AI agent that ties them all together.” The competitive question for agencies is what value remains in being the layer above Google’s stack when Google’s stack increasingly includes the layer.

The practical move for the next two quarters is to test Ask Advisor against your standing client report templates, validate Meridian’s MMM outputs against any third-party MMM you currently run, and decide where Asset Studio’s creative generation fits in client workflows. We expect to recommend Meridian for clients without existing MMM commitments, and Asset Studio for high-volume creative production accounts.

Google Ads Budget Misallocation Is Bigger Than You Think

How much Performance Max budget is quietly burning on the wrong channels? A useful diagnostic piece from Search Engine Journal this week on Google Ads budget misallocation, the kind of problem that’s frequently bigger than it looks and harder to spot than account managers expect. The analysis flags three common patterns. Performance Max budgets that absorb spend without clear attribution to specific channels. Smart Bidding strategies that over-allocate to low-value conversions because the model is trained on incomplete data. Shared budgets across campaign types that mask which campaigns are genuinely paying their way.

For agency-side audits, the practical move is a budget contribution analysis at the per-channel level inside Performance Max (using the available channel-mix reports), an audit of Smart Bidding’s conversion goals against the actual revenue events, and a review of which campaigns share budgets and what that masks. We’ve found these audits typically recover 8 to 15 percent of effective spend on lower-attention accounts.

WordPress 7.0 Launches With Native AI Integration

WordPress 7.0 launched on 20 May with native AI integration baked into the editor and the platform’s content workflows. This is the response to the Astro market share question raised in the 8 May roundup. WordPress is leaning hard into AI-assisted content production as the differentiator.

The headline features. AI-assisted content generation directly inside the block editor (powered by a configurable model layer that supports OpenAI, Anthropic, Google, and self-hosted models). AI-suggested internal linking based on the site’s existing content. AI-generated alt text and image captions. A new “topical authority” view in the dashboard that maps site coverage against a topic graph.

For the WordPress install base, this matters most for SMB and mid-market site owners who don’t have in-house engineering or content engineering teams. The features that previously required a paid plugin or a custom integration now ship in core. Yoast’s AI Content Planner launched two weeks ago into a market that just got more competitive overnight; expect the WordPress SEO plugin ecosystem to consolidate fast over the next two quarters.

Strategic Direction: Where Search Marketing Is Heading

Three threads from this week’s announcements tie together.

One. Google’s I/O 2026 is the loudest signal yet that the company is converging Search and AI into a single product, not running them as parallel surfaces. The Intelligent Search Box, Gemini 3.5 Flash as default, Search Agents launching this summer, SynthID verification in Search, Universal Cart across surfaces. Each one of those is a meaningful product update. Together they say Search, as agencies have known it, is being rebuilt on AI foundations. Optimisation strategy needs to reflect that.

Two. AI visibility measurement is finally getting actionable data layers. Microsoft Clarity now shows grounding queries. Google Analytics added the AI Assistant channel two weeks ago. iPullRank’s Personal Intelligence study gives agencies a structural reason to invest in owned audience touchpoints. Victorious’s 90% zero-mention finding gives agencies a sharp pitch for AI visibility services. The measurement infrastructure that was missing in 2025 is showing up now.

Three. The PPC surface is fragmenting in ways that benefit Google’s tighter products and pressure agency margins. AI Mode ad formats sit inside AI answers. Asset Studio generates creative inside the platform. Ask Advisor handles cross-product questions. Meridian models attribution end to end. Each of these reduces the surface area where agencies add value through the tool layer and increases the surface area where agency value is in strategy, judgement, and creative direction. That’s a coherent shift, not a passing trend; agency value propositions need to align.

The May 2026 core update may dominate the next two weeks of client conversations, but the larger story this week is structural. Search is becoming Search Plus AI, advertising is becoming AI Plus Advertising, and agencies need to plan their next 12 months around what that means.

Key Takeaways

• Document client rankings now ahead of the May 2026 core update settling. Hold off on remediation decisions for seven to ten days minimum, and avoid attributing ranking changes definitively to either the core update or the I/O changes until late June.
• Treat the Intelligent Search Box, Gemini 3.5 Flash default, and Search Agents as a single composite event that shifts the AI Mode optimisation surface. Longer queries, multimodal inputs, and agentic flows need their own measurement and creative track in client strategies.
• Audit which clients have Personal Intelligence opt-ins reachable through their newsletter and transactional email programmes. Owned email investment now has measurable second-order returns on AI Mode brand visibility for opted-in users.
• For retail and DTC clients not yet integrated with UCP, model the competitive cost of staying out as Universal Cart rolls into Canada, Australia, and the UK through 2026. Shopify merchants get this integration relatively easily; mid-market retailers should evaluate.
• Pull Microsoft Clarity grounding queries reports for any clients with Bing or Copilot-driven visibility. Use the grounding queries to refine content strategy beyond the user-typed keyword.
• For ad accounts running Performance Max, conduct a budget misallocation audit this quarter. Recoverable spend is typically 8 to 15% on lower-attention accounts.
• Add llms.txt to the technical SEO audit checklist as low-cost hygiene, not a ranking lever.

Frequently Asked Questions

How worried should I be about the May 2026 core update?

Not unusually worried, but vigilant. It’s a standard core update, the second of 2026, and Google’s guidance hasn’t changed: helpful, people-first content, no specific recovery actions for sites that lose ranking. The complication this time is that it landed the same week as Google I/O and AI Mode product updates, making clean attribution of ranking changes harder. Hold off on definitive interpretation until mid-June.

What does the new Intelligent Search Box and Gemini 3.5 Flash default change for SEO?

The query surface is shifting toward longer, more conversational, more multimodal queries. Gemini 3.5 Flash being the default means lower latency for AI Mode responses, which (combined with the new search box accepting images, files, and Chrome tabs) reshapes what queries look like in practice. SEO strategies that have optimised for short-keyword search need a parallel track for conversational query patterns, multimodal answer-readiness, and (from this summer) agentic discoverability.

Should brands worry about AI Mode ads cannibalising organic citations?

Yes, in the medium term. The embedded ad format in particular places ads inside the AI recommendation set rather than alongside it, which competes for the exact slot that was previously a free organic citation surface. Conversational Discovery ads sit alongside answers, so they’re less directly competitive with organic. Both formats are still in testing, so the practical impact is uncertain, but PPC and SEO planning need to coordinate on AI Mode strategy.

What’s the practical action on iPullRank’s Personal Intelligence finding?

For consumer-facing brands (especially preference-led verticals like apparel, beverages, and electronics), audit your owned email programme for brand-seed quality. Transactional emails, receipts, newsletters, and product recommendations are now influencing AI Mode brand visibility for users who opt into Personal Intelligence. Useful content tied to your brand, sent regularly to an engaged email list, has a measurable second-order benefit on AI visibility that didn’t exist 12 months ago.

Is WordPress 7.0 enough to stop the Astro drift?

Partially, for the SMB and mid-market install base where engineering capacity is the constraint. The native AI features remove the need for several paid plugins and lower the barrier to AI-assisted content production. For high-performance content sites with engineering teams, Astro’s static-first architecture is still the technical winner. The WordPress drift will probably slow rather than reverse.

Conclusion

This week was structural. Google’s I/O and Marketing Live announcements together describe a Search and AI Mode that is materially different from what existed seven days ago, and the May 2026 core update on top of all that is going to give the industry plenty to argue about through June. Measurement infrastructure for AI visibility is showing up across Microsoft Clarity, GA4 from two weeks ago, and Google’s own analytics stack. Personal Intelligence is changing brand visibility in ways that reward investment in owned audience touchpoints. Universal Cart is collapsing the retail buying journey onto Google’s surfaces. WordPress is leaning into AI as the platform differentiator.

For agencies, the practical move is to update client strategies for the AI Mode surface (longer queries, multimodal inputs, agentic flows), refresh measurement frameworks around the new data layers, and use the May core update window to set realistic expectations on attribution until the dust settles. None of this is panic territory. All of it is now.

Need help adapting your search strategy for the AI era? Contact the Anicca team for expert SEO and PPC guidance.