aiFeatured

Google used this week's I/O developer conference and Google Marketing Live to execute its most comprehensive AI-driven product overhaul in years

Friday, May 22, 2026·5 min read·19 sources

The Signal

Google used this week's I/O developer conference and Google Marketing Live to execute its most comprehensive AI-driven product overhaul in years. The company unveiled two new ad formats for AI Mode — conversational discovery ads and highlighted answer ads, both labeled "Sponsored" — alongside Shopping ads integration, a Business Agent for Leads, and an expansion of the Direct Offers pilot, all slated for AI Mode rollout in the coming months. On the search infrastructure side, Google launched agentic capabilities including information agents and a Universal Cart, fundamentally repositioning Google Search as a transaction layer, not just a discovery surface. Simultaneously, Google began rolling out the May 2026 Core Update on May 21 — its second broad core update of 2026, following March's update — with a completion window of up to two weeks. The update carries Google's standard framing: better surfacing of relevant, satisfying content across all site types. Beyond its own platforms, Google's AI momentum is reaching into the broader ecosystem. Digital Turbine announced a deepened Google Cloud partnership, embedding the Gemini Enterprise Agent Platform into the intelligence layer powering its ad and publisher solutions across more than one billion devices and 80,000-plus apps. Separately, Meta is forecast to surpass Google in net worldwide ad revenue this year — $243.46 billion versus $239.54 billion — the first time that threshold has been crossed.

Why It Matters

The real story here isn't Google's new ad formats — it's that Google is restructuring Search into a toll road, and most marketers are still debating the scenery. Conversational discovery ads and highlighted answer ads aren't incremental additions to the existing auction model; they're Google's bid to own the transaction layer before users ever reach a brand's own surface. The Universal Cart and Business Agent for Leads confirm this: Google isn't just surfacing intent anymore, it's capturing conversion. What surface-level coverage misses is the pressure driving this. Meta surpassing Google in net ad revenue this year — for the first time, ever — is not a footnote. It's the competitive gun pointed at every product decision Google made at I/O this week. Google's AI Mode buildout is, in large part, a defensive maneuver dressed up as innovation. When your core revenue base is threatened by a platform that's mastered social commerce and lower-funnel performance, you don't tinker — you rebuild the monetization architecture entirely. The May Core Update landing simultaneously is worth noting too. Two broad core updates in eight weeks, alongside a fundamental rewiring of how ads surface in AI Mode, means the ground is shifting on multiple axes at once. SEO strategies calibrated to March 2026 results are already outdated. For decision-makers, the strategic question isn't whether to run ads in AI Mode — that's table stakes within the next two quarters. The real question is whether your brand can afford to let Google become the primary relationship owner in your customer's purchase journey. Every dollar flowing through Universal Cart is a data point Google keeps and you don't.

The Exploit

🎯

Opportunity

Brands with first-party transaction data and direct purchase infrastructure are positioned to use Google's AI Mode rollout as leverage, not a dependency. Teams that can quantify customer lifetime value outside of Google's attribution window — and pressure-test that math before Universal Cart scales — can negotiate media allocation from a position of strength through Q3 2026.

⚠️

Risk

Performance marketers who migrate budget into AI Mode conversational and highlighted ad formats without auditing what conversion data stays with Google are building a compounding blind spot. Within two to three quarters, those brands will have optimized toward Google's outcome metrics, not their own, and won't have the first-party data infrastructure to course-correct when costs rise or formats shift again.

🚀

The Move

Before August 2026, the CMO should commission a single deliverable: a customer journey audit that maps exactly which conversion signals are currently shared with Google versus retained internally, with a dollar-value estimate of the data asymmetry at stake. Own that number before the Q3 2027 budget cycle opens — it becomes the anchor for every AI Mode investment decision that follows.

First-Mover Advantage

Gains

Risks

Window

Winners & Losers

Winners↑

↑

Google AI Mode Paid Search Teams and Performance Agencies

The introduction of conversational discovery ads and highlighted answer ads creates a new, early-mover premium inside AI Mode — teams that build fluency with these formats in Q3 2026 will establish auction positioning and quality score history before broader competition arrives. The mechanism is straightforward: new ad inventory in a low-competition environment typically yields superior CPCs and conversion rates before saturation. Performance agencies should be piloting AI Mode campaigns now, documenting format-specific creative learnings, and positioning that capability as a premium service offering to clients.

↑

Agentic Commerce and Checkout Optimisation Specialists

Google's Universal Cart and Business Agent for Leads signal that transactional infrastructure is being built directly into Search, creating immediate demand for specialists who understand how to optimise product feeds, lead flows, and checkout signals for agentic environments. The mechanism of advantage is expertise scarcity — very few practitioners currently understand how to structure data and conversion architecture for agent-mediated purchases rather than click-through journeys. Consultants and agencies that develop this capability through Q3 2026 will be positioned as essential intermediaries as Universal Cart scales into mainstream advertiser adoption.

↑

Mobile Ad Intelligence Platforms with On-Device Signal Access

The Digital Turbine and Gemini Enterprise Agent Platform partnership illustrates a broader pattern: Gemini is becoming the intelligence layer embedded across ecosystem partners, extending Google's AI reach to over one billion devices and 80,000-plus apps. Platforms that already hold proprietary on-device or in-app signal environments — similar to Digital Turbine's Ignite Graph — gain compounding advantage as Gemini integration amplifies their targeting and optimisation capabilities. The strategic response for any mobile ad platform with rich first-party signal infrastructure is to accelerate Google Cloud or Gemini partnership conversations before this integration pattern becomes a standard competitive requirement.

Losers↓

↓

Organic SEO Content Practices Calibrated to Pre-2026 Algorithms

Two broad core updates in eight weeks — March and May 2026 — layered on top of AI Mode's expansion means that SEO strategies built on 2025 or early 2026 ranking signals are operating on an invalidated baseline, with no stable measurement window between updates. The mechanism of pressure is compounding: each core update reshuffles organic visibility, while AI Mode simultaneously reduces click-through opportunity for even well-ranked organic results by resolving queries inside the SERP. SEO practitioners must treat their current performance data as unreliable, pause strategy commitments that depend on ranking stability, and begin rebuilding content architecture around satisfying, people-first depth rather than keyword-optimised surface coverage.

↓

Brand Direct Customer Relationship Owners in E-Commerce

Google's Universal Cart and Business Agent for Leads are structural threats to brand-owned purchase journeys — every transaction completed inside Google's ecosystem is a conversion event that Google records and the brand does not, systematically eroding first-party data accumulation and customer relationship ownership over time. The mechanism is a gradual substitution: as AI Mode handles more of the discovery-to-purchase sequence natively, brands become fulfilment vendors rather than relationship owners, weakening their ability to personalise, retain, and upsell independently of Google's platform. The defensive response is to accelerate investment in owned channels — email, loyalty programmes, direct app experiences — that capture post-purchase data and rebuild customer relationships outside Google's transactional layer before Universal Cart reaches mainstream adoption scale.

Strategic Outlook

Google's pivot to owning the transaction layer isn't a sudden move — it's the acceleration of a strategy that's been visible since the introduction of Shopping Graph in 2021 and the quiet rollout of Buy on Google features that predated it. What's different now is velocity and competitive necessity. For years, Google has operated on an implicit gentleman's agreement with the commerce ecosystem: surface intent, take a toll at the click, let brands own the conversion. The Universal Cart and Business Agent for Leads formally retire that arrangement. The more instructive historical parallel here is what happened when Amazon built its advertising business — brands that dismissed it as a retailer eventually woke up to a platform that owned the customer relationship, the purchase data, and the ad inventory simultaneously. Google is attempting the same consolidation before Amazon, Meta, or an AI-native alternative can claim that position in search-driven commerce. The Digital Turbine-Gemini partnership signals something equally important: Google's infrastructure ambitions extend well beyond its own surfaces. Embedding Gemini into a platform spanning a billion devices means Google's AI monetization layer is becoming an industry utility, not just a search product. Two broad core updates inside eight weeks — March completed April 8, May beginning May 21 — suggests Google is comfortable compressing its algorithm adjustment cycles as AI Mode demands a fundamentally different content quality baseline.