Google's Gemini pushes search and commerce into the operating system

Gemini Intelligence, introduced at the Android Show on May 12, runs across Android devices with the promise of automating complex tasks, summarizing web content, and simplifying form filling. Google is moving search and shopping out of the browser and into the layer beneath the operating system. For SEO agencies, visibility is shifting from ranking pages to being legible to software that can act for the user.

Search is becoming an operating-system feature

Gemini Intelligence is a proactive AI layer for Android, not a standalone chatbot bolted onto a search box. Google says it will appear across phones, watches, laptops, and glasses, and it introduced Googlebook as a new laptop category designed specifically for Gemini Intelligence and synced to an Android phone. Smart glasses with Gemini will launch later in 2026, and Wear OS 7 devices will get Gemini Intelligence later in the year.

The user journey no longer has to start with a browser tab. An agent can understand what is on the screen and take action, whether that means scheduling a meeting from an email date or helping someone visualize furniture in a living room.

The scale is already large enough to change strategy

Google says the Gemini app is now used by more than 900 million people every month across 230 countries and more than 70 languages. The same app now includes Project Astra capabilities such as screen sharing and live video streaming on Android, which gives Gemini more context than a text prompt ever could.

Google says Gemini can connect to apps including YouTube, Google Maps, Gmail, Spotify, Calendar, Keep, Tasks, Messages, Phone, and others.

SEO now has to speak machine and human at the same time

Product structured data can help product information appear in richer Search results, including Google Images and Google Lens. Structured data has to follow Google’s guidelines to be eligible for rich results, and eligibility does not guarantee appearance.

For ecommerce, product data has to be clean enough for machines to consume without guessing. For lead generation, service pages need clear entity signals, location data, hours, appointment paths, and pricing language that an assistant can parse quickly. For content sites, that means tighter topical organization, explicit naming, and a site structure that makes it easy for an agent to pull the right answer without falling back to a vague summary.

Commerce is moving from pages to protocols

Google’s Universal Commerce Protocol, or UCP, makes the direction even clearer. UCP is an open standard for agentic commerce, with a common language and functional primitives for commerce journeys between consumer surfaces, businesses, and payment providers. It is a transaction framework designed for software that can compare, prefill, and complete a purchase on the user’s behalf.

Google has also announced conversational shopping updates in Search and agentic checkout tools for retailers. Product visibility, merchandising, and conversion can no longer be owned by separate teams that barely talk to each other. If the agent can move from discovery to checkout inside a single flow, then product feeds, inventory accuracy, pricing consistency, and payment readiness become part of organic strategy.

The search stack is already being rebuilt

In May 2026, Gemini 3.5 Flash became the default model in AI Mode in Search. Google is not treating AI search as an overlay. It is replacing the model under the hood, which changes how results are generated, summarized, and acted on.

Google had already been testing Gemini in Search since December 2023, and by 2025 it had added Deep Search and agentic local calling features.

What agencies need to change now

A client’s content, product data, and local presence have to survive in a world where an assistant decides whether the brand is usable.

• Treat structured data as core site architecture, not a final SEO task.

• Audit product pages, service pages, and location pages for machine readability: clear entities, consistent naming, precise descriptions, and crawlable markup.

• Align Merchant Center, site feeds, local listings, and CRM data so pricing, availability, hours, and contact details do not conflict.

• Build conversion paths that work when the user never sees a classic SERP, including fast checkout, saved carts, direct booking, and click-to-call.

• Measure visibility beyond rankings, including appearances in AI Mode, Images, Lens, shopping flows, and local actions.