On May 19, 2026, Google announced around 100 updates at I/O.
To understand why this moment matters, consider the scale it lands on: AI Mode has already surpassed 1 billion monthly active users, and AI Overviews reach 2.5 billion monthly – more than half the global internet.
In reality, most of them don't matter for your business. But a handful change everything.
In select categories, Google is rolling out the ability for its agents to call businesses and book appointments.
Universal Cart means consumers can buy something from inside Search, Gemini, and YouTube without ever visiting a brand's website
Search box (unchanged for 25 years) now dynamically expands to accommodate longer queries, uses AI to anticipate intent beyond autocomplete, and is now multimodal – it takes text, images, files, videos, and Chrome tabs as inputs.
But the announcement getting the least attention may matter the most: the introduction of information agents. Google is now letting any search become a persistent background agent – one that monitors a topic, a provider, a product, or a decision continuously and alerts when something changes. Think of it like a flight price tracker.
Now imagine that for everything.
This isn't Google catching up to OpenAI and Perplexity. It signals the end of something brands have built around for two decades.
What are the changes by vertical
For many years, search has run on a crawl-index-rank model. A search engine visited websites, stored what it found, and ranked that cached content when someone searched. SEO was built on gaming that index.
What Google is deploying now is the opposite: agents that check sources in real time, continuously, and surface what's current. The question isn't whether you ranked. It's whether your data is accurate and fresh enough that an agent checking right now will surface your brand… and keep surfacing you, every time it checks back.
And what makes Google's agentic announcements different from similar claims by OpenAI or Anthropic is infrastructure.
Google already has Google Pay embedded across Search, Maps, and YouTube
It has Universal Cart live with major retailers
It has business listings, CTAs, and booking links woven into billions of daily interactions
The agentic calling capability has been in development and testing ahead of this rollout
The agents aren't arriving in a vacuum. They're arriving to a platform that was already built for action. For brands, that's the distinction that makes these announcements worth taking seriously right now.
But the implications aren't the same for every industry.
For healthcare systems, the path from symptom search to booked appointment is collapsing into a single AI conversation.
Agents are approaching the point of being able to book care on a patient's behalf. Machine-readable scheduling data, structured provider profiles, and consistent listings across every directory Google trusts aren't a best practice. They're the condition for being in the conversation at all.
AI Overviews are already answering the informational part of the patient journey (referral requirements, consult lengths, insurance questions) before patients reach a health system's site. The systems earning citations there have clean provider pages, structured FAQs, and consistent data, everywhere.
Review quality has become a gate: a weak profile doesn't just hurt conversion, it removes a provider from consideration entirely.
For brands in retail, restaurant, or hospitality, the transaction has left the building.
Universal Cart means a customer can complete a purchase from inside Search, Gemini, or YouTube without visiting a brand's site.
Accurate hours, real-time menu data, and consistent location information aren't search signals anymore. They're the conditions for being included in that transaction at all.
For financial services organizations,
Google's new Business Agent for Leads embeds a conversational AI agent directly inside an ad, replacing the static lead form with a live chat that answers questions based on the brand's own website.
That matters because, from a wealth management perspective, the potential clients most likely to engage are already AI-first: among U.S. households earning $150k or more annually, AI has overtaken traditional search as the starting point for local search.
For financial brands, the first touchpoint with a potential high-value client is increasingly AI-mediated and deeply personalized – which means what AI knows about your branches, advisors, and products determines whether that moment builds trust or loses the client.
What the data shows
The brands being included in AI responses across every vertical share one thing: their data is verified, consistent, and distributed (in real time) across every surface AI systems treat as a trusted signal (website, listings, reviews, local pages). Not because they've done something "for AI", but because they've built the data foundation that AI now requires to act confidently on a customer's behalf.
How do we know? Yext Research has analyzed 17.2 million AI citations from Gemini, Claude, Perplexity, and ChatGPT. The key finding is that citation behavior is predictable – and it runs on brand-managed sources.
The full picture is in Search Is Everywhere. Winning Brands Are Too, but here's the short version: brands that are cited are the ones whose data is ready. The ones that aren't, aren't.
Yext is built for exactly this moment
While citations are key to making sure your brand is part of the AI conversation, this next step to agent monitoring and agent-assisted transactions means brands need to think beyond mere "visibility," and that calls for a different type of platform.
Most platforms address one part of this.
Some manage listings. Some handle reviews. Some help brands build local pages.
None of them connect those pieces into a system where the data an AI agent finds on your website matches what it finds in your listings, local pages, and reviews – and where all of it is kept accurate (and syndicated) everywhere, in real time.
The Yext Knowledge Graph is the verified source of truth for your brand: structured, maintained by data agents, and readable by every AI system that looks for it.
That data foundation fuels the content built on top of it. Yext's content generation agents create and scale structured, locally-relevant content for pages (FAQs, location descriptions) in a format AI engines can read, cite, and act on. Brands don't have to choose between content at scale and content that's accurate. Yext does both from the same source of truth.
That same data is syndicated directly to 200+ publishers – no aggregator in between. What Yext pushes is what AI sees, the moment it checks. Not what was cached last week, and not filtered through a third party that introduces lag or inconsistency.
Scout is Yext's AI visibility agent, monitoring your citation presence across ChatGPT, Gemini, Perplexity, and Grok – surfacing where you're being included, where competitors are outranking you, and what's driving the gap. And with the launch of Scout MCP, your team can now bring your own AI tools directly into that intelligence layer, connecting your LLM of choice to Scout's competitive data without leaving your existing stack. That's agentic marketing executed your way, with Yext's intelligence working inside whatever workflow you're already running.
Reviews and social aren't a reputation function here. Instead, they're trust signals – the layer AI systems use to decide whether to recommend your brand when the query is subjective. Yext manages them as part of the same data foundation, not as a separate channel.
This is data infrastructure for agentic marketing. Not a collection of point solutions or tools, but instead one platform where intelligence, verified data, and distribution work from the same source of truth.
What to do now
The Google I/O announcements don't require brands to do more marketing. They require brands to get their data right – everywhere AI looks, all at once, in real time.
The index model that search was built on is ending. Agents are now checking your data all the time, multimodal queries are replacing keyword searches, and personal context is shaping what gets recommended.
What wins in this environment isn't better optimization for a static ranking. It's verified, structured, real-time data.
That's what Yext is built for. And the brands that build that foundation now will be the ones AI recommends next – and keeps recommending, every time the agent checks back.
