I left Google Search Central Live in Toronto with a clear takeaway: Search is becoming more selective about what content gets indexed, not just how it ranks, and that changes the role of marketing. It’s no longer about producing more content. It’s about creating content that is distinct, structured, and clear enough for machines to trust and use.
We’re not just optimizing for a system that retrieves links anymore. We’re optimizing for systems that interpret content, summarize it, and decide what actually gets surfaced.
This came up repeatedly throughout the day from speakers such as Danny Sullivan, Martin Splitt, and Ryan Levering. Across topics like AI, indexing, and structured data, the message was the same: Clarity, accuracy, and differentiation matter more than ever.
Here’s what stood out to me and how we at Schema App are thinking about it.
Structured data remains critical for AI and search
In true Martha fashion, let’s start with some good news!
There’s been a lot of noise in the industry about whether structured data still matters in an AI-first world. Google addressed this directly, and it was reassuring.
They continue to strongly recommend structured data, not just for SEO but also for AI readiness. And the reasoning was practical. They highlighted four specific reasons why structured data matters:
1. Precision
Structured data provides high precision for complex content. It delivers more accurate understanding than relying on large-scale LLM extraction alone.
2. Detail
It allows you to define information that isn’t visible on the page, like identifiers, metadata, and structured relationships. This additional context helps systems better understand your content.
3. Cost Efficiency
Parsing structured data is significantly more efficient than repeatedly extracting meaning using LLMs.
4. Focus
Structured data highlights what matters, reducing noise and helping machines avoid pulling in irrelevant information.
What stood out to me is that none of this is new. It’s just more important now. Yes, LLMs can generate and consume schema markup. But they also hallucinate. They make assumptions. And when it comes to how your brand is understood, “probably right” isn’t good enough.
If you care about how machines interpret your content, you need to be explicit.
Google also shared that they will begin publishing more validation rules for rich results using SHACL and ShEx. For us at Schema App, this was a familiar direction. We’ve been applying these types of rules at scale for years to ensure accuracy and governance.
Content still matters, but the standard is higher
One thing that came through very clearly at the event is that content still plays a central role. That hasn’t changed. What has changed is the standard.
Google talked a lot about what they called “non-commoditized content,” and I found that framing helpful. They’re looking for content that brings something distinct, whether that’s a specific perspective, a clear answer to a user need, or an angle that isn’t already well covered elsewhere.
In practical terms, that means being more intentional about what we create. Not just publishing to have coverage, but asking whether the content actually adds something new.
AI is raising the bar here. It doesn’t need another summary of what already exists. It needs content it can rely on and use with confidence.
They also reinforced something I think is important to keep in mind as we all experiment with AI. It’s a great tool for research and ideation, but the perspective and insight still need to come from people. That’s what makes the content valuable in the first place.
If your content isn’t unique, it may not be indexed
What really stood out to me was how this connects to indexing.
There was a moment when Google reframed the conversation simply: ranking only happens if you’re indexed. And indexing is where the real filtering is happening now.
If your content is too similar to something that already exists, there’s a good chance it won’t be included at all. Google doesn’t need another version of the same answer. That’s where this idea of non-commoditized content becomes more than just a best practice. It’s a requirement.
It also puts into perspective the opportunity in search. With 15% of queries being new every day, there’s no shortage of things to create content about. But the opportunity isn’t in producing more. It’s in producing something different and useful.
That’s really the shift I took away. It’s not just about creating content. It’s about creating content that earns its place.
AI has changed what “good traffic” looks like
Another shift that stood out is how we think about performance.
Google shared that when users click through from AI Overviews, they tend to spend more time on the site and arrive with more context. That aligns with what we’re starting to see as well.
It suggests that AI is doing some of the qualification work upfront. So while traffic patterns may change, the quality of that traffic can actually improve. This is where I think a lot of teams will need to adjust their mindset.
If we’re still measuring success purely on volume, we might miss what’s actually happening. Engagement, conversions, and outcomes will matter more.
See our case study on how Entity Linking increased AIO visibility by 19.72%
Google debunked common SEO myths
I appreciated that Google took the time to address some of the circulating myths and assumptions around performing in AI search. Things like chunking content for AI, optimizing for conversational keywords, or relying on markdown as a strategy.
Their guidance was refreshingly simple. Write for humans. Structure your content clearly. Focus on what actually helps users. They also confirmed that things like LLMs.txt don’t currently have an impact, and that Google is comfortable handling JavaScript.
The takeaway for me was that we don’t need more tactics. We need better fundamentals.
So what should we focus on?
When Google summarized where to invest, it wasn’t complicated. Focus on:
- Creating non-commoditized content
- Maintaining strong SEO fundamentals
- Delivering a great page experience
- Implementing robust structured data
They also pointed to growing opportunities in areas such as shopping, local, video, and image search, where structured data can play a larger role.
What’s next for AI Search? Prepare for the Agentic Web
There was some discussion about the “agentic web,” which is where things are heading.
The message here was clear. It’s still early.
Google shared a couple of examples, such as Business Agents and the Universal Commerce Protocol (UCP), particularly for shopping experiences. Business Agents is a new way for shoppers to chat with brands within Google Search. This is currently available for some US brands and is activated in Merchant Center.
UCP will soon power a new checkout feature on eligible Google product listings in AI Mode in Search and the Gemini App.
That said, they were also clear that they will continue to guide us on how to participate as this evolves.
That’s something we’ve been deeply focused on at Schema App, helping our customers build a strong semantic data layer as their foundation and preparing their content to be more accessible and interoperable with AI systems. This includes aligning to emerging standards like Model Context Protocol (MCP), which are shaping how agents will access and use structured information.
Because as these agentic experiences evolve, the brands that will lead are the ones that can show up clearly, accurately, and authoritatively. If you’re thinking about what this means for your own strategy, we’ve put together a practical guide on how to prepare.
Download our free eBook: The Agentic Web: What Marketers Must Do To Prepare.
Key Takeaway From Google Search Central Live Toronto
If I had to summarize the day, it would be that search is moving from retrieval to understanding.
That raises the bar for all of us. It’s no longer enough to publish content and hope it performs. We need to be intentional about how our content is understood, not just how it’s written.
At Schema App, this reinforces what we’ve been seeing with our customers.
The brands that are going to win in this next phase are the ones that:
- Create content with a clear point of view
- Structure their data so it’s easy for machines to interpret
- Maintain a consistent, accurate, brand-controlled source of truth
Because in a world where AI is interpreting your content, your visibility depends on how well you’re understood.
Thank you to the Google team for a great event and open conversations throughout the day.
What I always find most valuable about these sessions isn’t just what’s new, it’s the validation. When you’re working closely with customers and forming a point of view on where search is going, it’s helpful to hear how that aligns with how Google is thinking about the future.
For me, this event reinforced that we’re on the right path. The focus on structured data, content clarity, and helping machines truly understand information isn’t just a theory; it’s where search is actively heading.
Martha van Berkel is the co-founder and CEO of Schema App, a semantic technology company that helps enterprise brands control how they are represented in AI and search. By building a semantic data layer, Schema App enables organizations to improve visibility, protect their brand, and drive more efficient organic growth. She focuses on helping marketing teams navigate the shift from traditional SEO to AI-driven discovery by establishing a structured source of truth for their content.