The internet is radically changing. With Google AI Overviews rolling out globally and AI Mode giving users a dedicated way to search through generative AI, search is entering its biggest shift since Google’s launch. Users can now get synthesized answers directly in search – and while the clickthrough path may shrink, the foundations of search engine optimization (SEO) are more important than ever.
To cut through the noise, we spoke with SEO expert Ian Pereira, founder of position1, who has been helping global brands adapt their strategies for AI-powered search. Ian explains what AI Mode really is, how it works, and why SEO fundamentals remain the bedrock of AI visibility.
Proofed: Ian, can you start by giving us a quick overview of what Google’s new AI features mean for SEO right now?
Ian Pereira (IP): It’s clear that SEO fundamentals haven’t gone away. In fact, they’re more critical than ever. Google’s AI Overviews and AI Mode rely on the same signals we’ve always worked on: accessible websites, structured data, strong site architecture, and authoritative content.
Google has long been able to highlight specific paragraphs in snippets or link users directly to a section of a page. What’s new now (and not just at Google, but also with agents like ChatGPT, Perplexity, and Claude) is the scale and synthesis. These systems can break a query into dozens of sub-questions, pull together passages from multiple sources, and then generate a contextual answer. That means every section of your content should be optimized to stand on its own, because any passage could be the one these AI agents select.
Proofed: For clarity, what exactly is AI Mode?
IP: Google announced AI Mode in March 2025. It’s a dedicated tab in Google Search that uses generative AI (the Gemini model) to deliver synthesized answers with citations, follow-up prompts, and exploration tools. Where a user might once have run three or four different searches, AI Mode breaks down the query into many sub-queries – sometimes dozens – and gives the user a comprehensive, conversational answer. It’s not replacing traditional search results, but it’s offering an alternative mode for more exploratory queries.
Proofed: Some people say AI search is highly personalized – even pulling from Gmail or Calendar. Is that true?
IP: By default, no. Google does not use your Gmail or Calendar data in AI answers unless you explicitly opt in, which is a feature coming soon. Ultimately, AI might help you surface personal information like flight bookings or reservations, but outside of that opt-in, personalization is based mainly on your location and recent searches. For other AI agents, such as ChatGPT, results using the live web search tool are not personalized to you, although it may take into account any previous conversations you’ve had with the AI.
Proofed: How are AI results different from traditional search results?
IP: Traditionally, Google matched queries to pages. For years, it has been able to highlight specific paragraphs through featured snippets and ‘scroll-to-text’ highlights. What’s different with AI Mode is how far this goes: Google doesn’t just show a snippet – it performs a “query fan-out”, breaking your question into many sub-queries, retrieving passages from multiple sites, and then reasoning across them to generate a synthesized answer. You’ll often see summaries, bullet points, and suggested follow-ups. Links are still there – but they’re embedded in a broader, AI-constructed explanation rather than a simple list of results.
Proofed: There’s been a lot of hype that SEO is “dead” because of AI. What’s your view?
IP: That’s an overstatement. If anything, AI is showing how much SEO underpins these systems. Technical SEO, structured data, content clarity, and authority are what make your site AI-referenceable. AI Mode doesn’t invent answers out of thin air – it relies on trusted, optimized sources. In practice, Google, Claude, Perplexity, and ChatGPT combine retrieval (traditional search results) with generation (its AI model synthesizing the information). If your site isn’t JavaScript-crawlable, clear, and authoritative, you won’t appear in traditional results or in AI answers.
Proofed: So how should businesses adapt their content strategy for AI-driven search?
IP: Think in terms of passages, not just pages. Each paragraph should be concise, self-contained, and answer a specific intent. Avoid fluff. Structured formatting – headings, lists, schema markup – makes it easier for AI to pull your content. And think about the “query fan-out”: If someone searches “best proofreading service,” Google might also spin out related questions on price, industry fit, or comparisons, and then use the results of these additional queries to inform its synthesized response. You want your site to cover those angles too.
Proofed: What about different content formats – is text enough?
IP: Text is still the foundation, but content diversity matters. Google has said that AI overviews and AI Mode are multimodal (i.e., they can process and understand multiple types of data simultaneously, such as text, images, audio, and video), so you should support your text with high-quality images, videos, and even tools when relevant. That helps with user experience, but also makes your site more comprehensive.
Proofed: How do we measure performance if AI Mode doesn’t send as much referral traffic?
IP: Measurement is the hardest part right now. Early on, clicks from AI results often showed up as ‘Direct’ in analytics because the referrer wasn’t passed through. Google has since updated Search Console: AI clicks are counted under the Web search type – but they aren’t broken out separately, so attribution is still limited.
This is why the industry is moving toward AI visibility tracking: tools that measure how often your content is cited in AI answers, not just where you rank in the 10 blue links. New metrics will include things like:
Together, these will give us a truer picture of visibility in an AI-first search landscape.
Proofed: If you had to give one piece of advice to businesses worried about AI search, what would it be?
IP: Don’t abandon what works. SEO isn’t dead – it’s the foundation of AI search. Double down on clarity, technical excellence, and authoritative content. Then, adapt it. Structure your content so that every section could be the one that gets quoted in an AI Overview. That way, whether a user clicks through from the classic blue links or gets an AI-generated answer, your brand is still visible.
Ian’s insights reveal a search landscape that rewards precision over volume and relevance over generic optimization. For businesses and content creators, several actionable strategies emerge from this conversation:
While the specifics of AI-powered search continue to evolve, Ian’s core message remains clear: The businesses that will thrive are those that prioritize user value over gaming algorithms. By focusing on creating genuinely helpful, well-structured content that directly addresses user needs, companies can position themselves to succeed regardless of how search technology continues to develop.
The transition to AI-driven search represents both a challenge and an opportunity. For content creators willing to adapt their strategies (to emphasize clarity and relevance), this new landscape offers the potential to develop deeper, more meaningful connections with their audiences.
As we navigate this transition, the most successful approach may be the most traditional one: to create content so valuable and well crafted that both humans and AI systems naturally want to recommend it.
Absolutely not. As Ian emphasized, technical SEO fundamentals are more important than ever because AI systems require clean, accessible data to work with. Keep doing the basics – just adapt your content strategy for passage-level optimization.
This is challenging right now. AI referrals often show up as “Direct” traffic in analytics, which makes attribution difficult. Consider persona-based testing – searching as your target customer would while logged into their typical accounts – to see how often you surface.
Instead of thinking about entire pages ranking for keywords, focus on individual paragraphs or sections that directly answer specific questions. Each passage should be self-contained and valuable, capable of standing alone as a complete answer.
For core content, yes. While some boilerplate elements are inevitable (e.g., navigation or legal disclaimers), you should specifically tailor your main content to each audience rather than just swapping out keywords in the template copy.
Very important. Google’s patents suggest AI Mode favors sites with rich content ecosystems – videos, photos, infographics, and interactive tools – not just text. This diversity signals authority and provides multiple ways for AI systems to reference your expertise.
They’re already struggling with personalized AI results. You’ll likely need to supplement traditional tools with broader brand monitoring, direct traffic analysis, and persona-based testing to form a complete picture of your search visibility.
Want to bite down on your content crunch?
Let’s talk about the support you need.
Δ
We use cookies to give you the best possible experience with Proofed. Some are essential for this site to function; others help us understand how you use the site, so we can improve it. We may also use cookies for targeting purposes.