Artificial intelligence has become central to Google Search. What was once limited testing has now transformed into AI Overviews (SGE), AI Mode, and a fundamental shift in how the search engine responds to user queries.
The architect of this evolution, Robby Stein, head of AI Search at Google, explained it clearly: query behavior has evolved, user preferences have changed, and SEO must adapt to a new reality where AI serves as both a discovery engine and a preliminary layer before visiting websites.
This article analyzes Stein’s statements in detail and their implications for publishers, agencies, and brands relying on organic traffic.
From Long-tail to Ultra-specific: How Queries Are Changing
For years, content creators focused on long-tail SEO: long, specific keywords rather than generic ones. But, according to Stein, that approach is now insufficient.
“People no longer just search for things to do in Nashville, but for outdoor terrace activities in Nashville with a dog and allergies.”
This is the new landscape: long, conversational queries with multiple conditions. Users no longer hesitate to be specific because they trust that Google, powered by AI, can understand the complexity and deliver relevant results.
This textual shift is complemented by another equally powerful one: visual and multimodal search. Features like Google Lens or Circle to Search enable users to utilize their camera and screen as input methods. According to Google, use of this mode grows at a 65% year-over-year rate, especially among younger users who naturally navigate among photos, voice, and text.
Are Clicks Going Away? Google Insists They’re Not
The industry’s main concern is clear: if Google responds with AI, why would users click? Stein addressed this directly:
- Connecting to websites remains a core design principle.
- Google still directs “billions of clicks daily” to sites.
- AI doesn’t replace traffic; it filters and qualifies it: users who click after an AI Overview are better informed and have clearer intent.
In fact, Sundar Pichai announced that AI Overviews had over 1.5 billion monthly users by early 2025. Some outlets reported figures over 2 billion in summer, though Google hasn’t officially confirmed this. What’s certain is that no search feature has seen such rapid adoption in such a short time.
Ranking and Signals: Nothing Changes… and Everything Changes
Stein was clear: AI systems use the same ranking signals as traditional search. There are no alternative algorithms or secret factors.
This means SEO continues to rely on:
- Content quality and usefulness.
- Topic authority and E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness).
- Backing with verifiable sources.
- Satisfying the search intent.
The difference now is in how users formulate queries and the fact that the search engine now includes a layer of AI that summarizes, contextualizes, and selects which links to display as references.
What Does “Useful Content” Mean in 2025?
Stein’s interview also clarified what Google considers useful content in this new era:
- Specific: vague is not enough; it must answer the exact question.
- Reliable: based on facts and trustworthy sources.
- Clear: well-structured, unambiguous.
- Cited and supported: with external references backing it up.
- Direct: must address the query without unnecessary filler or detours.
In short: less “SEO stuffing,” more verifiable, cited, and differentiated content.
The End of “AI Slop”: Goodbye to Low-Value Content
One of the most notable parts of the interview was the confirmation that Google has strengthened its systems to detect and penalize what the industry calls “AI slop”: mass-generated, low-value content designed to manipulate rankings and generate traffic without real utility.
- Google does not differentiate between human or AI-generated content; it only measures usefulness.
- Recent updates make it much harder for filler pages to manipulate results.
- Many sites abusing this model have seen sharp drops after the June core update and the August spam crackdown.
Google aims not just to respond faster but to respond better and more personally.
Deep Personalization
Stein shared that Google is working on experiences to “deeply understand the user” to deliver tailored responses based on their context: travel, shopping, history, or preferences.
For example: when planning a trip with his young daughter, Stein asked Google for accessible beaches with compatible schedules and parking. The AI returned a comprehensive, personalized itinerary.
AI Mode: Conversational Search
This mode allows keeping a dialogue with the search engine. Users can ask follow-up questions, refine queries, and explore complex needs.
Deep Search: Advanced Investigation
Designed for queries with multiple variables: comparing products with technical criteria, analyzing a market, or understanding multifaceted purchase decisions.
Multimodality: The New Normal in Search
The interview transcript revealed how search is now multimodal:
- Long, conversational text.
- Photos and screenshots.
- Real-time voice (live voice search, in testing in the US).
The pattern is clear: search is no longer limited to a text box. It adapts to how we think and ask questions in real life.
SEO Keys to Survive (and Thrive) in 2025
1. Optimize for Compound Queries
Think about users combining multiple requirements in a single search. Develop content with FAQs, filters, and concrete examples.
2. Build Credibility (E-E-A-T)
- Authors with recognized expertise.
- Verifiable source citations.
- Transparency about methods and limitations.
3. Add Value After AI Overviews
Expect Google to handle the basics. Your advantage lies in what follows: interactive guides, calculators, in-depth comparisons.
4. Manage Visual Content
Camera searches demand clear, optimized, and descriptive images (attributes, variants, sizes, materials).
5. Measure Beyond CTR
Assess session quality: users arriving via AI Overviews tend to bounce less and convert better.
An Inevitable Transition for Publishers and Brands
Some publishers report drops in organic traffic, though Google emphasizes that AI expands discovery rather than diminishes it.
The reality is that media outlets and brands must Diversify their channels (newsletters, social media, podcasts, subscriptions) and simultaneously optimize content for the new AI Search ecosystem.
Conclusion: SEO Evolves, It Doesn’t Die
Robby Stein’s message is clear: Google AI Search doesn’t eliminate SEO; it transforms it.
- Clicks remain essential, but now they come after a layer of AI that adds context.
- Useful, specific, and trustworthy content remains key to ranking.
- What changes is how users ask questions and how we should respond.
The future of SEO lies in understanding user specificity, strengthening credibility, and designing experiences that provide value beyond AI Overviews.
FAQs
What is Google AI Search and how does it impact SEO?
It’s the evolution of Google Search with integrated AI (AI Overviews and AI Mode). It doesn’t eliminate SEO, but it shifts how to optimize: users now ask longer, multimodal, contextual queries.
How much has visual search grown on Google?
According to Robby Stein, visual search with camera and Circle to Search has grown 65% year-over-year, mainly among younger users.
What does Google consider useful content in 2025?
It must be specific, reliable, sourced, clear, backed, and directly address the user’s question.
What is AI Slop and how does Google fight it?
It’s low-value, mass-produced content generated with AI that offers little real utility. Google has strengthened its systems to reduce its visibility in results.
via: SEO and Google news