Google is pushing AI Overviews further toward a conversational experience, with follow up questions now opening directly into AI mode.
At the same time, Google is rolling out Gemini 3 as the default model powering AI Overviews worldwide.
Implications for the search experience
The primary piece notes that when AI Mode overlays the search results, sources can become less prominent or presented differently, which is a meaningful concern for brands that rely on Search as a primary acquisition channel.
The “second query” is increasingly happening inside Google’s own interface rather than across the open web. Follow-up questions, where users typically refine their intent and generate high-value clicks, are now more likely to open in AI mode, keeping a greater share of the journey within Google.
This isn’t just a UI update, it reflects a broader product strategy. Multiple sources describe Google narrowing the gap between search and chatbot experiences, allowing users to move from a “quick snapshot” to a deeper conversation without changing content.
Improvements in answer quality won’t automatically translate into traffic gains. While Gemini 3 may improve response quality, more complete in-SERP answers can still reduce clicks when user intent is fully satisfied in search.
Key viewpoints: points of alignment and divergence
Where sources align
Search is becoming conversational by default. Google, Search Engine Journal, and The Verge all describe a smoother progression from AI Overviews into dialogue, moving users from a high-level snapshot to deeper interaction without a context switch.
Gemini 3 now underpins AI Overview quality. Across coverage and in Google’s own announcements, Gemini 3 is consistently positioned as the core model driving improvements in response quality.
Follow-ups act as the bridge to deeper exploration. Google explicitly emphasizes context carryover, while its help documentation outlines AI mode’s “fan-out” approach, reinforcing follow-up questions as the mechanism for expanding intent.
Where perspectives diverge
Search Engine Land takes a more direct stance, warning of fewer clicks and calling out UX choices that may suppress outbound traffic.
Google’s framing positions AI Overviews and AI Mode as tools for exploration, emphasizing “prominent links” and higher-quality clicks rather than volume
The Verge frames this as part of a broader shift away from links as the primary organizing principle of Search.
Google, by contrast, presents the change as an expansion of Search’s ability to support more complex journeys, not a replacement for the open web.
From Search Results to Search Dialogues: What Google’s AI Mode Handoff Changes for Acquisition
Historically, the first query is imprecise. Users start broadly, scan results, then refine. That refinement is where intent sharpens and where brands often earn the click by aligning with what the user is actually looking for.
Now, Google is routing those follow-ups directly into AI Mode from AI Overviews, creating a more seamless “overview to conversation” flow. At the same time. Gemini 3 is becoming the default model behind AI Overviews globally, raising the bar for response quality and consistency.
For marketers, the feature mechanics matter less than the behavioral shift: the second question is increasingly handled within Google’s AI interface rather than distributed across the open web.
Why does this change the economics of visibility?
AI Overviews have already reduced many of the “easy clicks” tied to informational questions. This update extends that pressure further down the funnel, into the exploratory loop where users move from a broad answer to questions like, “Okay, but what about…?”
Search Engine Land articulates the concern most marketers intuitively feel: if follow-up questions are increasingly handled inside AI mode, publishers and brands should expect fewer opportunities to earn the click.
Google’s counter-position, repeated across its own materials, is that AI-enhanced journeys can drive higher-quality visits, arguing that when AI helps users land on the right page, those clicks are more intentional and more valuable, even if they’re fewer overall.
What “ranking” looks like in a conversational Search journey.
Ranking in search used to be relatively linear: one question, one results page, a finite set of links competing for attention.
AI mode breaks that model.
In Google’s own framing, AI Mode uses a fan-out approach. A single question is deconstructed into multiple subtopics, several searches are run in parallel, and the response is synthesised with supporting links. Context also carries forward across turns. The result is a conversational journey, not a sequence of isolated questions.
For brands, that changes what visibility actually means. You can appear in an initial overview and then disappear entirely based on how well your content supports the next question, not just the first one.
Two implications marketers need to internalize
You’re no longer optimising for a single question. You’re optimising for a chain of questions. If your content only answers the headline question, you may gain visibility early, only to drop out as the user refines their intent. Pricing, comparisons, limitations, alternatives, and regional availability, these are no longer edge cases, they’re the natural continuation of the conversation.
Coverage gaps get exposed faster. Conversational Search is unforgiving of thin coverage. If a competitor has a solid “X vs Y” page and you don’t, AI Mode has an easy reason to route the user elsewhere.
Strategy: how marketers should respond (without panic)
This shift doesn’t invalidate SEO fundamentals. It does, however, require a more realistic view of how discovery now works.
This means proactively addressing the questions people almost always ask next: trade-offs, ideal use cases, who it’s not for, setup time, cost ranges, integrations, compliance and privacy considerations, common mistakes, etc.
The goal is simple: when Google fans out into subtopics, your site has a page or at least a section ready to be cited.
Treat citations as a new top-of-the-funnel moment.
In conversational Search, the click often follows trust. Visibility now depends on being a source the model returns to repeatedly across the journey.
In practice that usually correlates with:
- Clear authorship and first-hand credibility signals
- Fast, scannable structure, with headings that map cleanly to real questions
- Distinctive data, examples, and frameworks that can’t be replaced by generic summaries.
Optimize for comparison and choice, not just awareness.
Google explicitly positions AI Mode as being helpful for comparisons and complex decisions. That’s not accidental. This is the cue to invest in content that helps users choose: “best for”, “alternatives”, “vs”, pricing breakdowns, templates, checklists and implementation playbooks. These pages are not only more likely to be cited, but they’re also the ones most likely to convert as overall traffic becomes scarcer.
Rebalance measurement beyond last-click SEO
If users spend more time inside Google’s AI layer, last-click attribution will increasingly undercount Search’s influence.
Google is turning Search into a dialogue, and follow-up questions are the gateway. That’s not “the end of SEO,” but it is the end of SEO that only targets the first query.
The teams that win will be the ones who design content for the entire question chain and who build enough brand gravity that, when Google summarises the web, it can’t tell the story without them.
Find out more about our AEO services.
