Google’s AI Mode — its most significant SERP experiment since the knowledge panel — fundamentally changes the click dynamics for a meaningful class of queries. For keyword research and content strategy, the implications are real, even if the affected query volume is smaller than the coverage suggests.
What AI Mode Actually Changes
AI Mode generates comprehensive answers to complex, multi-part queries that would previously have required clicking through multiple search results. A query like “what’s the best way to structure a content audit for a 500-page e-commerce site” now receives a detailed, synthesised response that incorporates multiple sources — without the user needing to visit any of them.
The effect on click-through rates varies significantly by query type:
Heavily affected: Research-intent queries, definition queries, “how does X work” queries, and simple factual lookups. These are the queries where Google’s AI response fully satisfies the intent without requiring additional information.
Lightly affected: Transactional queries, local queries, navigational queries, and queries with strong brand intent. Users searching for “buy running shoes” or “James Dagmanseo pricing” are not satisfied by an AI answer. They need to reach a destination.
Positively affected (potentially): Citation sources. If your page is cited in an AI Mode response, you receive a visibility that didn’t previously exist — even if the direct click volume is lower than an equivalent organic ranking would generate.
The Keyword Research Implication
The practical implication for keyword research is that intent classification has become more important than search volume for informational topics.
A keyword with 2,000 monthly searches that is fully satisfied by an AI Mode response may generate fewer clicks than a keyword with 500 monthly searches that requires a destination visit. Prioritising high-volume informational queries as the foundation of a content strategy is increasingly unreliable.
The adjustment: weight keyword targets more heavily toward transactional, commercial, and navigational intent where AI Mode doesn’t absorb clicks. For informational content, focus on queries where the answer is complex, opinionated, or experience-dependent — the queries where users aren’t satisfied by a synthesised AI response and want to read an expert’s perspective.
What This Means for Content That Already Ranks
If you have informational content ranking in positions one through three for queries that AI Mode now dominates, your organic traffic from those positions may have dropped without any change in ranking position. This is the “ranking but not receiving traffic” pattern that is increasingly common in Search Console data.
The response isn’t to abandon those rankings — they may still contribute to AI Mode citation and to ranking for related queries. But the traffic model for informational content needs to account for zero-click absorption in a way it didn’t 18 months ago.
The Positive Case for GEO
The opportunity inside this disruption is citation. AI Mode surfaces sources, and being cited as a source creates a new form of visibility — more prominent in some ways than a traditional organic listing, because it appears as the foundation of a direct answer rather than as one of ten blue links.
The sites that will be cited consistently are those that produce content with the characteristics AI systems extract well: specific data, clear expert positioning, structured formats, and demonstrably original information that can’t be synthesised from generic sources.
This is why GEO is not a replacement for SEO — it’s an extension of producing genuinely authoritative content. The sites that rank well organically for their topic area tend to be cited by AI systems in the same topic area. Building topical authority is the underlying work that supports both.
A Practical Adjustment to Keyword Research
Rather than abandoning informational keyword research, the practical adjustment is to layer an intent-saturation assessment on top of traditional search volume and difficulty analysis.
Before committing to an informational article, run the target query in AI Mode. If the response fully and accurately answers the question with no meaningful gaps, the click capture potential is low regardless of ranking position. If the response has gaps, is demonstrably incomplete, or would benefit from expert opinion the AI can’t provide — that’s a content opportunity that is likely to rank, be cited, and drive meaningful traffic.
The keyword research process hasn’t changed fundamentally. The evaluation criteria have just expanded.