What it is
Generative AI in search refers to search engines using large language models to create synthesized answers instead of just returning a list of links. Google's AI Overviews (AIOs), Bing’s CoPilot Search, ChatGPT's search capabilities, Perplexity AI, and similar tools use algorithms trained on massive amounts of content to predict and generate responses based on user queries.
Unlike traditional search that matches keywords to documents, generative AI search produces direct answers, summarizes information from multiple sources, and can engage in conversational follow-ups.
Why it matters
Generative AI fundamentally shifts how users interact with search results and where your traffic comes from. Instead of scanning ten blue links, users get an AI-generated answer that may or may not send clicks to your site. This changes what "ranking" means and forces you to optimize for inclusion in AI-generated responses rather than traditional SERP positions.
The commercial implications are real: 1 in 10 U.S. internet users already use generative AI for online search, and adoption is generational. For example, if you target college students (like Chegg does), you're facing an audience that has rapidly shifted to ChatGPT for answers. Meanwhile, if your audience is older professionals in traditional industries, you might have years before meaningful AI search adoption hits your traffic.
The key insight: You need to ask your actual customers whether they use AI to search, not assume based on industry hype.
How to use this knowledge
Test how AI search engines answer queries in your space. Run your core keywords through ChatGPT, Perplexity, Google's AI Overviews, and Bing Chat to see what results appear and which sources get cited.
Use reverse-engineering to understand which sources AI pulls from and why.
Optimize for citation and attribution by making your content more structured, authoritative, and clearly sourced.
Track referral traffic from AI platforms (ChatGPT, Perplexity) in your analytics to measure actual impact rather than relying on assumptions.
Segment your strategy by audience: Invest heavily in AI search optimization if your users skew young or tech-forward, but maintain traditional SEO if your audience hasn't adopted AI search yet.
Growth Memo guidance
"The focus right now is on creative tasks but it's not far-fetched that AI will soon help with more commercial tasks. The key for SEO is still to be visible in AI Overviews and other AI results. But I expect a big load of top-of-the-funnel traffic to fall away in favor for better converting traffic."
"It's a lot harder for Google to fine-tune AI answers than classic search results. Before AI, Google was able to rearrange snippets and sometimes SERP features. Now, it needs to improve the angle, presentation, accuracy and diversity of answers, which is a lot more complicated."
"The adoption of AI is generational. Young people are at the frontlines. The question is not whether Chat GPT replaces Google, but when your target audience will adopt it."
AI Overviews — Google's public-facing generative AI search feature that replaced SGE beta testing
Search Generative Experience (SGE) — Google's original beta version of AI-powered search results with conversational capabilities
Hallucination rates — the percentage of AI-generated responses that contain false or fabricated information, ranging from 3% in GPT-4 to 69-88% in legal queries
Query fan-out — when AI search systems break a single user query into multiple sub-queries to gather comprehensive information
Multi-modal AI — generative AI that can understand and create content across different formats like text, images, and video