Leading researchers conducted a groundbreaking study on how AI-powered search engines like Google’s SGE, Bing Chat, and Perplexity.ai respond to user queries. Their key takeaway? Generative AI transforms search, but most creators still don’t know how to rank in AI-generated answers.
This is where Generative Engine Optimization (GEO) comes in.
As someone who’s tested AI search trends firsthand, I’ve seen how quickly the landscape is evolving. In this deep dive, I’ll break down the study’s findings, share real-world data, and reveal how you can adapt your SEO strategy for the AI era.
How the Study Worked (With Surprising Results)
The researchers followed a structured approach:
- Collected 1,200+ search queries across different topics (mostly informational).
- Built a custom AI search engine mimicking Bing Chat’s responses.
- Tested rankings on Perplexity.ai to compare results.
- Applied GEO techniques to select URLs.
- Tracked citation frequency for Did AI start referencing optimized pages more?’
Key Findings:
This aligns with what I’ve seen in my own tests—AI search engines prioritize credibility over keyword density.
Why Traditional SEO Isn’t Enough Anymore
Google’s Search Generative Experience (SGE) is already rolling out to millions. According to a BrightEdge study, AI-generated answers could impact 65% of all search queries by 2025.
But here’s the catch: Only 3% of marketers are actively optimizing for AI search. Most still focus on old-school SEO tactics that don’t work as well in generative results.
What Google’s SGE Patent Reveals
The study assumed AI search behaves predictably, but Google’s official SGE patent suggests otherwise. Their AI considers:
- User intent (Are they researching or shopping?)
- Past search behavior (Personalized answers)
- Related queries (Broader context)
This makes GEO far more complex than traditional SEO.
The Biggest Risks of AI-Powered Search
- Misinformation & AI Hallucinations
The study claims AI search is “rapidly replacing traditional engines.” But is it reliable? 42% of ChatGPT’s answers contain inaccuracies (Stanford Study). Google SGE sometimes ignores the query and delivers generic summaries - Manipulation by Low-Quality Content
If simple tricks (like stuffing expert quotes) boost rankings, we could see a flood of spammy AI-generated content, just like the early keyword-stuffing days. - Traffic Drops for Publishers
Forbes reported a 40% decline in organic traffic after AI answers rolled out in some niches.
Small blogs could lose up to 60% of their search traffic if they don’t adapt.
How to Future-Proof Your Content (GEO Strategies)
Based on the study + my tests, here’s what works:
- Cite Trusted Sources
Link to .gov, .edu, and industry-leading sites (AI loves authority). Example: Instead of saying “studies show…” link directly to NIH or Harvard research. - Use Expert Quotes & Interviews
AI rewards firsthand expertise. Case Study: A health blog saw a 33% increase in AI citations after adding doctor interviews. - Add Data & Stats
Use charts, percentages, and research-backed numbers. Example: “72% of marketers say AI impacts SEO” (with a source). - Optimize for “People Also Ask”
AI pulls heavily from PAA sections. Pro Tip: Answer common follow-up questions in your content.
For more advanced tactics, check out my Complete Guide to AI SEO.
Despite the hype, traditional search isn’t dead yet:
87% of users still prefer Google’s classic links for complex queries (HubSpot Data). AI answers often lack depth—users still click through for details.
Google is also fighting back:
EEAT (Experience, Expertise, Authority, Trustworthiness) now plays a bigger role.
Firsthand experience ranks higher than generic AI content.
Final Thoughts: Should You Focus on GEO Now? Yes—but don’t abandon traditional SEO.
AI search is growing fast, but it’s still evolving. The best strategy? Combine GEO tactics with EEAT principles to rank in both AI and classic results.
Want to Dive Deeper?
What’s your take? Have you seen AI impact your search traffic? Let’s discuss in the comments!