Google now shows two different answers on a single search results page: an AI Overview at the top and the traditional list of blue links below it. Most people assume the AI Overview summarizes those results beneath it. A new benchmark study from B2B SaaS agency DerivateX reveals that assumption is wrong. Across 100 software-buying queries and 1,259 AI Overview citations, only 35% of the sources cited in the AI answer also rank in Google's own top 10 organic results for the same query.

What's Actually Happening Inside Google's Two-Layer Search Results?

The research, titled "Two Googles, One Query," analyzed AI Overview citations against organic search results and product recommendations captured in June and July 2026. The findings reveal a stark divergence between what Google's AI layer recommends and what Google's traditional ranking algorithm surfaces.

The gap widens when you look at product recommendations, which is where buying decisions actually happen. Only 28% of the products named in Google's AI Overview also appear in the search results for the same query. In practice, the shortlist a buyer reads in the AI answer is mostly a different shortlist from the one Google ranks directly below it.

The study uncovered several structural differences between the two surfaces:

• Source Overlap: Only 35% of AI Overview citations also rank in Google's top 10, meaning 65% are exclusive to the AI answer.
• Product Recommendations: Only 28% of AI Overview-recommended products appear in the classic search results, with 72% named without ranking.
• YouTube Preference: The AI Overview cites YouTube in 51 of 100 queries, while Google's own search results rank YouTube in just 7 queries.
• Reddit Reversal: Reddit ranks in 82 of 100 search results but is cited by the AI Overview in only 43 queries.
• Category Divergence: Of 20 product categories tested, 19 diverge from Google's traditional rankings, with only QuickBooks hosting showing close alignment at 62% overlap.

When the two surfaces do agree, they agree at the top. Where a source appears in both the AI Overview and the search results, 72% of the time it ranks in Google's top 5, with a median position of number 4. Ranking well is not sufficient to be cited by the AI Overview, but among shared sources, it remains the only reliable way in.

Why Does This Matter for Businesses and Content Creators?

The implications are significant for anyone trying to reach buyers through search. The AI Overview is becoming the primary discovery layer for many users, especially those asking comparative or informational questions. If your content doesn't appear in that AI answer, you're invisible to a growing segment of searchers, even if you rank well in traditional results.

"Ranking #1 on Google still matters. It just no longer matters for the same reason it did in 2023. The AI Overview is a separate discovery layer with its own content preferences. Treating it as a byproduct of good SEO is the fastest way to disappear from the surface most buyers now see first," said Apoorv Sharma, co-founder of DerivateX.

Apoorv Sharma, Co-founder at DerivateX

This shift introduces a new optimization challenge: answer engine optimization, or AEO. Unlike traditional SEO, which focuses on ranking a page as a clickable link, AEO focuses on being extracted and cited as the answer itself, whether or not a click follows. The two disciplines share foundations but have different end goals and different success metrics.

How to Optimize for Both Search Layers

• Treat Video as a First-Class Surface: The AI Overview shows a strong preference for video content, citing YouTube far more often than traditional search results do. This represents the single largest structural gap between the two ranking systems, making video a priority for AI visibility.
• Earn Placement in Third-Party Listicles: The AI Overview reads and cites third-party comparison articles and roundups heavily, which account for 63.4% of its citations. Getting featured in industry listicles and review sites is now part of the optimization work.
• Measure AI Citations Separately from Rankings: Stop assuming that good SEO rankings automatically translate to AI citations. Track AI Overview mentions and citations as a distinct metric using prompt testing and citation tracking tools, separate from your rank reports.
• Structure Content for Extraction: Answer engines reward content that states the answer plainly in the first line or two, then elaborates. Burying the answer three paragraphs down is how you rank in traditional search but still get skipped by AI systems.
• Build Off-Site Reputation: Backlinks still help, but reviews, mentions, and third-party corroboration are what answer engines quote and trust. Your reputation off your own domain is now part of the optimization work.

The research points to a fundamental shift in how search discovery works. For the past 15 years, SEO has been about winning a position on a results page and earning the click. Now, success increasingly means being the answer the machine gives, cited and trusted, whether or not a click follows. The two strategies are not rivals. AEO is built on top of SEO, and the businesses that win in 2026 do both.

DerivateX is now offering what it calls GEO (generative engine optimization) agency services to help B2B SaaS companies navigate this shift. The agency helps clients get found, cited, and recommended inside ChatGPT, Perplexity, Gemini, Google AI Overviews, and Claude, alongside maintaining organic Google rankings. Client results include Gumlet, which attributes roughly 20% of inbound revenue to AI discovery, and REsimpli, which became the most recommended and cited real estate CRM on ChatGPT within 90 days.

The honest answer for most businesses is that you need both strategies, in proportion to your query mix. Traditional SEO keeps doing most of the work on transactional, brand, and navigational searches, the ones where a shopper already knows roughly what they want and is close to buying. AEO is the layer for the informational and comparative slice of the funnel, the "which is best for X" and "what should I buy" questions that increasingly get asked to an assistant instead of a search bar.