Here’s a data point that should change how you think about rankings.
The share of AI Overview citations coming from top-10 organic results dropped from 76% to 38% between late 2024 and early 2026. That’s a massive shift. In just over a year, Google’s AI went from overwhelmingly citing the highest-ranking pages to pulling nearly two-thirds of its citations from pages outside the top 10.
Some citations now come from pages ranking at position 40 and beyond.
How Google’s AI finds its sources.
The reason is a process Google calls “query fan-out.” When a user asks a question, Google’s AI doesn’t just look at the top results for that exact query. It breaks the question into sub-queries and related searches, then pulls relevant information from the best sources across all of those secondary searches.
A page that ranks #35 for your main keyword might rank #3 for a specific sub-query that Google’s AI uses as part of its answer. That page gets cited — not because it ranks well for the primary term, but because it answers a specific piece of the puzzle better than anything else.
AI Mode takes this even further.
Google’s newer AI Mode feature (a more conversational, deeply AI-integrated search experience) shows even more diversity in its sources. According to ALM Corp research, only 13.7% of citations in AI Mode overlap with standard AI Overview citations. The two features pull from substantially different pools of content.
This means a single page could earn citations across multiple AI features, each drawing from different queries and ranking positions. Your page doesn’t need to dominate one keyword. It needs to be the best answer for specific questions.
This is an equalizer.
For years, SEO has been heavily weighted toward whoever could reach position 1-3 for a target keyword. Those positions still matter — they still get the most traditional clicks. But AI citations have introduced a parallel system where the quality and specificity of your content can earn you visibility that raw rank position alone wouldn’t provide.
If you’re a smaller site competing against big brands for head terms, this is your opening. You probably can’t outrank them for “best CRM software.” But you can create the definitive page on “CRM data migration for agencies under 50 employees” and get cited when Google’s AI needs that specific answer.
Engagement signals still matter here.
The shift toward broader citation sourcing doesn’t eliminate the importance of engagement signals. Google’s query fan-out process still evaluates which sources to cite based on quality and relevance indicators. Pages that demonstrate real user engagement — clicks, time on page, low return-to-SERP rates — signal to Google that the content actually satisfies the query.
The formula is straightforward: create highly specific, authoritative content on topics within your expertise, then make sure Google sees the engagement signals that confirm your content delivers.
You don’t need to rank #1 anymore. You need to be the best answer for the right questions.
SerpClix uses an army of over 400,000 real human clickers to boost your organic CTR. Get started with a free trial or log in to your dashboard to set up your next click order.
Please note: there are no guarantees in search engine optimization, ever. There are innumerable factors that can affect search engine rankings. All SEO efforts can involve an element of risk. SerpClix employs real human clickers, so we think our service is far less risky than trying to use automated or robotic click methods. But, like all SEO strategies, there is an element of risk because Google’s algorithm is unknown and subject to change at any time. For more information please see our FAQs.
