Insights
1 in 7 Top-Ranking Pages Vanished From Google
According to SE Ranking's analysis of Google's December 2025 core update, 15% of pages that were sitting in the top 10 completely disappeared from the top 100.
Their research analyzed 100,000 keywords across 20 niches, comparing results from November 10, 2025 to January 5, 2026.
We can't know exactly what Google changed, but we can make some educated guesses about what might have happened.
1. A content quality shift
It's possible Google started prioritizing freshness and depth differently. Pages with outdated information or thin content could be losing ground to more comprehensive, recently updated alternatives. There's also speculation that Google's gotten better at detecting poorly edited AI content (though that's just a theory).
2. E-E-A-T signals are a priority
Pages without clear author credentials, expert sources, or authoritative signals might have lost rankings, especially in YMYL (Your Money Your Life) niches. Google could be weighing expertise and authority more heavily, especially for sensitive topics.
3. Search intent changes are always in play
Google might have re-evaluated what users really want for certain queries. Your informational content may have dropped simply because Google now thinks users want to make a purchase instead. For time-sensitive topics, even accurate content might get penalized if it looks stale.
4. Your competitors got better
Competitors simply published significantly better, more in-depth, or more recent content that better serves user needs.
Note the changes across industries:
- E-commerce and retail saw the most volatility with over 23% of the top 3 URLs being new entries
- Healthcare was more stable with only ~8% of top 3 URLs replaced (3x less volatile than e-commerce)
This suggests Google might be applying different standards across niches—more aggressive with commercial content, more conservative where misinformation can cause real harm.
Of course, there are multiple reasons for websites to lose their rank, but if a page stops being the 'best answer,' analyzing what replaced it and how search intent may have shifted allows you to improve it accordingly.
Related: Amsive's analysis of the December 2025 Core Update winners and losers.
Query Length Strongly Correlates With AIO Share
Two pieces of data stood out in Serpstat’s 11-month study on 1 billion keywords and 35 million AI Overviews (AIOs).
In Part 1, I mentioned that Featured Snippets may be disappearing from Google.
Part 2
Query length strongly correlates with AIO share, supporting long-tail keyword research.
Queries with 14+ words have an AIO share of 80.56%, compared to only 24.27% for 1-word queries.
This indicates AIOs handle complex, specific queries that traditional search results struggle to satisfy. While previous studies confirm this pattern, it demonstrates how LLMs have changed search behavior.
Since the launch of ChatGPT, users have been “trained” to enter natural language or “voice-style” queries. Longer queries reflect conversational phrasing that AI systems are optimized to interpret and respond to directly.
In 2025, short queries (1-2 words) dropped from 42% to 31% of total searches, while 3-4 word queries became more dominant.
The introduction of AI Mode allowed for more complex, multi-modal, and conversational questions. By August 2025, reports indicated that AI Mode queries were trending towards 10-11+ words, indicating that users are increasingly using natural language and providing more context.
The key takeaway: Long-tail keywords typically have low search volumes, so target specific, descriptive phrases that reflect how your audience talks about a topic.
My advice is to stick with what has always worked: focus on intent, and structure your content well.
Related
Robby Stein, Google VP, says Google shows AI Overviews in Search largely based on whether users engage with them — and removes them when they don’t. This gem came from Stein's recent interview with CNN where he answered questions about Google's newest features.
The Overlooked Risks in AI Content Plans
It surprises me that marketing leaders would prioritize "AI content generation at scale."
They're confusing output with authority.
Chasing topical authority through AI at scale is walking a very thin line between strategic advantage and self-inflicted damage.
Google doesn't penalize content solely because it is AI-generated, but they will specifically target the mass production of unoriginal, low-effort AI content intended to manipulate search rankings (scaled content abuse).
That distinction matters more than most CMOs realize.
When speed becomes the goal, quality gets treated as a box to tick instead of the core product. Large volumes of AI-generated pages that simply rephrase what already exists, flatten nuance, or miss search intent send the wrong signal entirely. Instead of building authority, the site starts to look shallow and repetitive.
The compounding downside
Once a site is associated with scaled content abuse, recovery is slow and expensive. Rankings dip, and trust erodes at the domain level.
CMOs betting on AI velocity are assuming that future algorithm updates will be neutral or forgiving. History suggests the opposite. Google tends to let a behavior spread before they aggressively correct it.
The biggest irony is that this strategy will undermine the very goal they're chasing.
Authority isn't proven by coverage alone. It's proven by insight, experience, and usefulness. Yes, AI can support that work, but it can't replace editorial intent or human judgment.
These marketing leaders are painting a target on their sites if they believe scaling content with AI is the answer to better performance across Search and AI engines this year.
Note: I pulled the image above from a CMO Investment Report by Conductor. You can download it to see more investment and strategy priorities from enterprise-level organizations.
Related
A LinkedIn post from Joshua Squires at Amsive raises some questions about how closely Google looks at a website for AI disclosures:"It appears that the standard legal loophole that sites use to manage legal liability when using AI generated or partially AI generated content can cause Google quality raters to flag site content as Lowest Quality / Untrustworthy and Lowest EEAT."
Do you follow me on LinkedIn? I share regular tips and stories I don't have room for here. Come and join me.
