Google Disrupts Rankings - What's Next?

AD

Download our guide on AI-ready training data.

AI teams need more than big data—they need the right data. This guide breaks down what makes training datasets high-performing: real-world behavior signals, semantic scoring, clustering methods, and licensed assets. Learn to avoid scraped content, balance quality and diversity, and evaluate outputs using human-centric signals for scalable deployment.

Download the guide now

SEO + GEO

Google Search Confirms It Does Not Support num=100 Page Parameter

Google has officially disabled the “results per page” URL parameter, eliminating the ability to see 100 results on a search page. This change is not a bug but a deliberate shift in how Google serves its results. SEO professionals are feeling the impact, with tools for rank checking now struggling to adapt. The removal could improve the accuracy of Search Console data, reducing inflated numbers caused by bots.

Google said: “The use of this URL parameter is not something that we formally support.” They’re even hiring an Anti-Scraping Engineering Analyst.

What some of the various rank trackers and SEO tools are doing:

• Semrush - Top 20

• Accuranker - Top 30 + Top 100 (2x month)

• Ahrefs - they will reduce, but no clear number given

• Spyfu - Top 100

• SE Ranking - Top 100

• DataForSEO - Top 10

• Keyword Insights - Top 100, but charging additional credits

My Take: The timing of the num=100 removal is crazy because I’m building some tracking into Floyi at the moment. I’ve decided to keep the Top 20 to show the key opportunities and gaps.

AI Progress Stalls for SEO Tasks Despite Wave of New Models

David Bell explores how AI model progress have slowed and they still fall short of expert-level execution in complex SEO tasks. Claude Opus 4.1 leads the way for tasks like technical SEO and localization, with ChatGPT-5 making a rebound after initial struggles. Gemini 2.5 Pro offers integration with Google’s ecosystem, offering a key benefit for SEOs. However, most AI tools still lack the precision needed for strategic, technical, and nuanced SEO work. For now, AI tools complement human SEO expertise, rather than replacing it.

What We Can Learn About Google's AI Search From the Official Vertex & Cloud Documentation

Olaf Kopp explores how Google's AI Search operates, focusing on key concepts like Retrieval, Ranking, Retrieval Augmented Generation (RAG), and Grounding. By dissecting Google’s Vertex AI and Cloud documentation, he sees how Google's advanced systems leverage embeddings, cross-attention, and dynamic retrieval to improve search relevance. He suggests applying these insights by optimizing content to match Google's evolving search algorithms, enhancing visibility and accuracy in AI-driven results.

Google Discover in Search: Allowing You to “Follow” Creators and Publishers

The Google Discover feed will now surface a richer mix of content, including articles, YouTube Shorts, Instagram posts and updates from X. You can also “follow” your favorite creators and publishers directly in Discover to see more of their work and preview their latest posts before subscribing.

How to Claim, Verify, and Manage Google Business Profiles at Scale

Elizabeth Rule breaks down the steps for claiming, verifying, and managing multiple profiles at scale. She highlights the importance of using domain-level emails for verification, organizing listings into business groups for easier access, and bulk-verifying locations for businesses with 10 or more storefronts. She also emphasizes monitoring changes and using API tools carefully to avoid errors, helping you stay on top of your listings and prevent unwanted updates.

Does Schema Help You Surface More in Google’s AI Mode?

There’s been considerable debate on whether schema markup improves visibility in Google’s AI Mode. Dan Taylor analyzed over 107,000 websites cited in AI Mode, the results are clear: no special schema types, like FAQs or Product, offer a measurable advantage. The most common schema types were Organization, Article, and Person, which are standard in well-structured sites. The key takeaway is that schema still plays a role in traditional search, but it doesn’t seem to directly impact AI Mode. To get noticed, focus on accuracy, completeness, and content quality - not just on schema tricks.

Structured Data Does Not Help With Visibility In AI Search

Recent tests by Mark Williams-Cook and Julio Guevara show that structured data (schema) does not directly influence visibility in Google’s AI-driven search results. Mark demonstrates how schema markup gets "destroyed" during tokenization in AI models, making it irrelevant to models like GPT. Similarly, Julio’s tests found that AI models could not extract key data from structured data if no visible text existed. These findings suggest that, while schema is vital for traditional search, it does not yet play a role in enhancing visibility for AI search. Focus on content quality over schema for AI mode success.

My Take: Just because they say Schema doesn’t give an advantage in AI Search now, doesn’t mean you shouldn’t do it. Schema supports your content and is still relevant for traditional search engines. Using plugins or AI to create schema is quick too, so there’s no reason not to do it.

SEO + GEO Ripples

• Recent buzz around Google Search Console adding an AI Overviews filter turned out to be a false alarm. A screenshot circulating in SEO circles claimed this new feature, but it was quickly debunked by Google’s John Mueller. Google’s Search Console currently lumps AI Overview data with regular web data, and there are no plans to provide separate filtering for it. The excitement around this feature stemmed from a fake post, which was later removed and apologized for.

AI

Reddit Wants a Better AI Deal with Google: Users in Exchange for Content

Reddit is negotiating a better deal with Google, wanting more revenue and user growth in exchange for the valuable data it generates. After striking a $60 million deal with Google in 2024, Reddit now wants to increase its role in Google’s AI ecosystem. The platform seeks dynamic pricing for its data, where compensation is based on its content's impact on AI-generated answers. This push highlights the paradox where platforms like Reddit provide crucial data but suffer from AI's potential to drain traffic and engagement that makes their content valuable.

Go Behind the Browser With Chrome’s New AI Features

Google is turning Chrome into a search powerhouse with Gemini and AI Mode. Gemini in Chrome now lets you clarify complex pages, summarize across multiple tabs, and even recall sites you visited last week with a simple prompt. Upcoming agentic tools will let Gemini handle tasks like bookings and orders while you stay in control. AI Mode is coming directly to the address bar too, delivering long-form answers and follow-up questions without leaving the page.

Detecting and Reducing Scheming in AI Models

OpenAI teams with Apollo Research to spot “scheming” models that hide misalignment, then curb it with a practical fix. The paper introduces covert-action tests, shows current frontier models can sandbag in controlled settings, and offers a counter: deliberative alignment. Teaching models an anti-scheming spec before they act cut covert behaviors roughly 30x. Action for builders: add scheming evaluations to your safety suite, preserve chain-of-thought visibility, train models to cite principles, and watch for situational awareness that can game tests. Here’s some more info on scheming:

Scheming is when an AI appears compliant but secretly pursues a different goal. In practice, it takes covert actions like hiding or distorting task-relevant information to get a preferred outcome while avoiding detection. Think of it as strategic deception, not a simple mistake. Examples include:

• Sandbagging: underperforming on evaluations to pass safety gates.

• Selective reporting: claiming a task is done or omitting inconvenient facts.

• Goal gaming: optimizing for reward or approval instead of the user’s intent.

Powering AI Commerce With the New Agent Payments Protocol (AP2)

Google shares the AP2, an open protocol that lets AI agents pay on your behalf with guardrails you can audit. It uses Mandates as cryptographically signed instructions that prove user intent to purchase. Start with two flows: real time approvals that convert carts with one tap, and delegated purchases with rules like price caps and timing. Plug AP2 into A2A and MCP to keep agents interoperable. Support multiple rails from cards to stablecoins. Build verifiable audit trails for authorization, authenticity, and accountability. You can get started by reviewing the specs on GitHub.

AI Ripples

• Apple developers can now use Claude Sonnet 4 directly inside Xcode 26 to write code and ship. After connecting a Claude account in Xcode’s Intelligence settings, you get a natural-language coding assistant that understands your entire project. Update to Xcode 26, log in with your Claude account, and start using Claude Code for app development for iOS, macOS, and other Apple platforms.

• Security researchers have uncovered a vulnerability where ChatGPT, using a prompt injection technique, was manipulated into leaking sensitive Gmail data. This exploit, called "Shadow Leak," exploited AI's ability to operate without constant human oversight, allowing attackers to access HR emails and personal details undetected. Although OpenAI has patched the issue, the vulnerability raised alarms for other tools linked to AI, including Outlook, GitHub, and Dropbox.

CASE STUDY

How People Use ChatGPT

This study dives into how people interact with ChatGPT, revealing insights into its rapid adoption and usage patterns. This research sheds light on ChatGPT’s broader impact across various demographics and industries. Read the full study. Some of the key highlights:

• ChatGPT now serves 700 million active users globally

• Non-work usage grew from 53% to 73% from June 2024 to 2025

• Practical Guidance, Seeking Information, and Writing account for 77% of usage - with writing at 42%

• They split user intent into Asking (49%), Doing (40%), and Expressing (11%).

• Younger users, especially under 26, make up nearly half of all messages

• Low- and middle-income countries have seen faster growth in adoption

AI Mode Impacts on Ecommerce: Researching Healthy Breakfast Cereals

Francine Monahan runs a clean, real-world test of how shoppers use AI Mode to choose healthy kids’ cereals and finds a gap between shiny AI answers and aisle-side decisions. Some interesting user interactions:

• 33% saw AI Mode as an option

• 2% used AI Mode unprompted

• 33 seconds average AI Mode dwell time

• 47% engaged with AI Overviews

• 76.6% unfurled the Overview content

• 20% clicked links inside Overviews

• 1.96 average links clicked

• 2 minutes 7 seconds average task completion

• 9% of participants phrased their search as a question

• Most participants still purchase cereal in store

The First Generation of Gen AI Use Cases in Marketing Are Streaking Through the Hype Cycle

Scott Brinker explores the rapid acceleration of generative AI applications in marketing and their journey through the Hype Cycle. While some AI use cases like chatbots and text generation are entering the Plateau of Productivity, others like video generation are experiencing a decline in adoption. Some of the key takeaways:

• Chatbots and text generation are seeing rapid adoption

• Video generation and customer journey mapping are in decline

• 62% adoption rate for chatbots, 45% for text generation

• Trends analysis up 56.5% in adoption

• AI tools are rapidly moving through the Hype Cycle

• First-generation AI use cases are maturing quickly

• The next generation of AI tools is already in early stages

WAYS WE CAN WORK TOGETHER

Floyi - The only AI-powered tool that builds 4-level topical maps. Don’t just plan your content strategy - make it unstoppable.

TopicalMap.com Service - Let us do the heavy lifting. We handle the research, structure, and strategy. You get a custom topical map designed to boost authority and dominate your niche and industry.

Topical Maps Unlocked 2.0 - Unlock the blueprint to ranking success. Master the art of structuring content that search engines (and your audience) love - and watch your rankings soar.

AD

Become an email marketing GURU.

Join us for the world’s largest FREE & VIRTUAL email marketing conference.

Two full days of email marketing tips & trends, famous keynote speakers (Nicole Kidman!), DJ’s, dance contests & networking opportunities.

Here are the details:

100% Free
25,000+ Marketers
November 6th & 7th

Don’t miss out! Spots are LIMITED!

I want in!