- Digital Surfer
- Posts
- Google Says Trust Us. SEOs Say Show Us the Data.
Google Says Trust Us. SEOs Say Show Us the Data.
Wave 318; Google Advice Backlash; AI Giving Coupons; Knowledge Graph Update; Trust Through Real People; Pricier Click B2B Ads; and Much More!
FIRST …
Google impressions jumped 86x on a YMYL health site during the Nov. core update. We didn't build a single backlink.
We started rebuilding the whole site and content in Floyi on July 1st. The last pages went live Sept. 28th.
Every change happened on the site itself. Floyi built a new topical map, we rewrote or pruned every old page against it, and Floyi wrote the briefs and drafts. More details on LinkedIn post.

If you want to run the same play on your site or a client’s, Floyi's Black Friday deals start today.
50% off your first 2 months on Monthly plans - Promo code: 50BFCM2026
20% off your 1st year on Annual plans - Promo code: 20BFCM2026
Both codes expire Nov. 30th at 11:59pm US ET.
When inside Floyi, head over to Settings and the Plans & Upgrades tab. Then enter the Promo Code in the field and click Subscribe Now.

Most Black Friday and Cyber Monday deals start next week. I’ll send out the BF issue mid-next week with the best ones I see.

SEO + GEO
Danny Goodwin reports that SEOs are pushing back on Google’s claim that AI search expands discovery, citing lost traffic, its content advice and incomplete reporting.
Google’s Brendon Kraham cites an Ipsos survey in which AI-assisted shoppers visited 12 touchpoints, nearly three times as many as shoppers who didn’t use AI.
Three in four shoppers who used Google AI Overviews or AI Mode said the features helped them make faster, more confident decisions, according to Ipsos.
Daniel Foley Carter says AI Overviews use publisher content while hurting traffic, and Pew found traditional-result clicks on 8% of visits with an AI summary versus 15% without one.
Pew found cited sources received clicks on just 1% of visits with AI summaries.
Seer found organic click-through rates for informational queries with AI Overviews fell 61% between June 2024 and September 2025.
Mathias Flink questions whether the additional touchpoints happen on Google properties, a breakdown Google’s published survey summary does not provide.
Lily Ray questioned Google’s advice about generic listicles, while MosiRaz says comparison and top-10 posts still appear in AI answers.
Joe Youngblood says missing prompt-level and click data makes Google’s advice to measure business outcomes difficult to follow.
Recipe publisher Inspired Taste says Google presented inaccurate versions of its recipes under its brand name.
Barry Schwartz reports that Google AI Overviews and ChatGPT display coupon codes directly in answers. Metehan Yesilyurt spotted ChatGPT cards with copy buttons and says coupon sites could lose visits and affiliate commissions when shoppers can take a code without clicking through. Glenn Gabe saw cards in the free version of ChatGPT and plain codes in the paid version. Barry notes that Google ads have shown promotions for years and merchants already supply offers through structured data.
Dan Lauer reports that automated rank trackers are receiving plausible Google results pages where the usual ranking sites have been replaced. Nozzle found a tracked site was about 20 times more likely to disappear from the top 10 on an affected page. Vendor estimates vary from AccuRanker’s 1–2% in October to Nozzle’s 27.6% average and 68.6% peak, with different systems and definitions. Google has not confirmed whether the results are intentional. Rankings can swing wildly while traffic holds steady, so Dan recommends treating Search Console as the source of truth before changing strategy.
Barry Schwartz reports that OpenAI may be testing a card-based ChatGPT interface that previews what users can do. Radu Oncescu spotted the design on desktop web for logged-out users and called it “Nova.” The cards suggest activities such as reading news, planning a gathering, building a travel itinerary and finding a book.
Jason Barnard reports that person entities fell 10.1% during Google’s November Knowledge Graph refresh, while companies barely moved. Among person entities scoring 24, 18.6% disappeared, mostly blank duplicates, while every tracked person’s own entity at that score remained. Google defines resultScore as a measure of how well an entity matches a request and has published no safety threshold. Jason points to Danny Goodwin, whom Google confused with a baseball player for more than a decade. He gained his own Knowledge Panel while his score stayed at 24.
Anna Crowe reports that publisher citation losses coincided with ChatGPT model launches without an OpenAI update notice, though the API-based tracking cannot establish the cause.
Journalism’s share of ChatGPT citations fell from roughly 33% in January to 22% in September, according to Generative Pulse.
Citations to AP, TIME, TechCrunch, the Los Angeles Times and The Atlantic fell 50% to 90% within a week of the July 9 GPT-5.6 launch.
Google top-three keyword counts for those outlets ranged from a 17% gain to an 8% loss between June 15 and July 15.
Four of those outlets lost 17% to 22% of their Google top-three keywords around May’s core update, while ChatGPT citations held flat through June.
In a 102-question panel, government and NGO sources rose from 10% to 35% of ChatGPT citations, while brands’ own sites rose from 14% to 20%.
AP, TIME and The Atlantic lost citations despite their OpenAI licensing deals.
Only 200 to 250 of more than 20,000 sites cited across four models appeared in all four, according to Generative Pulse.
Repeat ChatGPT answers shared just 41% of cited domains, Generative Pulse found.
Lily Ray says investing in identifiable experts is more durable than chasing search shortcuts, noting that Google warned against low-value automated pages as far back as 2003. After a medical client was hit by the 2018 Medic update, her team rebuilt the site around doctors, reviewers and original insights, with recovery taking more than three years. In Lily’s analysis of 64 sites, core-update winners were 258% more likely than losers to use real experts, a correlation rather than proof of cause. A Semrush study of 325,000 prompts found 59% of LinkedIn content cited in ChatGPT Search and Google AI Mode came from individual members. She recommends identifying a company’s real experts, building their presence on its site and helping them earn a footprint beyond it, with several voices instead of one.
SEO + GEO Ripples
Bing is testing a Web Mode filter that shows web page results without other features, including Copilot answers. it looks like more of a test rather than a full rollout.
Google is rolling out a collapsible Search sidebar on desktop globally. It gives users a way to return to recent AI searches.
Adobe approached Semrush to expand into AI search, former Semrush president Eugene Levin says. Learn more about what happened, because he says Semrush wasn’t looking to sell when Adobe made its approach.
Microsoft Ads adds “Made with AI” labels for AI-generated creative. Advertisers can declare AI-generated images or video and choose a visible watermark where supported.
A new .seo domain extension may arrive following an application by XYZ. It isn’t available for public registration yet, and Google says new generic extensions generally receive no special ranking benefit.
Google explains how sites can allow Google-Agent through a CDN, firewall or bot-detection service. The agent acts on user requests, and the new guide covers identification and troubleshooting.
Google is testing a Search sign-in prompt that appeared after a first query for one user. It’s another attempt to block scrapers. Google’s Rajan Patel said he would investigate and did not believe it was intended to appear that early.
The FBI linked SEO poisoning attacks to ransomware payments exceeding $28 million. The agency identified more than 1,000 infected WordPress sites and 216 controlled domains as of October 2026.
Search Console’s API adds three more reporting options, including branded-query classification and multimodal web data. It also supports filtering multiple countries or devices in one request, with branded data limited to eligible properties.
AI
Rand Fishkin analyzes anonymized ChatGPT conversations from Verso’s opt-in US panel to examine what people ask the assistant.
The dataset contains 118,163 prompts across 41,961 conversations from 12,321 panel users.
Images and visual design accounted for 16.1% of the average user’s categorized activity.
Information and entertainment accounted for 14.0%, while everyday life and administration accounted for 12.2%.
Writing and communication represented 11.9%, and health and wellness represented 10.3%.
Shopping, brands and local services represented 5.7% of activity.
Technology support and coding represented 4.4%.
Brand, product or company recommendations appeared in 12.1% of responses.
OpenAI introduces GPT-6.2 Astra for complex professional work and GPT-6.2 Sol for everyday tasks. OpenAI says Astra follows professional templates more closely when creating slides, spreadsheets and documents, while both models can turn written briefs into interactive browser experiences. On OpenAI’s internal GDPval v2 evaluation, Astra improved 6.2% over GPT-6 Astra while using 22% fewer output tokens. Astra’s API pricing is $8 per million input tokens and $40 per million output tokens, while Sol retains GPT-6.1 Sol’s token prices at $2/m input and $10/m output.
Google released Gemini 4 Neon for complex coding, analytical reasoning and enterprise workflows. Google says the model can work across spreadsheets, PDFs, charts, audio and video within a one-million-token context window. API pricing is $3 per million input tokens and $15 per million output tokens.
Google is rolling out Gemini Tasks free in the US, Japan, Korea, Canada, Australia and India using Gemini 4 Hydrogen. Tasks is a rebrand of Gemini Spark, its 24/7 personal AI agent. The feature can carry out multi-step requests such as calling businesses, organizing information and helping with purchases. Google says users can set app permissions and that Gemini is designed to ask for confirmation or pre-approval before critical actions such as buying. Advanced features are starting in the US, and phone calling has additional geographic restrictions.
Microsoft AI introduces their new model, MAI Code 2, for tasks such as debugging and multi-file changes, alongside Code 2 Flash for routine fixes and rapid iterations. Microsoft says Code 2 averaged 98% of GPT-6.1 Sol’s performance at 40% of the cost across four coding benchmarks. The company notes that some benchmark cost comparisons use promotional pricing and that results vary by workload and configuration. Both models are available in GitHub Copilot, while Code 2 Flash is in limited-access preview in Microsoft Foundry.
AI Ripples
Gemini computer use is available for Mac. It works across local files, desktop apps and the web in an isolated background window, with user-selected app permissions and approval for key actions.
Meta’s Muse agent faces early safety concerns after Reuters reviewed internal reports of wrongly sent messages and tested access to Signal messages. Meta called the examples isolated incidents and said many apparent account mix-ups were hallucinations.
Anthropic has closed its initial public offering, with shares trading on Nasdaq under the symbol ANT. The company reports approximately $119.5 billion in combined gross proceeds from the offering and concurrent private placement, including shares sold by existing stockholders.
MARKETING
Matt G. Southern shares a Grey Matter study of 14 B2B advertisers that found the same non-brand Google Ads keywords cost an estimated 21% more year over year after statistical controls. The estimate’s range was 1% to 44%, reflecting the small client sample. The analysis tracked 1,349 keywords and found an association between higher campaign spend and higher costs per click. It did not report conversion rates or cost per lead, so the figures cannot establish whether campaign outcomes improved or worsened.
Reza Moaiandin shares a SALT issue when they found genuine contact-form enquiries recorded by its marketing platform but missing from its sales CRM. From August 1 to November 20, 69% of genuine enquiries reached the CRM automatically, 20% were added manually and 11% never arrived. In the first 20 days of November, the automation missed 73% of genuine enquiries. Reza says AI agents, autofill, password managers or another technical issue could be involved, but SALT has not established the cause.
CONTENT
Benu Aggarwal explains how teams can connect customer questions with verified business information to guide content decisions. She recommends combining permitted signals from search, sales and customer service, then building briefs around the customer’s goal, constraints and evidence. Her approach makes passages understandable on their own and assigns owners and review triggers to facts that can change. She also separates mentions, citations and recommendations when measuring AI visibility.
TOOLS AND RESOURCES
Shelley Walsh reviews W3Techs figures distinguishing share among sites with a known CMS from usage across all surveyed sites.
WordPress’s known-CMS share was 60.2% at the start of 2026 and declined to 57.9%. It appeared on 39.2% of all surveyed websites.
Shopify held 7.9% of the known-CMS market and appeared on 5.4% of all surveyed sites.
Wix held 6.2% of the known-CMS market and appeared on 4.2% of all surveyed sites.
Squarespace held 3.6% of the known-CMS market and appeared on 2.4% of all surveyed sites.
Joomla held 1.6% of the known-CMS market.
Webflow and Tilda each held 1.2% of the known-CMS market.
No CMS was detected on 32.3% of surveyed sites.
WHAT I'VE BUILT FOR YOU
Floyi - Build Topical Authority that wins in Google and AI Search. Don’t just plan your content strategy - make it unstoppable.
ReplyDeck - Be active on Reddit without living on Reddit. Schedule replies, stay present in the threads that matter, and get your time back.
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.
LIKE DIGITAL SURFER?
Find me and others in the Digital Surfer Discord community.
I’d also love to know what you think and if you have any ideas for the newsletter. Reply or email me at [email protected].
I’d also appreciate it if you shared it with fellow digital surfers.
{{rp_personalized_text}}
Have a great week taking your SEO and digital marketing to another level!
And don’t forget to drag the Digital Surfer emails to your Primary Inbox 🌊


Reply