- Digital Surfer
- Posts
- Four levels in. Three levels out. 8.9% said hold my beer.
Four levels in. Three levels out. 8.9% said hold my beer.
Topical Map Depth Study; Discover Update Done; ARENA Framework; AI Search Prompts Inside GSC; Public Companies; How ChatGPT Uses SEO; and Much More!
FIRST …
The number that proved right 300+ topical maps.
We just published a study at Floyi. 1,256 brands.
Three stages of the mapping workflow:
Topical Research - explore the subject
Topical Maps - turn research into a SERP-validated hierarchy
Authority Maps - ongoing governance
Every stage tracked.
The headline finding: 48.5% of brands build to four levels during topical research. After SERP validation, 87.7% consolidate to three. Zero remain at four.
That's not ambition failing. That's the system working.
I built 300+ topical maps at TopicalMap.com before building Floyi. We always allowed up to four levels. Some branches of a subject went there. Most settled at three.
Not because we ran out of ideas, but because the SERP evidence said two nodes shared the same intent. Force them onto separate pages and they cannibalize each other. Merge them and the map gets stronger.
Then something interesting happens in execution. As teams get deeper into a subject, 8.9% earn their way back to four levels.
Not by forcing depth, but because genuine expertise reveals nuance the algorithm couldn't see. A branch that looked redundant at validation turns out to deserve its own page when you know the subject and audience well enough.
Four levels is where you explore.
Three levels is where most teams execute.
And that 8.9% who return to four?
They didn't ignore the data.
They added to it.

AD
How can AI power your income?
Ready to transform artificial intelligence from a buzzword into your personal revenue generator
HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.
Inside you'll discover:
A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential
Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background
Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve
Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.
SEO + GEO
Matt G. Southern reports that Google's February 2026 Discover core update finished rolling out on February 27 after roughly 22 days, about 8 days longer than Google's original two-week estimate. This was the first time Google publicly labeled a core update as a Discover core update. NewzDash data comparing pre and post-update windows found fewer unique domains in top placements in the US and California, increased regional personalization with local domains appearing more often in local feeds, and X.com posts from institutional accounts climbing from 3 to 13 items in the US top 100. An analysis of over 400 news publishers found Discover's share of Google-sourced traffic nearly doubled from 37% in 2023 to 68%, while traditional web search traffic dropped from 51% to 27%.
Marie Haynes analyzes four sites that recovered from the December 2025 core update and identifies what changed. A medical ecommerce site that had been suppressed since August 2024 recovered after improving customer service and shipping logistics, which repaired its trust reputation. An affiliate site recovered after purchasing products for original reviews, adding photos and videos, breaking up walls of text, and using Microsoft Clarity to identify user frustrations. An ecommerce site in a YMYL niche recovered after improving site navigation, checkout flow and UX. A city guide recovered after adding original photography and video, writing about actual experiences rather than paraphrasing, and adding decision-making help like "This place is best for..." recommendations.
My Take: The common thread across all four recoveries is not technical SEO. It is removing reasons for Google to distrust you. Trust issues from customer service problems, commodity content that AI could write, poor UX that frustrates users. Fix those and you might recover.
Yoyao Hsueh (that’s me!) introduces the ARENA Framework for diagnosing why brands disappear from AI answers. The five-step pipeline covers Access (eligibility), Retrieval (context selection), Extractability (quote-friendly passages), Name (brand attribution) and Authority (reinforcement over time). AI search optimization is a selection problem, not a checklist, and that most teams fail because they cannot pinpoint which step in the pipeline is broken. The piece includes a diagnostic table for rapid triage, a 30/60/90 implementation plan, and a breakdown of how optimization differs across AI Overviews, AI Mode and chat engines.
Chris Long explains how to mine Google Search Console for prompt-like queries using a regex filter that surfaces searches with 10 or more words. He notes that last November, ChatGPT queries were found leaking into Search Console reports, and that AI Mode data will be available in Search Console as well. The method involves filtering Search Console Performance data with the regex ^(?:\S+\s+){9,}\S+$ to find longer, conversational queries that read like prompts.
Barry Schwartz reports on a newly granted Google patent that describes a system for generating AI-created landing pages tailored to individual users. The patent, filed about a year ago, outlines a process where Google evaluates a landing page score and, if it exceeds a threshold, generates a custom AI page that looks like the organization's site but is dynamically built in real time.
Metehan Yesilyurt reverse-engineers Google Discover's client SDK telemetry and maps how content moves from ingestion to feed delivery. The key takeaway is that several hard gates run before any ranking model does, including collection-level blocks that can remove an entire publisher from eligibility. He outlines a 9-stage pipeline: ingestion, Open Graph parsing, classification, collection gate, interest matching, server-side pCTR ranking, feed assembly, delivery, and feedback.
In this Search Off the Record episode, Gary Illyes and Martin Splitt explained why resource hints like dns-prefetch, preload and preconnect do not affect Googlebot. Gary said Google's infrastructure does not need DNS prefetching because it can talk fast to all cascading DNS servers, and Google caches page resources separately rather than fetching them in real time. The discussion also covered why meta robots and canonical tags must appear in the head, and why HTML validity is not a ranking signal since validity is binary and a missing closing span tag makes HTML technically invalid without changing anything for users.
Barry Schwartz reports that Microsoft updated the Bing Webmaster Guidelines with a new design and revised introductory language that now explicitly includes Copilot and grounding API results. The updated intro states that the guidelines describe how Bing discovers, crawls, indexes, evaluates and surfaces content across Bing search experiences, Copilot and grounding API results. Following the guidelines now helps ensure URLs are eligible for indexing and ranking, grounding results and citations, and sustained visibility.
This Search Engine Land analysis of 13 months of data across a customer base found four key patterns in LLM referral traffic.
LLM referrals still account for less than 2% of total referral traffic on average, ranging from 0.15% to 1.5% across ChatGPT, Perplexity, Gemini and Claude.
But LLM traffic grew 80% on average comparing the first half of 2025 to the second half, with some companies seeing 300% increases.
Aggregate referral traffic grew 3x from January to December 2025.
LLM referrals convert at roughly 18%, higher than paid shopping, SEO and PPC, despite accounting for the lowest percentage of total traffic.
Glen Allsopp notes that companies are growing more optimistic about AI search based on recent earnings calls, even as search traffic declines have been devastating for some businesses. The shift reflects higher conversion rates from AI referrals offsetting some traffic loss, though opportunity varies significantly by business type. Many companies have diversified to social, increased ad budgets and newsletters, but others genuinely see AI Overviews and chat experiences as an opportunity rather than just a threat.
SEO + GEO Ripples
Google is preparing to test changes to search results in Europe that would show results from rival vertical search services by default alongside its own. They want to avoid a fine under the Digital Markets Act, which can reach up to 10% of global annual revenue. Google has accumulated $11.5 billion in EU antitrust fines since 2017. Previous smaller-scale tests with stripped-down results saw businesses report up to 30% drops in free direct booking clicks.
SerpAPI filed a motion to dismiss Google's lawsuit, arguing that retrieving publicly visible search results does not constitute unauthorized access or hacking. The company cites legal precedents where browsewrap terms were found unenforceable and calls out Google's hypocrisy for scraping the open web while restricting others from doing the same. The case could determine whether independent SERP access tools have legal grounding.
AI
Emma Roth reports that Anthropic claims DeepSeek, MiniMax and Moonshot misused Claude to improve their own models through "industrial-scale" distillation campaigns involving roughly 24,000 fraudulent accounts and more than 16 million exchanges with Claude. DeepSeek allegedly held over 150,000 exchanges targeting Claude's reasoning capabilities and used it to generate "censorship-safe alternatives to politically sensitive questions." OpenAI similarly accused DeepSeek of "ongoing efforts to free-ride on the capabilities developed by OpenAI and other U.S. frontier labs" in a letter to lawmakers.
Lavall Chichester breaks down how generative AI platforms use SEO for growth. OpenAI was offering $310,000 to $393,000 for a content strategist with SEO experience and separately hired for a growth role focused on SEO, CRO and web strategy, suggesting a $410,000 to $600,000 annual investment in SEO talent. ChatGPT has an authority score of 99, ranks for roughly 287,800 keywords, and gets 45.5 million monthly searches for its brand term. Perplexity ranks for 184,800 keywords with an authority score of 81, while Claude ranks for 36,000 keywords with an authority score of 75. The analysis applies a 3Cs framework (code, content, conversions) and finds ChatGPT wins on technical foundations like robots.txt and semantic URLs, while Perplexity and Claude have optimization gaps including unindexed pages and missing meta tags.
Google releases Nano Banana 2 (Gemini 3.1 Flash Image), which brings the advanced capabilities of Nano Banana Pro to the speed of Gemini Flash. The model supports subject consistency across up to five characters and 14 objects in a single workflow, precise text rendering and translation within images, and production-ready resolutions from 512px to 4K. It also pulls from Gemini's real-time web search for grounded visual generation of specific subjects.
AI Ripples
OpenAI announced $110 billion in new funding at a $730 billion pre-money valuation, including $30 billion from SoftBank, $30 billion from NVIDIA and $50 billion from Amazon. The company disclosed that ChatGPT now has more than 900 million weekly active users and more than 50 million consumer subscribers, with more than 9 million paying business users. Weekly Codex users have more than tripled since the start of the year to 1.6 million.
Microsoft released a joint statement confirming its partnership with OpenAI remains unchanged after the Amazon announcement. Microsoft retains its exclusive license to OpenAI IP and Azure remains the exclusive cloud provider for stateless OpenAI APIs. The revenue share arrangement continues unchanged.
ChatGPT's ad ecosystem is expanding, with ad intelligence firm Adthena spotting more brands including Best Buy, AT&T, Pottery Barn, Enterprise, Qualcomm and Expedia. Based on 1,500+ prompts analyzed, most ads appear on the first prompt but some only trigger on the third or fourth repetition of the same query. High-intent modifiers like "best" and "new" carry significant weight. Best Buy secured two ad placements in a single response for iPhone-related queries.
Anthropic clarified how its three Claude bots work: ClaudeBot crawls for AI training, Claude-User fetches pages when users ask Claude questions, and Claude-SearchBot indexes content for search optimization. Blocking one does not block the others. All three respect robots.txt directives, but IP blocking may not work reliably since the bots use public cloud provider IPs.
OpenAI defeated xAI's trade secrets lawsuit, with a US District Judge granting its motion to dismiss with leave to amend. The judge wrote that xAI "does not point to any misconduct by OpenAI" and that allegations about eight former xAI employees who left for OpenAI showed no indication that OpenAI directed their actions. The case is part of a broader legal conflict between Elon Musk and OpenAI, including a separate lawsuit over OpenAI's transition to a for-profit entity scheduled for jury trial in April.
MARKETING
Kenneth Hamner outlines how AI Overviews are reshaping paid search economics. Seer Interactive data shows paid CTR on queries with AI Overviews fell 68% from 19.7% to 6.34% between June 2024 and September 2025, with the steepest drop occurring in July 2025 when paid CTR collapsed from roughly 11% to 3% in a single month. Brands cited in AI Overviews see a 91% paid CTR lift, suggesting informational queries can still be strategic assets for cited brands. The authors recommend four pivots:
monitor informational intent performance rather than blanket-excluding those keywords
prioritize product feed quality since Google's Shopping Graph relies on structured data
craft creative that differentiates rather than informs since users already consumed AI answers
build first-party audience data since keyword targeting becomes less reliable.
CONTENT
Tomek Rudzki warns that companies rushing to publish raw AI content at scale are damaging their visibility in both Google and AI search engines. His analysis of AI content generation tools found that 36% to 75% of brands featured in their case studies had suffered significant Google traffic drops. When Grokipedia lost Google rankings in early 2025, ChatGPT, AI Overviews and AI Mode citations declined at the same time.
My Take: This confirms what I have seen across client sites. The brands getting hit are not the ones using AI to help them write better. They are the ones using AI to replace expertise entirely. If you cannot answer yes to whether you would read your own article or whether it adds something competitors cannot copy from an LLM, do not publish it.
SOCIAL MEDIA
Reddit announced it is testing an AI-powered shopping feature that surfaces product carousels in search results for a small group of US users. The carousels show products mentioned in community discussions, with pricing, images and direct buy links.
Social Ripples
Snapchat is testing Creator Subscriptions in the U.S. with select Snap Stars, giving fans access to exclusive content, priority replies and ad-free Stories for a monthly fee. Snapchat reported 946 million monthly active users in Q4 2025, with U.S. users posting to Spotlight growing 47% year-over-year.
TOOLS AND RESOURCES
Paul Andre de Vera launched SearchQualityRater.com, a free tool that evaluates whether Google would consider content "high quality" based on the 182-page Search Quality Rater Guidelines. Users paste a URL and receive a quality score matching Google's scale, a breakdown of E-E-A-T signals (Experience, Expertise, Authoritativeness, Trust), and prioritized improvement recommendations. The tool requires no account and was built to automate the same evaluation process Google's human reviewers use.
My Take: This is useful for a pre-publish quality check or content audit. Run your key pages through it to see where E-E-A-T signals are weak before Google's raters or algorithms make that determination for you.
BUSINESS IDEAS
Katlyn Edwards breaks down how to build a sustainable SEO freelance practice by treating it as a service business, not just a flexible job. She says that generalists compete on price while specialists compete on expertise and outcomes, and that productization (turning custom work into repeatable deliverables) is what separates freelancers who scale from those who burn out. The piece covers three pricing models (hourly, project-based and retainer), with project-based pricing recommended for productized work and retainers only advised after you have delivered the same project multiple times and can firmly hold scope.
Jeff Bullas breaks down the February 3, 2026 market selloff that erased $285 billion from software stocks after Anthropic announced Claude CoWork, a plugin system that lets AI agents execute core business workflows autonomously. He argues the event marks the collapse of the per-seat SaaS model and the start of a new opportunity for solo operators.
Claude CoWork launched with 11 plugins covering productivity, sales, customer support, product management, marketing, legal, finance, data, enterprise search, bio-research and plugin management.
The side hustle economy is projected to grow from $556 billion to $1.8 trillion by 2032, with 41.8 million solopreneurs in the US contributing over $1.3 trillion annually.
Businesses using AI report 25 to 55% productivity gains and $3.50 to $4.00 in return for every dollar spent on AI solutions.
84% of content creators already use AI-powered tools in their workflow, and the AI-in-creator-economy market is projected to reach $12.85 billion by 2029.
NEWS
LowEndBox reports that GoDaddy updated its Universal Terms of Service Agreement to reclassify all 21 million users as "business customers," removing protections for private, personal or household use. The change eliminates consumer rights like EU 14-day withdrawal periods, weakens protections against unfair contract terms, and shifts disputes to commercial arbitration with higher fees. The updated terms also waive jury trial rights and restrict evidence options.
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
How Marketers Are Scaling With AI in 2026
61% of marketers say this is the biggest marketing shift in decades.
Get the data and trends shaping growth in 2026 with this groundbreaking state of marketing report.
Inside you’ll discover:
Results from over 1,500 marketers centered around results, goals and priorities in the age of AI
Stand out content and growth trends in a world full of noise
How to scale with AI without losing humanity
Where to invest for the best return in 2026
Download your 2026 state of marketing report today.
What Did You Think of This Week's Wave? |
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.
You currently have 0 referrals, only 3 away from receiving LinkedIn Shout-out.
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