Let's Talk SEO Testing with an AWS Senior PM

Bing Shares AI Insights; NerdWallet's EEAT; AI Search Grounding; Open Knowledge Format; Anthropic's AI Survey; Copilot Cowork; Meta and Pinterest New AI Tools; and Much More!

Sponsored by

FIRST …

I’ll be on the next IMG Roundtable on June 25th, 10:00am EDT with Kaitlin McMichael, Sr. Product Manager at Amazon Web Services 🤩 .

We’ll be talking about SEO Testing, which lines up with the Hackathon that they’re doing now and I’m a judge for.

I’m looking forward to it just for Kaitlin herself and hear everything she has to say!

AD

One AI. Every Tool Your Store Actually Needs.

Most e-commerce sellers are paying for 6 to 8 separate tools that don't talk to each other — and spending hundreds of dollars a month just to keep up. StoreClaw replaces your entire stack with one autonomous AI engine that monitors competitors, optimizes listings, automates marketing, and tracks real profit across Shopify, Amazon, and beyond.

It doesn't wait for you to ask. It runs 24/7 in the background, so you wake up to a full dashboard instead of a list of things you forgot to check.

Connect your store, and StoreClaw gets to work — no prompts, no complex setup, no six-app stack.

Free to start. No credit card required.

SEO + GEO

Bing Webmaster Tools added four AI Performance preview features: Intents, Topics, Citation Share, and Compare. Intents classifies grounding queries into categories such as informational, commercial, navigational, local, research, and creation. Topics groups related grounding queries into thematic clusters, Citation Share estimates the percentage of citations a site receives for a grounding query, and Compare overlays prior time periods against the active view.

My Take: This moves AI visibility measurement closer to how SEOs actually plan work: topics, intent, and share of citations instead of one query list. Use it to spot where Bing already sees authority, then map those cited pages against your topical coverage and internal links. GSC - your move…

In the latest Search Off the Record podcast, Martin Splitt and John Mueller say normal HTML remains the safest format for SEO and AI discovery because crawlers already parse HTML and use page structure to understand a site.

  • HTML is still the discovery layer: Mueller says normal HTML pages are critical because crawlers can see links in headers, footers, sidebars, categories, and related pages that raw Markdown may strip away.

  • Chunking is weaker than architecture: Google’s take is that isolated text blocks are not enough. Crawlers need the surrounding page structure, internal links, and site context to understand how the content connects.

  • Markdown does not solve crawling: Splitt notes that crawlers already had to solve HTML parsing at web scale, so converting a normal site to Markdown does not fix a problem search engines have.

  • Machine-only versions create drift: Mueller warns that separate Markdown, text, or LLM versions add maintenance risk because a broken machine-only page may fail silently.

  • llms.txt is not an SEO shortcut: Mueller says llms.txt was not designed to help search engines or LLM systems discover all of a site’s content. It may help once an agent already knows the site and needs instructions for what to do there.

  • Developer docs are the exception: Mueller says Markdown versions can make sense for APIs, code samples, and technical docs, especially when the same Markdown source generates the HTML version users see.

My Take: It’s Google, so take it all with a grain of salt. Do your own research and testing too. They speak only for Google and none of the other search engines. And even then, we’ve found reality is the opposite of what they say is true.

Ethan Crump shows how NerdWallet turns visible trust signals into search and AI visibility: named authors, credentialed reviewers, public rating methods, editorial independence, and clear correction policies. Foundation’s data with AirOps found only 11.4% of FinTech AI answer citations come from brand-owned sources, which leaves almost nine in ten citations flowing to third-party sites like NerdWallet and Bankrate. NerdWallet’s nearly 9,000 pages draw more than 7 million organic visits a month, and its AI visibility now shows roughly 79,700 ChatGPT appearances, 54,300 AI Mode appearances, and 52,500 Google AI Overview appearances.

Louise Linehan and Ahrefs analyzed 137,210 domains and found 28% publish an llms.txt file, but 97% of those files received zero requests in May 2026. Among the small share that were fetched, 96% of requests came from bots, only 19.5% came from named AI tools, and AI retrieval bots accounted for just 1.1% of total requests. The file looks more useful for agentic tools and developer workflows than for winning AI search citations.

Dan Petrovic shows how different AI systems ground the same query by running “best ai seo agency 2026” through Google Gemini, OpenAI, and Anthropic on the same day. Gemini received 7 pages and cited all 7, OpenAI received 39 pages but cited only 2, and Anthropic received 14 pages and cited 9.

Barry Schwartz reports on Google Search Central Live Milan, where shared slides and attendee notes covered query fan-out, sitewide quality, AI reporting, subscriptions, and AI Overview clicks. The notes say 15% of daily Google searches are net-new, complex queries can trigger fan-out into parallel sub-searches, and forcing paragraph “chunking” for AI is useless because organization should follow human readability. Google also pointed to subscription linking through Reader Revenue Manager, with internal case studies showing a 34% engagement lift, and said AI Overview clicks tend to spend more time on site without giving absolute numbers.

Frank Olivo reports that AI search visibility is splitting by engine, audience, and content type, which makes one-size-fits-all GEO work less useful. The Zero Click NY takeaways point to a shift: teams need to decide whether they are optimizing for ChatGPT, Claude, or another engine before they decide what content to build.

  • Citation overlap is tiny: ChatGPT and Claude share only about 8% of their citations, with ChatGPT leaning more on Reddit, Quora, and forums while Claude favors listicles and opinion content.

  • Claude is stronger in B2B than traffic charts show: Ramp’s AI Index says 34.4% of tracked businesses now pay for Anthropic, ahead of OpenAI at 32.3%, while roughly 85% of Anthropic revenue comes from enterprise and API usage.

  • ChatGPT ads are already part of the board: Ads trigger on about 20% of prompts even though only 14% of real user prompts show commercial intent, and about one in five placements appears against a direct competitor mention.

  • Claude has a visible optimization path: Profound testing found 79.2% of Claude citations came directly from Brave’s top 10 results for the equivalent search, but Claude only searches the web on about a third of prompts.

  • Prompt type changes the work: Claude searches on about 81% of recency-framed prompts, 67% of ranking prompts, 55% of location prompts, and 51% of comparison prompts, while definitional and how-to prompts often stay inside model memory.

  • Marketing engineering is becoming real work: Olivo notes Google, Figma, RBC, and Autodesk have moved toward marketing engineer roles, with Figma posting a $295,000 base salary for the role.

SEO + GEO Ripples

  • Bing Webmaster Tools May Support More Country Reporting: Barry Schwartz reports that Bing Webmaster Tools may expand country-level reporting inside the Search performance report. Microsoft’s Fabrice Canel said Bing supports all countries and regions worldwide, but the specific report currently surfaces only up to 11 countries and should be extended soon.

  • Google updated its site move documentation to tell site owners to submit Change of Address requests for all domain variants during migrations. Google says domain migrations work best when all variants are migrated properly, including subdomains plus www and non-www versions, even if some are not actively used. The variants also need to be verified in Search Console before the move.

AI

Google presents OKF v0.1, an open spec that turns internal context into a directory of Markdown files with YAML front matter. Each concept gets its own file, the file path becomes its identity, and normal Markdown links turn the bundle into a graph that agents and humans can read without a vendor SDK. Google is also publishing an enrichment agent, a static visualizer, and sample bundles for GA4 e-commerce, Stack Overflow, and Bitcoin public datasets.

My Take: Knowledge has to live in a hierarchy before an AI system can use it well. That’s why I’m a fan of topical maps - it’s not just a list of keywords. It is a structured tree of topics, entities, intent, pages, and internal links. OKF points to where AI search and agent workflows are going: the teams with clean topic hierarchy and portable knowledge files will give machines better context than teams with scattered docs.

OpenAI says it is investing $150 million in the OpenAI Partner Network to help consultancies, systems integrators, technology firms, and data partners build and deploy enterprise AI solutions. The program launches with Select, Advanced, and Elite tiers, planned specializations in areas like Codex, cybersecurity, and agents, plus a Forward Deployed Experts pilot for complex deployments. OpenAI also says it aims to train and enable 300,000 certified consultants by the end of 2026, while Paychex reports an OpenAI and Bain deployment cut wait time by 80% and human-reviewed effort time by 30%.

Anthropic reports results from a national survey of 51,993 Americans conducted in November and December 2025, creating a baseline for how the US public thinks about AI benefits, risks, and governance.

  • Health leads public hope: 48% ranked curing diseases like cancer or Alzheimer's among their top three hopes for AI, followed by helping people with disabilities at 36%.

  • Job loss is the top fear: 64% worry AI will displace jobs, and it was the leading fear in every state surveyed.

  • Hands-on use changes the threat model: 54% of people who use AI at work every day worry about job loss, compared with 70% of people who do not use AI at all.

  • Government oversight has broad support: 71% say government should be involved in AI development and regulation, including 79% of Democrats, 68% of Republicans, and 69% of Independents.

  • Trust in AI companies is thin: Only 15% trust AI companies to make decisions about how AI is developed and used, below every other institution Anthropic tested.

Google announces ARD, an open spec for publishing, finding, and verifying AI capabilities across the web. The model uses domain-hosted catalogs and federated registries so agents can find MCP servers, A2A agents, OpenAPI tools, and other resources, then verify the publisher before connecting through the tool’s native protocol. Google is tying ARD to Agent Registry inside Gemini Enterprise Agent Platform, with governance controls for namespaced URNs, egress policies, pinned tools, and trust manifests for compliance needs like HIPAA.

Microsoft announces global availability for Copilot Cowork, Microsoft’s agentic system for long-running, multi-tool tasks inside Microsoft 365. Microsoft says more than half of the Fortune 500 used Cowork during the three-month Frontier preview, and internal June 2026 tests found Copilot Cowork averaged 30 to 40% lower cost per prompt than Claude Cowork with the Microsoft 365 connector across 125 runs. GA adds usage-based billing, tenant and user spending controls, usage reporting, nine partner plugins, Edge browser use in Frontier, and Microsoft 365 security controls for prompts, responses, and artifacts.

AI Ripples

  • New OpenAI Academy courses for the next era of work: OpenAI introduces three free Academy courses: AI Foundations, Applied AI Foundations, and Agents and Workflows. The courses move teams from basic prompting and review habits into repeatable workflow plans, then into agent-assisted work with defined context, boundaries, outputs, and human review.

  • LinkedIn and Adobe Launch Global AI Skills Initiative for Marketing Professionals: LinkedIn and Adobe are launching AI Essentials for Marketers, a free short-form LinkedIn Learning program available for the first 12 months. The launch includes four role-based courses across digital marketing, content and creative, social and communications, and data and analytics, with courses available in 47 languages after LinkedIn found marketing job posts requiring AI literacy rose 113% year over year.

MARKETING

Anu Adegbola reports that Google is broadening its Limited ad serving policy on Search, giving itself more room to restrict impressions for advertisers it sees as unqualified or confusing to users. The rollout starts this month and will continue gradually through 2028, with user feedback, brand clarity, and advertiser identity becoming stronger signals for ad eligibility.

Google introduces Ask Ad Manager, a Gemini-built agent for Google Ad Manager that helps publishers troubleshoot line items, generate custom reports, and move to the right platform settings from a conversation. The beta launches this month, with more skills and developer tools planned later in 2026. Google says Yahoo is already connecting Ad Manager into custom agents for forecasting, line item creation, reporting, and campaign optimization.

Rob Glover compiles 2026 online advertising cost benchmarks across search, social, video, and display channels. WordStream reports:

  • The average small business spends more than $3,000 per month on Google Ads, with average Google Ads CPC at $5.42 and average Google Search Ads CPL at $66.69.

  • Facebook ads at $0.70 CPC and $27.66 CPL

  • Instagram ads at $0.40 to $2 CPC

  • LinkedIn ads at $4.50 to $12 CPC

  • YouTube ads at $4 to $10 CPM

  • Google Display Network ads at $3 to $10 CPM

SOCIAL MEDIA

Facebook says it is rolling out AI Mode, a Meta AI search experience that answers questions using public culture, opinions, and recommendations shared across Facebook, Groups, Reels, and other Meta apps instead of returning only links. The update also adds opt-in camera roll sharing suggestions, collage cutout templates, transition effects for video montages, and AI photo presets that can change clothing, hair, accessories, or a profile picture style.

Pinterest says it is adding Business Assistant, Pinterest MCP, a new Pinterest Performance+ creative model, and Ask Pinterest as AI changes how people search, shop, and plan. Business Assistant is in closed US beta inside Ads Manager and mobile, giving advertisers visual trend data, performance prompts, and optimization guidance. Pinterest MCP connects campaign, analytics, and keyword data into partner copilots, while the new Performance+ creative model increased click volume by 7.5% in testing against the previous single-variant model.

Reddit and WPP Media report that consumers use Reddit as a validation layer before acting, especially when they need to compare options, test claims, and hear from people with direct experience. The research says 63% of redditors feel more confident in decisions based on Reddit community interactions than on other social platforms. It also found 44% say ads feel most relevant when clearly disclosed as sponsored, 58% agree that seeing a brand respond directly makes them more likely to trust that brand, and 52% prefer educational or problem-solving content.

Social Ripples

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.

AD

For a limited time, we’re giving away a whole folder of resources for getting started with TikTok Shop, including an internal case study on how went from $0 to $90K/month in 90 days. 

Claim yours:

What Did You Think of This Week's Wave?

Login or Subscribe to participate in polls.

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

or to participate.