AI search visibility is no longer just a website problem. Google's platform properties connect social and video reporting with search performance. This Visibility Report covers platform properties, AI opt-out trade-offs, AI ads, product feeds, and AI citations. https://thevisibilityreport.com

ChatGPT ads now have conversion bidding, geo exclusions, and bulk controls. The paid side of AI search is starting to look like performance marketing. This week's Visibility Report covers ads, citations, crawler access, entity consistency, and reviews. https://thevisibilityreport.com

Google says AI search sends billions of clicks each week. Publishers still cannot verify the number or isolate the traffic. This week's Visibility Report covers AI clicks, AI Overviews, share of voice, tool notes, and Duane Forrester's GEO tools survey. https://thevisibilityreport.com

AI search makes the old higher ed marketing boxes less useful. SEO, content, reputation, and enrollment visibility have to work together. https://edtechconnect.castos.com/episodes/paula-french-from-clicks-to-conversions-a-practical-playbook-for-ai-search

Google Search Console now reports social and video content performance. That matters for AI visibility because the cited surface is bigger than your website. New Visibility Report: off-site visibility, AI-shaped B2B shortlists, RAG, AI shopping, and entity clarity. https://thevisibilityreport.com

For AI search, your website is not the only evidence source. Reviews, directories, outcomes, faculty expertise, and consistent facts all matter. https://edtechconnect.castos.com/episodes/paula-french-from-clicks-to-conversions-a-practical-playbook-for-ai-search

AI visibility is getting more technical. This week's Visibility Report covers Safari MCP for AI debugging, markdown pages for AI bots, what agents skip, and Lighthouse checks for llms.txt. If agents cannot parse the page, they may not cite it. https://thevisibilityreport.com

Your page can rank on Google and get skipped by AI Mode. Here's why: Google says it breaks a question into "a multitude" of sub-queries (outside analyses put it around 8-12), searches each one. Learn more: https://www.searchinfluence.com/blog/what-is-query-fan-out-in-ai-search/

AI search makes a simple higher ed content problem more obvious: program pages need to answer real student questions. Clear beats clever. Specific beats generic. https://edtechconnect.castos.com/episodes/paula-french-from-clicks-to-conversions-a-practical-playbook-for-ai-search

Freedom of speech. Freedom of worship. Freedom from want. Freedom from fear. Franklin D. Roosevelt, Four Freedoms Address, 1941

“Ubiquity equals probability” has been a useful shorthand. If AI tools are going to understand who you are and what you know, one perfect page is not enough. Repeated, attributable signals matter: site, profiles, video, social, examples, quotes, mentions.

AI citations may shape what people search next. Similarweb tied AI brand mentions to later direct visits and branded search. New Visibility Report: Similarweb, schema for AI search, citation decay, Google's spam update, and Reddit as source material. https://thevisibilityreport.com

A login is not onboarding. It is account provisioning. The useful AI work starts when a team chooses one real workflow, gives the model context, sets boundaries, and builds a review loop. The tool is not the hard part anymore.

I’m going to spend a few posts this week on AI onboarding. Main point: access is not adoption. A team can have the tools and still not have a workflow. The real work is context, boundaries, review, and using corrections to improve the process next time.

Traffic can go down before demand does. That is the uncomfortable AI search shift Paula French talked about on The Signal. Some student research now happens before the click. https://edtechconnect.castos.com/episodes/paula-french-from-clicks-to-conversions-a-practical-playbook-for-ai-search

"Every useless help is an obstacle to development." Maria Montessori

AI visibility is an off-site evidence problem too. Reddit belongs in GEO, AI citation tracking, sentiment analysis, and modern off-page SEO because buyers are already doing the talking there. https://www.searchinfluence.com/blog/reddit-ai-search-strategy-ai-citations/

AI visibility data is more useful when it leaves the dashboard. This week: MCP connectors, Slack workflows, prompt discovery, agent-ready websites, and Google/Microsoft backing agent discovery standards. https://thevisibilityreport.com

Google’s AI Mode report: keywords became task briefs. Top terms include Find, Identify, Explain, Summarize. Optimize for the job, not just the phrase. Windsock, not GPS. SMX class: https://searchengineland.com/smx/master-classes/topics/claude-code-for-seo

ChatGPT gets the conversation. Google has the habit. If most users meet AI inside the search results they already use, AI Overviews needs its own measurement plan. New Visibility Report: Peec AI's 500K-prompt study, MCP data stacks, and agent discovery. https://thevisibilityreport.com

My SMX Master Class is Thursday: Claude Code for SEO. If you read the SEL piece on Claude Code as an SEO command center, this is the deeper live class. June 25 + on-demand. Friend15 = 15% off. https://events.searchengineland.com/smx-master-classes-june-2026/register?ref=will

SEO work is context switching: GSC, GA4, Semrush, crawls, spreadsheets, AI visibility, briefs and QA. Claude Code can connect it. I’m teaching that at SMX, live June 25. Friend15 = 15% off. https://events.searchengineland.com/smx-master-classes-june-2026/register?ref=will

My Search Influence colleague Paula French was on The Signal talking AI search for higher ed. Traffic can drop before demand does. The research step moved upstream. https://edtechconnect.castos.com/episodes/paula-french-from-clicks-to-conversions-a-practical-playbook-for-ai-search

Search is moving from rankings to answers to actions. My Search Influence colleague Cory Agular explains MCP SEO, a visibility layer for the AI assistant era. SEO: where do we rank? GEO: are we mentioned? MCP SEO: are we the source? https://www.searchinfluence.com/blog/what-is-mcp-seo/

Most SEO teams don’t need another dashboard. They need a way to connect the data they already have. That’s what my SMX Master Class covers: Claude Code as an SEO command center. Live June 25. Friend15 = 15% off. https://events.searchengineland.com/smx-master-classes-june-2026/register?ref=will

"You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete." Buckminster Fuller

ICYMI: the durable AI SEO asset is not the model. It is the learning system: sources, prompts, rubrics, structured facts, human review notes, and citation tracking. Visibility #21: https://thevisibilityreport.com

I’m teaching an SMX Master Class next week: Claude Code for SEO. Live online June 25. We’ll build an SEO command center for GSC, GA4, Semrush, Ads, AI visibility, briefs + QA. Use Friend15 for 15% off: https://events.searchengineland.com/smx-master-classes-june-2026/register?ref=will

Ranking #1 is useful. It is not a guarantee that AI systems will cite you. AI search pulls from a messier source layer than most dashboards show. I pulled the SMX Advanced data into a short carousel here: https://willscott.me/2026/06/04/smx-advanced-2026-less-busywork-more-geo/

I don’t think AI means everyone becomes a manager. A lot of the work is still work: checking, editing, improving, and deciding whether the output is worth using. That’s less management and more QA.