The GEO Show

Welcome to episode 28 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, sits down with Simon for a long-form conversation. They cover GrokBot agent OS staffing, social citation coverage, Neil Patel's AI+editorial experiment, Gary Vee's volume thesis, the GEO Show daily machine, generational AI pushback, Astra video editing, agency leverage, the future of ChatGPT ads as sponsored conversations, credit addiction, 2026 product bets on GrokBot and Meta Muse, podcast guest outreach ROI, and SynthID / ElevenLabs voice clones.

In this episode:

• 🤖 GrokBot agent OS staffing
  Simon has about seven GrokBot staffers. Cloud VMs keep working with the laptop closed. Paris calls GrokBot the AI product of the year for mid-2026.
• 📅 Social calendar agent win
  One employee-style prompt reorganized about two months of Stitch AI social content: overlaps, gaps, and channel differentiation in minutes.
• 📣 Citation coverage of social
  YouTube leads AI Overviews and AI Mode citations. LinkedIn articles work as stacked indexable URLs. Instagram, Facebook, TikTok, and X rarely appear. First-hour YouTube engagement may feed citation likelihood.
• 📝 Neil Patel: AI+human editorial ~400% uplift
  Pure AI slop underperformed AI content with human editorial. Do not hide SynthID. Ground drafts in proprietary knowledge, keep light editorial, and let engagement validate.
• 📣 Gary Vee 395 posts/day
  Volume across ~20 networks via team, camera, and AI remix. Engagement demotion is the risk if quality slips. LinkedIn reach is still a fraction of connection count.
• ⚙️ GEO Show daily machine
  AI brief → 15–20 min record → Descript → Sandy/GrokBot kit → YouTube, Transistor, Webflow, LinkedIn by lunch.
• 🖨️ Generational AI pushback
  About half of one early-20s cohort is against AI. Printing-press analogy. Knowledge-work career prospects without AI skills are effectively nil.
• 🎬 Astra + DaVinci + Higgsfield
  ChatGPT-6 Astra with computer use edited a talking-head into a cut timeline with AI animations. Videography for this content class is being solved.
• 🏢 Agency leverage / headless ops
  Posting workflows collapse to “here’s the link, go.” Bots can run agendas. Relationship time stays human.
• 💬 THROUGHLINE: ChatGPT ads as sponsored conversations
  Not banners or keyword Google Ads. Intent detection, soft sales transition, in-chat disclosure, payment authority close. Commission (~Amazon 16–18%) replaces CPM/CPC. Same rate across advertisers. Merchant feeds in; websites optional. 1–2 years out, maybe sooner.
• 💳 Credit addiction patterns
  Session shrink + gap expand after addiction. GrokBot generous limits + Grok 4.6 frontier quality as the counter.
• 🔮 2026 bet: GrokBot + Meta Muse
  Paris: top AI products of 2026 will not come from OpenAI, Anthropic, or Google. Work chief of staff + personal life assistant.
• 🎙️ Podcast guest outreach ROI
  GrokBot pitched Decoded to 20–25 shows. One asked $15k/episode. Trust and long shelf life can beat Google Ads MQLs for the right show.
• 🔊 SynthID / ElevenLabs / faceless podcasts
  Voice clones, AI-arguing hosts, NotebookLM stock voices vs custom ElevenLabs. Open question on voice watermarking.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the platform changes and citation mechanics that actually move AI visibility.

💬 Question: If ChatGPT closes the sale inside a sponsored conversation, which part of your funnel still needs a website, and which part becomes a merchant feed?

Welcome to episode 27 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, runs ten stories. He covers AirOps native CMS and project-management integrations, OpenAI licensed financial datasets inside ChatGPT, the ChatGPT Work data agent across major warehouses, reported ChatGPT ad refusals for competing image and audio products, Similarweb ecommerce 2026 AI referral findings, August LLM web-share shifts with Claude at 9.3%, Anthropic crawl-to-refer ratio spreads of about 31x, Chattopic persona and journey simulation, Hexagon's 50K shopping-recommendation study, and Good Firms' directory of 1,178 GEO agencies.

In this episode:

🔌 AirOps: publish without opening your CMS
AirOps launched native integrations with seven CMSs (Webflow, WordPress, Contentful, Sanity, Contentstack, Ghost, Strapi) and four PM tools (Asana, ClickUp, Monday.com, Airtable). The pitch is insight to published content without manual handoffs. Like Profound, AirOps is pushing a marketing system-of-record play that abstracts CMS and project management.

📚 OpenAI answers move onto licensed data
ChatGPT for financial services now includes premium datasets from Daloopa, PitchBook, and LSEG News. OpenAI says it indexes and hosts partner data itself with granular citations. Commercially important evidence may live in licensed datasets GEO cannot influence through conventional public-web optimization.

💬 ChatGPT Work data agent: dashboards become conversation
OpenAI's data agent connects BigQuery, Databricks, Snowflake, Redshift, and ClickHouse. Users ask why a metric moved and get investigation in natural language. Standalone analytics UIs get less defensible as work shifts into the LLM via MCP connectors.

🚫 OpenAI ads vs competing gen products
PYMNTS, citing The Information (Sep 10), reported OpenAI told some partners it would no longer accept ChatGPT ads for image and audio generation products that compete with OpenAI. Adobe Firefly was reportedly affected. Separate paid AI visibility from organic GEO.

🛒 Similarweb ecommerce 2026: AI referrals +200% YoY
Direct referrals from AI platforms grew more than 200% year over year but remain small in volume. An AI recommendation gave up to a 2:1 purchase advantage in some cases. 89% of AI shoppers also use conventional search, often ending in branded bottom-funnel queries.

📈 Claude is no longer a rounding error
Similarweb August web traffic: ChatGPT 55.5% (was 73%), Gemini 25.6% (was 13%), Claude 9.3% (was 1.9%), DeepSeek 3.4%, Grok 2.4%, Copilot 1.6%, Perplexity 0.9%. Add Claude to AI visibility tracking.

📉 Crawl-to-refer ratios ~31x apart
Search Engine Journal documented Anthropic crawl-to-refer ratios via Cloudflare from 71,000:1 down to 2,200:1. Crawl volume rises while human referrals drop. Measurement windows and missing referrers explain some spread, not the whole trend.

🧭 Chattopic: persona-first visibility
Chattopic positions AI visibility as starting with the user, not a static prompt: buyer simulation, persona-level visibility, persona chat, and AI focus groups. Gumshoe-style workflow: build personas first, then align prompts, for more runs and better statistical relevance.

📊 Hexagon 50K shopping recommendations
Hexagon analyzed 50,000 shopping recommendations across ChatGPT, Perplexity, and Claude. Authoritative content, third-party validation, and structured data accessibility allegedly explain 73% of citations. DA60+ roundups and structured data correlated with recommendations, with debate versus earlier episodes.

🏢 Good Firms: 1,178 GEO agencies
Good Firms (Sep 10) lists 1,178 GEO agencies across 80 countries, median $37/hr, 9% with 50+ people. Directory classification, not a pure-play census, but the SEO-to-GEO rebrand wave is real and early-mover advantage is largely gone.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the platform changes and citation mechanics that actually move AI visibility.

💬 Question: If OpenAI answers increasingly cite licensed partner datasets, which evidence sources in your category can you still influence, and which sit outside conventional GEO?

Welcome to episode 26 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, runs ten stories. He covers Writesonic data showing ChatGPT-6 Astra searching less and citing brand-owned pages at a 71.1% first-party rate, a brief Google AI Mode citation bug after a Gemini Flash swap, exact and phrase match search ads entering AI Mode on direct intent, Brand Ghost's 12.6% cross-engine source overlap, Highwire's Acro Answers bringing AI visibility into PR, HubSpot extending MCP so GEO workflows run inside ChatGPT and Claude, Matter reframing reviews as AI discovery, Soupa Intent's ChatGPT ads library, Jessica Bowman on AI visibility as an operations problem, and Meta buying Stilla.ai for agentic commerce in messaging apps.

In this episode:

• 📊 GPT-6 Astra favors brand-owned pages (Writesonic)
  Writesonic tested fifty prompts across sixteen categories. Astra averaged 6.1 fan-outs versus 10.3 for ChatGPT-5.6 Sole, read 50 results versus 82, directed 83% of search activity at brand-owned pages, and hit a 71.1% first-party citation rate. Shallower fan-out plus branded preference can reset GEO share of voice when models update.
• 🐛 Google AI Mode briefly lost citations
  After Gemini 3.8 Flash entered AI Mode on September 2, PinMeTo and others saw broad queries return fewer citations or none. Google Search VP Robbie Stein said it was not working as intended; source links returned after a fix. Sudden citation drops can be platform bugs, not content failure.
• 📢 Search ads enter Google AI Mode on direct intent
  Ginny Marvin confirmed a small experiment: exact and phrase match keywords can serve text ads in AI Mode when intent is explicit and direct. Complex conversational intent stays with AI Max and Performance Max. Paid and organic now compete on the same AI surface, so split those metrics.
• 🔀 Brand Ghost: still no single AI search reality
  The September 8 observatory run found 12.6% cross-engine source overlap, 44.7% of comparable sets with zero brand overlap, and 40% of dated citations older than a year across ChatGPT, Claude, Gemini, and Perplexity. Blended scores hide engine-level behavior.
• 📣 Highwire Acro Answers pulls GEO into PR
  Highwire launched Acro Answers to track ChatGPT, Gemini, Claude, and Perplexity on visibility, prevalence, perception, and topic adjacency, claiming 267-plus prompts and 30,000 AI responses per brand per month mapped to earned, owned, social, and paid actions. GEO is taking share from brand and corporate communications budgets, not only SEO.
• 🔌 HubSpot: GEO workflows inside ChatGPT and Claude
  HubSpot's September 9 docs update extends its MCP connector so users can retrieve visibility and citations, manage tracked prompts, review recommendations, and trigger content generation from ChatGPT or Claude. The dashboard is becoming optional; AI search intelligence needs to be callable by agents.
• ⭐ Matter reframes reviews as AI visibility infrastructure
  Matter's Review Visibility Management connects review capture, reputation monitoring, local profiles, search visibility, and AI discovery. Vendor causal claims aside, reputation is the next phase after visibility: how engines describe you, not only whether they mention you.
• 📚 Soupa Intent ChatGPT ads library
  Soupa Intent lists 3,151 unique creatives and 7,984 appearances, with Ad Radar tracking advertiser, message, landing page, prompt, and position across ChatGPT and Google AI Mode. An appearance is not an impression, and there is no spend or conversion data. Paid AI competitive intel is becoming its own GEO-adjacent category.
• 🏢 Jessica Bowman: AI visibility is an ops problem
  Search Engine Land quotes Bowman: AI can pull customer complaints, release notes, earnings calls, case studies, and employee reviews. Marketing hits a ceiling when the narrative problem is operational. Route remediation by source class to the right department.
• 🤖 Meta buys Stilla.ai for agentic commerce in chat
  Stilla is joining Meta to strengthen Business Agent commerce in WhatsApp, Messenger, and Instagram, with more than one million businesses already using it. The next visibility surface may be an agent that chooses and transacts, not an answer page with citations.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the platform changes and citation mechanics that actually move AI visibility.

💬 Question: If HubSpot can run GEO workflows inside ChatGPT and Claude, which task would you call from chat first, and which would you still keep behind a dashboard?

Welcome to episode 25 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, runs ten stories with slides. He covers a GEOforge sixteen-week graded-answer case study where a blended GEO score hid three contradictory engine outcomes, Yext finding four AI engines agree on the number one recommendation only 4% of the time, Semrush manufacturing data that separates mention, citation, and traffic, fashion treating AI visibility as a PR metric, Claude's B2B-skewed Similarweb audience, Similarweb data inside Perplexity Computer, AirOps repositioning AI search as bigger than SEO, Cited Index's 114-tool catalog with 97 monitoring-only products, Surfer MCP, and Surva.ai's $39 monitoring-plus-content bundle.

In this episode:

📊 GEOforge: one blended GEO score, three contradictory engines

A sixteen-week graded-answer study ran thirty buyer prompts forty times each against ChatGPT, Google AI Mode, and Google AI Overviews. Share of voice moved differently on each surface while the blended score rose. Engine, topic, and recommendation content change the verdict, so divergence belongs in product and category positioning.

🎯 Yext: four engines agree on #1 only 4% of the time

Yext's research brief finds four AI engines agree on the number one recommendation only four percent of the time. OpenAI leaned on directories; Anthropic used review platforms more often. There is no single AI search ranking reality, which is why engine consensus and divergence scores matter.

🏭 Semrush: mention, citation, and traffic are different layers

AI Overviews rose from 38% to 57% of tracked manufacturing search volume from January to July, while AI Mode and AI assistance still drove only a small share of site sessions. Visibility starts at mention and citation, moves through recommendation and visit drop-off, and ends at revenue.

👗 Fashion turns AI visibility into a PR metric

Launchmetrics tracks ChatGPT, Claude, Gemini, and Perplexity across ten markets and event windows, covered in Vogue. GEO is escaping the SEO department into PR, communications, and brand budgets because what AI says often comes from earned media.

📈 Claude's audience skews B2B (Similarweb)

Claude's worldwide mobile audience grew roughly 14x versus about 31% for ChatGPT, with users more likely to visit business-service and programming sites. Absolute ChatGPT size understates Claude for SaaS, cybersecurity, and developer tools.

🔌 Similarweb data live inside Perplexity Computer

Proprietary traffic, audience, and market-share data is now callable through Perplexity Computer. AI answers increasingly draw from connected datasets, not only public web retrieval, which further fragments where brands must be represented accurately.

📣 AirOps: AI Search is Bigger than SEO

AirOps' September 9 webinar positions AI search as a core buying channel where content visibility alone no longer explains discovery, citations, or buying decisions. The GTM move is upward: from content optimization toward AI decision influence.

🧰 Cited Index: 114 tools, 97 monitoring-only

Cited Index catalogs 114 AI visibility and GEO products; 97 are monitoring-only. Read-only dashboards without statistical rigor or an execution layer leave the hard work (citation building) to a small minority of the category.

🧩 Surfer MCP: SEO workspace inside ChatGPT and Claude

Surfer MCP lets users access and act on Surfer data from MCP-capable assistants. Large SaaS leaders that expose MCP will keep a seat; mid-tail tools risk being replaced by capabilities inside the LLMs.

🆕 Surva.ai: $39 monitoring, content, and publishing

Surva bundles thirty prompts across five platforms, content generation, citation analysis, and CMS publishing paths from $39/month, with docs updated September 8. New entrants are shipping full-funnel feature sets, not visibility alone.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the platform changes and citation mechanics that actually move AI visibility.

💬 Question: If four AI engines agree on number one only 4% of the time, which engine are you still treating as the "real" ranking, and why?

Welcome to episode 24 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, debates Simon Young of Question Marketing on whether the SEO industry is dead at last. They cover hopium retainers versus cited and trusted reference sites, Google's goto redirect layer, ChatGPT query fan-out, OpenAI's Labrador index, why keyword research and DA links stop mattering for GEO, share of voice vanity under Astra, OpenAI ads and Gemini voice, and why a 2026 Google ranking-factors survey already feels like the wrong ritual.

In this episode:

🎬 Question Marketing rebrand and 300 YouTube receipts
Simon re-released about 300 previously unlisted videos under Question Marketing, including one with 750,000 views, and is pulling transcripts into a knowledge base as proof he has talked about answer engines for years. question.marketing (July launch) saw Google crawl 150-plus pages and index fewer than ten.

🎙️ Episode pages, daily audio, and ChatGPT Visualize
Paris publishes full transcripts on episode pages for AI crawlability. Simon's clients record daily audio commentary from news feeds. Paris used ChatGPT Visualize to turn stories into a 10-slide deck after viewers asked for more than a talking head.

⚰️ Is SEO dead at last? Hopium to cited and trusted
After a decade of death claims, Simon argues the industry needed a professional body and sold hopium retainers instead. The shift is toward cited, trusted reference sites AI uses. Fundamentals remain; the no-click era may retire the phrase "search engine" itself.

↪️ Google goto redirects and ChatGPT query fan-out
Google is inserting redirects between SERP clicks and destinations, breaking pixels and conversion tracking. Paris walks how ChatGPT fans a prompt into roughly 20 to 30 searches and synthesizes answers from Google's results: the same harvest pattern Google used on publishers.

🐕 Labrador: OpenAI's own search index
OpenAI's Labrador project aims to cut ChatGPT dependence on Google and Bing. If that index becomes robust, classic Google ranking may become optional for ChatGPT-first GEO. Crawlable discovery still matters today; the destination index may not stay Google's forever.

🧱 Three SEO buckets: only crawlability clearly survives
Technical crawlable HTML still matters. Schema is a nice-to-have with weak LLM-citation evidence on prior show studies. Content shifts from human blog traffic to information gain for models. Keyword research is dead for GEO. Links give way to brand mentions beside trusted peers.

🎯 Personalization, SOV vanity, ads, and human premium
AI memory enables hyper-personalization Google never really delivered. Share of voice swings when Astra cites differently than Sol. OpenAI ads already exist with rough UX; Gemini claims 65% voice prompts on its mobile app. Astra and HeyGen-class video compress studio costs, while Paris expects a premium for real human proof.

🏁 Verdict: playbook dead, goals different, PR's opening
SEO optimizes for rank, traffic, and ROI. GEO optimizes for accurate, positive brand positioning in AI conversations (and later agents). BrightonSEO will show how much of the floor still sells link packages. PR agencies and exclusive pitched proof get a new lease on life. Closing beat: Cyrus Shepard's ranking-factors survey of 131 SEOs feels mismatched when the bar is "do not get skipped by AI."

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the platform changes and citation mechanics that actually move AI visibility.

💬 Question: If ChatGPT can answer without a click, are you still buying a ranking-factors playbook, or are you building sources AI will trust and cite?

Welcome to episode 23 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, records a September 8, 2026 solo news roundup. The lead story is Semrush and Exploding Topics finding that 57.5% of AI users said chatbot information caused them not to buy, then nine more beats on mentioned versus chosen, single-check misclassification, Reddit citation collapse, five-engine source overlap, Google AI Mode citation architecture, production UI collection APIs, Claude retrieval versus citation, historical AI data moats, and Mastercard's agent shopping forecast.

In this episode:

🛒 Semrush/Exploding Topics: AI talks buyers out of purchases
A vendor-sponsored Semrush and Exploding Topics survey of 2,338 US adults (Sept 7) found 57.5% of AI users said chatbot information caused them not to buy. 65% said AI had at least partially replaced Google for product research. Among weekly AI users, 73.6% bought from an organic AI recommendation. Paris's GEO take: purchase exclusion, reputation, and buyer risk now sit beside discovery.

📌  Latent Space: mentioned is not chosen
Latent Space's Frontier AI visibility tracker (6,762 answers, 161 categories) shows Kysely in 42/42 answers but first choice only once. GPT-5.6 Sol and GPT-6 Astra category leaders differed in 33/121 comparable categories. Mention share of voice can overstate commercial visibility; recommendation role and sentiment matter.

🎲 Surfaced by: one AI check can misclassify visibility
Across 536 combinations with at least five checks, 80% of appearance cases were inconsistent; a single check matched the majority only 72.2% of the time on inconsistent sets. Paris: target roughly 30 to 40 runs per prompt before volatility flattens.

📉 Surfaced by: ChatGPT Reddit citations 37.9% to 3.4%
In fixed cohorts across two 14-day windows, ChatGPT Reddit citations collapsed while Perplexity (50% to 44.7%) and Gemini (30.8% to 33.7%) stayed comparatively stable. Off-site GEO must be engine-specific.

🔗  Intender: only 8% source overlap across five engines

27,924 citations from 3,600 searches: 8% overlap across all five platforms; 71% of 4,625 domains on one engine only; 65 domains in all five. No single AI citation ecosystem.

🗺️ Nicholas Sitter: Google AI Mode citations flip by question type
Entity-wrapped citations: 87.6% for recommendation-style prompts, 19.8% for explanatory prompts, 0% for generic informational queries. Tag and report prompts by buyer stage and intent.

🔌 querying.ai: production UI collection as a GEO API
Real customer-facing responses across seven surfaces from about $20/month, not model API output. Paris: UI and API results diverge in testing; API-only trackers can mislead.

🧪 Is My Brand in AI: Claude retrieval vs citation
Named-brand sites retrieved in only 33% of run-brand pairs via Claude API web search; 62.7% of retrieved domains cited. Shaky design (API + three runs), but retrieval and citation still need separate metrics.

🏛️ Similarweb: historical AI visibility as a data moat
If nobody sampled an answer when it happened, replaying the prompt today does not reconstruct it. Historical depth separates data companies from prompt trackers.

🤖 Mastercard: AI agents shopping and paying by 2030
Forecast: 300M-plus shoppers may routinely use AI agents to shop and pay by 2030. Teen trust signals already lean agentic. Paris: build agent-ready sites across discovery, recommendation, selection, and transaction.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the studies, platform changes, and citation data that actually move AI visibility.

💬 Question: If AI is already talking buyers out of purchases, does your GEO reporting still stop at whether you were mentioned?

Welcome to episode 22 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, records a September 7, 2026 solo news roundup. The lead story is Peak AI reporting that OpenAI is building ChatGPT's own search index (Labrador) plus vertical indexes, then fourteen more beats on brand-memory GEO tools, Cloudflare crawler defaults, Google AI Mode at 1B users, prediction layers, anti-manipulation defenses, engine-specific sources, SaaS plugins, ChatGPT Ads, ad agents, and off-site citation acquisition.

In this episode:

🔍 Peak AI: ChatGPT is building its own search index
Peak AI (September 4) documents OpenAI developing Labrador plus specialized indexes for news, shopping, PDFs, YouTube, and academic content. ChatGPT can retrieve from indexed and cached web content rather than relying only on Bing/Google-style third-party results. Paris's GEO take: crawl and index behavior becomes a first-order AI visibility variable, reducing dependence on Microsoft/Bing.

🧠 Profound Context Manager = persistent brand memory
Profound's September 1 Context Manager ingests docs, transcripts, and conversations into six brand records (positioning, market dynamics, product, audience, strategic priorities, execution methods) for its AI Marketer agent. Live for enterprise; self-serve via demo. Move from monitoring toward agentic execution and a stickier system of record.

🎤 Semrush Spotlight after Adobe acquisition
Semrush (Adobe company since April 28, 2026) hosts Spotlight in London on October 13 for 1,000-plus CMOs, with speakers from OpenAI, LinkedIn, Wix, WPP, and Adobe. Follows Adobe Brand Visibility combining Adobe and Semrush AI visibility datasets (~300M AI prompts, 28B keywords, 43T backlinks). Paris speaks at BrightonSEO October 8 and may attend Spotlight; contact [email protected].

🛡️ Cloudflare Sept 15 mixed-use crawler defaults
New Cloudflare sites will default-block crawlers that mix training and model-building. Critical now that OpenAI bots also crawl to build ChatGPT's own index. Audit allowlists so you do not erase visibility by accident.

📈 Google AI Mode hits 1B monthly users
Google (I/O 2026 / TechCrunch Aug 11): AI Mode surpassed 1B monthly users; query volume more than doubles each quarter. Search Engine Roundtable: AI Mode capability continues into AI Overviews on a Gemini 3.7 Flash-class model. Azoma: audit citation source mix. Paris: AI Overviews as a bridge to AI Mode as default Google.

📊 AirOps: 86.6% prioritize AI search, budgets lag
AirOps 2026 Marketing Leaders Reality Index (300-plus leaders, vendor-commissioned): 86.6% prioritize AI search; 88.4% emerging/operational maturity; 75% higher targets; only 43% budget increases; under 25% full approval. Full-approval teams 2.3x more likely to expand headcount. Interest ahead of budget.

🎯 Metrisque: 92% recommendation predictability
Metrisque's pre-registered study: predictions before ~1,100 AI recommendations correctly identified territory 92% of the time across 1,916 products. Rankings and mentions explained little. Question shifts from frequency of appearance to why the model classifies relevance.

📉 Rankcaster citation prediction
September 6: Rankcaster ML predicts which URLs get cited using embeddings plus crawl freshness, schema, domain age, and link graph into a P-citation score. Next layer: opportunity, expected impact, action, validation.

⚔️ GEO Defender vs black-hat GEO
arXiv (Sept 2): GEO Defender cuts seven-attack average success from 50% to 6% while retaining 94% benign evidence use across five models. Early black-hat GEO rewriting patterns are here; engines will harden.

📚 Ahrefs: Reddit/YouTube differ by engine
Reddit: 16.8% ChatGPT, 28.5% Gemini top source share. Google AI Mode: Reddit 17.9%, YouTube 17.8%. No universal GEO publication list. Map sources per engine and prompt cluster.

🧩 Perplexity independent retrieval stack
Perplexity uses proprietary embedding/ranking on an exabyte-scale index with its own re-ranking infra. Google ranking is only one upstream signal. Optimize for semantic retrievability and evidence relevance; measure SEO separately.

🔌 OpenAI Zendesk + OneNote plugins
ChatGPT/Codex plugins let models use Zendesk tickets/knowledge and OneNote org knowledge. Visibility expands from recommend my software to use my software (alongside Salesforce MCP trends).

📣 Rebid adds ChatGPT Ads to the media stack
Rebid activates ChatGPT Ads beside Google, Meta, LinkedIn, X, and DB360. Earned GEO and paid visibility need joint management and separate attribution.

🤖 Microsoft Advertising MCP for AI agents
Microsoft Advertising MCP lets Copilot, Claude, ChatGPT, and peers query live campaign data. Stagwell audits; Conversios covers 120-plus accounts. Agency value shifts from daily execution to judgment, experimentation, and proprietary workflows.

🌐 First Page Agency: 800+ citations via off-site context
Zero to 800-plus citations/mentions in under four months across ChatGPT, Gemini, Perplexity, Google AI experiences, and Grok via LinkedIn, Medium, Reddit, and communities rather than backlinks alone. GEO authority as distributed corroboration.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the studies, platform changes, and citation data that actually move AI visibility.

💬 Question: If ChatGPT is building Labrador and its own vertical indexes, are you still managing crawl allowlists as if Bing were the only retrieval path that matters?

Welcome to episode 21 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, records a September 6, 2026 solo news roundup on Unification Day in Bulgaria. The lead story is NVIDIA acquiring Hugging Face for $12.93 billion, then ten more beats on model volatility, execution-layer GEO, llms.txt, citation freshness, publisher volume, ChatGPT ads, high-intent AI Overviews, and B2B buyer behavior.

In this episode:

💰 NVIDIA acquires Hugging Face for $12.93B
NVIDIA agreed to buy Hugging Face, whose platform hosts 3 million models, 500,000 datasets, and 1 million applications used by over 18 million developers. NVIDIA says Hugging Face will remain open and will not require NVIDIA compute. Paris's GEO take: this bolsters American open source and further fragments AI engines beyond ChatGPT, Google AI Mode, AI Overviews, and Claude, making per-platform visibility reporting more important.

⚡ ChatGPT-6 Astra raises model volatility for GEO
OpenAI launched ChatGPT-6 Astra with browsing, research, and multi-step agentic upgrades. Rollout has begun with selected organizations. OpenAI also rates Astra at a critical cybersecurity capability threshold. Major model changes can alter retrieval and recommendations even when a brand changes nothing, so annotate releases in GEO reporting.

🤖 Pepper Agent Atlas executes GEO workflows
Pepper launched Agent Atlas, connecting GEO tracking, Search Console analytics, competitive data, and CMS information, then recommending and executing workflows such as content optimization. The stack is moving from monitoring to diagnose, recommend, and execute. GEOforge connects a similar integration set and competitor share of voice.

🧭 HubSpot GEO Recommendations beta lands in the marketing stack
HubSpot's GEO Recommendations beta analyzes citation patterns across tracked prompts and recommends site content, third-party outreach, and social actions. It can generate gap content and track mention changes. Paris's take: HubSpot's CRM, CMS, social, and revenue measurement create a strong competitive edge as GEO embeds inside the marketing platform.

📄 41% of B2B SaaS have llms.txt without evidence
Treyci scanned 100 B2B SaaS companies and found 41 with llms.txt. No controlled comparison showed more recommendations or citations for adopters (NetContent SEO). Treat it as a cheap experiment and measure before-and-after citation rates rather than selling implementation as the outcome.

⚠️ llms.txt becomes a cybersecurity risk
Ars Technica: researchers scanned 6,214 corporate domains and found 120 llms.txt or llms-full.txt files referencing unregistered packages or domains. After registering abandoned names, coding agents including Claude, Codex, and Hermes executed proof-of-concept code. At least one misconfigured reference already pointed toward real malware.

📅 Brand Ghost: median citation age 263 days
Brand Ghost analyzed 6,762 AI citations. Median age was 263 days, 42.2% were more than a year old, and only 21.5% were under 90 days. That challenges freshness-only narratives. Protect durable research, benchmarks, docs, and evergreen assets instead of manufacturing date updates.

📉 USA Today: more content no longer grows audience
Search Engine Land reported USA Today's September 4 memo: producing more content is no longer as effective while platforms keep users. The volume model is structurally weakening. Optimize for information gain, citations, recommendations, branded demand, and pipeline, not articles-per-month.

📣 ChatGPT ads split paid vs organic visibility
OpenAI's ad tooling adds deeper custom audiences, inclusion/exclusion, bid adjustments, and conversion measurement. Conversational discovery now has earned GEO and paid layers. Paris has not seen platforms cleanly break out paid ChatGPT visibility, so share of voice may already mix apples and oranges.

🎯 High-intent B2B queries are AI Overview heavy
SER Interactive data via Search Engine Land: AI Overviews on 95.4% of X vs Y, 86.3% of review, 83.4% of price/cost/buy, and 81.3% of best-of queries. Start GEO programs with battle cards, comparisons, alternatives, pricing, and best-vendor pages before more top-funnel blogs.

📊 G2: 51% of B2B buyers start with AI chatbots
G2 finds 51% of B2B software buyers now start with AI chatbots more often than Google, up from 29% eleven months earlier. 69% changed vendor plans after chatbot recommendations. Roughly a third bought a previously unknown vendor. 85% view mentioned vendors more favorably. Paris's close: B2B SaaS needs to be GEO first.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the studies, platform changes, and citation data that actually move AI visibility.

💬 Question: After NVIDIA's Hugging Face deal, are you measuring open-source model visibility as a first-class GEO surface, or still optimizing like ChatGPT and Google are the whole map?

Welcome to episode 20 of The GEO Show, the briefing on generative engine optimization, LLM citations, and AI search visibility. Brought to you by GEOforge.

In this episode, Paris Childress, founder of Hop AI and co-founder of GEOforge, walks a September 5, 2026 solo news roundup whose biggest story is Google testing exact and phrase match ads inside AI Mode, then covers nine more beats on clicks, citations, creators, funding, product pages, entity accuracy, crawlers, and traffic redistribution.

In this episode:

📉 AI is not replacing Google users. It is replacing searches and clicks
Similarweb finds 95% of ChatGPT users also use Google even as generative AI visits grew 70% year over year. Baconi research found households with ChatGPT search access reduced traditional searches about 9.4%, while a randomized AI Mode experiment saw external click-through rate fall 18.8 percentage points. Reach, query volume, and clicks are different metrics.

📊 Nearly half of AI Overview citations miss the organic top 10
Organic KPI analyzed 1,208 citations across 160 commercial best-X queries. 77.6% of citations pointed to ranked lists, while only 53.6% of cited domains appeared in the same query's organic top 10. SEO rank and GEO citation eligibility keep decoupling.

Google is testing exact and phrase match ads in AI Mode
Ginny Marvin confirmed a small experiment letting regular search campaigns with exact and phrase match serve text ads in AI Mode when intent is explicit and direct. AI Max and Performance Max stay for complex conversational intent. Paris called this the biggest story of the day: keyword control returning to conversational search, with AI Mode treated as combined earned plus paid visibility.

🎬 Creator briefs are being designed for AI citations
Digiday reports Tinuity, DEPT, Jellyfish, and Ogilvy building AI discoverability into creator strategy. Tinuity says AI visibility grew from roughly 10% to about 25% of creator considerations. DEPT briefs creators on exact product names, verifiable claims, chapter markers, and clean transcripts. Favor SMEs with information gain over mega-follower reach.

💸 Amberos raises $5.5M and builds Creator Graph
Amberos announced a $5.5 million seed round (total funding $6.7 million). It monitors ChatGPT, Gemini, Claude, Perplexity, Grok, and DeepSeek with a brand knowledge graph covering 10,000 brands, and will fund Creator Graph to map which creators and publishers influence AI recommendations. Capital is moving from monitoring toward influence mapping and intervention.

🤖 Mention OS launches an answer-engine visibility agent
Mention OS launched an agent that monitors ChatGPT, Gemini, Perplexity, and Google AI daily, traces cited sources, and can draft fixes for brand-site publication. The broader trend is closed-loop GEO: measure, intervene on owned and earned channels, distribute, remeasure, and compound proprietary data.

 Product pages win 24.1% of B2B buyer-stage AI citations
10 Speed analyzed 7,387 citation appearances across 170 B2B buyer evaluation prompts on ChatGPT, Claude, Gemini, and Perplexity. Product pages generated 24.1% of citations, articles 17%, and Reddit plus YouTube plus forums only 4.2%. The Reddit-mention playbook is too simplistic. Product pages need comparison evidence and specifics.

⚠️ Thin evidence lets AI describe your competitor instead
Duane Forrester in Search Engine Journal: when evidence is insufficient, AI may fill gaps with category norms, stale information, or better-documented competitors. On-site audits may miss it because the failure appears in generated outputs. GEO is visibility and entity accuracy.

🕷️ Training crawlers and answer crawlers need different policies
Publishers increasingly block training bots while allowing search and answer bots. robots.txt is voluntary; CDN, server, and WAF controls enforce blocks. Paris's take: do not accidentally block training-oriented bots if you publish high information-gain content for GEO. Still get indexed.

📈 ChatGPT traffic +48.4% while Bing fell 50%
Search Engine Land / Semrush: ChatGPT.com US visits rose 48.4% year over year to 1.09 billion monthly visits. Bing fell 50%. LinkedIn grew 15.5%. Google itself grew 10.7%. Attention is redistributing unevenly across a fragmented discovery ecosystem, not simply replacing Google.

If this briefing is useful, subscribe to The GEO Show so we can keep walking through the studies, platform changes, and citation data that actually move AI visibility.

💬 Question: If exact and phrase match land in AI Mode at scale, does that change how you split earned GEO work versus paid AI Mode budget?

Welcome to episode 19 of The GEO Show. Paris Childress and Simon dig into information gain: net-new knowledge AI has not already ingested, the 0-100 information gain score inside GEOforge, e-commerce and review paths to lived experience, YouTube AI Overviews citations, and GEO expanding into social, UGC, and influencer go-to-market.

In this episode:

Information gain is the anti-slop ingredient
AI is hungry for net-new knowledge. GEOforge scores information gain from 0 to 100 as the primary forecast for content success.

Novelty is not truth
High novelty still needs corroboration before models cite a claim.

Experience, reviews, and video
Lived use cases and verified reviews beat manufacturer specs. YouTube is a major AI Overviews citation source.

Sites as AI crawler databases
Human traffic falls while bot visits rise. Structure for retrieval, then earn relevant mentions.

💬 Question: What is the highest-information-gain asset you could publish this month that no model has already trained on?