84% of AI-cited sources come from earned media, according to Muck Rack research highlighted in Force4's 2026 Market Influence Report. Paris Childress covers the week's GEO news under one theme, the GEO market splitting into measurement, activation and agent trust: the Getfluence and Meteoria partnership, a UCP profile audit, Proof and Circle's identity-gated agent payments, Microsoft's controls for enterprise agents, Rank Control, an AI visibility tracker comparison, a 30.5x spread in GEO market size estimates, and HubSpot's AI search masterclass.
84% of AI-cited sources come from earned media, not brand-owned sites. Paris Childress, founder of Hop AI and co-founder of GEOforge, covers this week's GEO news under one theme: the GEO market is splitting into measurement, activation and agent trust.
French GEO platform Meteoria and media placement company Getfluence announced a partnership that connects AI visibility measurement with editorial activation. Meteoria measures brand presence, sentiment, competitors and cited sources across major AI engines, and Getfluence provides access to more than 70,000 media outlets and contacts. Meteoria reports serving over 1,000 companies and agencies; these are company-reported figures. Paris, who met the Getfluence team at Brighton SEO, sees PR-style outreach to third-party media as maybe the most important aspect of GEO.
Force4 Technology Communications released its 2026 Market Influence Report, a synthesis of more than 50 previously published sources. Its headline finding, that 84% of AI-cited sources come from earned media, originates from Muck Rack research published in May 2026, not a new experiment. The report also draws on G2 buyer research, and the findings use different populations and methods, so they should not be combined into a single causal argument. Paris's view: classic PR outreach is probably the most important part of succeeding in GEO.
Massive tested 36 selected online storefronts for published Universal Commerce Protocol (UCP) profiles. It found published profiles on 11, no profile on 16, and nine were blocked or gave no response. Nine of the 11 detected profiles belong to Shopify brands. Paris's takeaway: the buzz around agent commerce is running ahead of observable deployment.
Proof and Circle announced plans to demonstrate identity-gated agent transactions at Money20/20, using x401 for verified identity and authorization credentials and x402 for payment through a Circle agent wallet. Microsoft expanded its enterprise agent security with dynamic agent risk and dynamic conditional access, which reassess an agent's permissions when risk increases. Both point toward continuously governed autonomy rather than unrestricted agent access.
Rank Control's launch listing describes a platform for finding AI citation gaps, creating and publishing content, and tracking outcomes, with publishing in 23 languages and access through MCP, command line tools and APIs. The listing describes an advertised offering, not demonstrated effectiveness. Paris sees measurement-only tools becoming commodities, with the execution layer, starting with content, now a must.
DerivateX updated its comparison of 11 AI visibility monitoring platforms, highlighting differences in prompt allowances, collection frequency, model version controls, historical depth and integrations. Paris notes that many tools run each prompt once per day, which is basically a coin flip. GEOforge runs each prompt 30 times before reporting share of voice, for 30 prompts every 7 days, to get the margin of error below 2%.
Cyana Marketing compared 10 published market estimates. Across the 8 with a 2026 value, the range was $160.9 million to $4.91 billion, a 30.5-fold difference driven mostly by different definitions of software, services and related SEO markets. The report is a secondary compilation, not a census of revenue. Paris remains bullish because GEO now has a seat at the table with brand and product marketing, budgets that are about 9 times bigger than SEO services by his research.
HubSpot scheduled a masterclass, Found or Forgotten: Why AI Search Is Your Biggest Growth Channel, aimed at smaller marketing teams. It is category education from an incumbent, not a software launch. Paris welcomes it: GEO is not SEO 2.0, and non-SEO specialists need to learn the channel.
Paris: Hey, everybody, and welcome to episode 66 of The GEO Show, brought to you by GEOforge, Full Self-Driving for AI Visibility. Got some great stories ahead. Let's dive right into it today. The theme is the GEO market is splitting into measurement, activation, and agent trust. Today's first story is an interesting partnership.
Paris: Getfluence and Meteoria connect GEO measurement to earned media execution. French GEO platform Meteoria and media placement company Getfluence announced a partnership connecting AI visibility measurement with editorial activation. Meteoria measures brand presence, sentiment, competitors, and cited sources across major AI engines.
Paris: Getfluence provides access to a network of more than 70,000 media outlets and contacts. Meteoria reports serving over 1,000 companies and agencies. These are company-reported figures, not independently verified customer or performance data. All right. An emerging competitive model is coming here that combines monitoring with the ability to influence sources that appear in AI-generated answers.
Paris: And I actually met with the team at Getfluence, during Brighton SEO a couple of weeks ago, and they are very focused on the PR aspect of GEO, which is very, very important. In fact, maybe the most important aspect of GEO, this is one of my takeaways from the conference, is PR-style outreach to third-party media companies to get your brand cited and placed in authoritative third-party sites.
Paris: So this is interesting. They're trying to put together the whole stack here with Meteoria's measurement and Getfluence's off-site work. So it's on-site plus off-site, which by the way, I'll just mention GEOforge has all of these within our four modules in our platform. We've got measurement, content generation grounded in a knowledge base, and a full suite for outreach.
Paris: All right, building on that last story, new B2B market influence report emphasizes earned media. Force4 Technology Communications released its 2026 Market Influence Report, synthesizing more than 50 previously published sources. Its headline finding is that 84% of AI-cited sources come from earned media.
Paris: Importantly, this figure originates from Muck Rack Research published in May 2026, rather than a new October experiment. The report also draws on G2 buyer research about AI-assisted software discovery and vendor selection. The findings have different populations and methodologies and should not be combined into a single causal argument.
Paris: All right, so this kind of solidifies what we just talked about with the partnership earlier between Meteoria and Getfluence, that 84% of AI-cited sources are earned media. So I do believe that earned media, particularly just classic PR outreach, is probably the most important aspect of succeeding in GEO.
Paris: Okay, next story. New audit finds limited adoption of agent commerce profiles. Massive tested 36 selected online storefronts for published Universal Commerce Protocol, or UCP profiles. Its findings: published UCP profiles were found on 11 storefronts, no published profile was detected on 16, and nine were blocked or gave no response.
Paris: So basically, nine of the 11 detected profiles belong to Shopify brands. But only a minority of these profiles that were identified in the study of the 36, only 11 had published their UCP profiles. The rest were either no publish detected or blocked or no response.
Paris: So really, I think that the Agent Commerce Protocol here and the announcements and the buzz around Agent Commerce Protocol for e-commerce is really running ahead of observable deployment across some major retailers. And continuing on with that theme, the next story, Proof and Circle connect verified identity with AI agent payments.
Paris: Proof and Circle announced plans to demonstrate identity-gated agent transactions at Money20/20. The proposed transaction uses two protocols: x401, where an agent presents verified identity and authorization credentials, and x402, the agent executes payment through a Circle agent wallet.
Paris: So this is interesting because agent-mediated transactions are increasingly requiring proof of who authorized an action, not just whether an agent can technically execute it. And these types of protocols are going to be very important for the human users of AI assistants and AI personal agents, for them to gain confidence in sending their agents out to make purchases when they know that there is a verified identity and authorization check that precedes the purchase. Next story, Microsoft introduces risk-adaptive security controls for enterprise agents.
Paris: Microsoft announced an expansion of Project Perception to address risks from AI agents operating inside enterprises. Capabilities include dynamic agent risk and dynamic conditional access. These systems use contextual signals to reassess an agent's permissions and potentially restrict activities when risk increases.
Paris: Microsoft emphasizes maintaining security boundaries independently of the AI model itself. So this is really important. I think enterprise AI adoption is still being held back a lot by these perceived risks, and through this move by Microsoft, of course a leader in enterprise software, enterprise AI execution is moving towards continuously governed autonomy rather than unrestricted agent access.
Paris: So I think this is a good move for the ecosystem that will, hopefully accelerate, AI adoption at the enterprise level. Next story, Rank Control emerges with integrated GEO content execution. Rank Control's current launch listing describes an integrated platform for identifying AI citation gaps, creating content, publishing to customer websites, and tracking outcomes.
Paris: Advertised capabilities include publishing in 23 languages, monitoring major AI engines, and access through MCP, command line tools, and APIs. The listing establishes its advertised offering and scheduled launch, not independently demonstrated effectiveness. So this is yet another new entrant that continues to adopt the same measurement-to-execution proposition that is pursued by lots of other established competitors, including GEOforge.
Paris: So I think the trend here is that measurement-only tools are no longer good enough. These are becoming commodities, and the execution layer is a must, and that starts with content. First of all, identifying what content is needed to close the share of voice or AI citation gaps, and then deploying that content through AI agents. Which leads us to the next story about a pure tracking tool.
Paris: New tracker comparison emphasizes methodology over engine coverage. DerivateX, I think I'm pronouncing that correctly, updated its comparison of 11 AI visibility monitoring platforms. The report emphasizes differences in prompt allowances, collection frequency, model version controls, historical depth, and integrations.
Paris: And the analysis was originally published in September and updated October 19. What is this saying here? That there is an inconsistency, a large inconsistency here across AI visibility monitoring platforms. Differences in how many prompts, the frequency that, these models are checking, the number of runs, are they controlling for model versions or not, the historical depth and other integrations.
Paris: So I think that people who are shopping for a reliable AI visibility tracker as a first step, you really have to look into all these things. How confident are they in what they're reporting? I think most of them are not doing enough sampling to establish a high enough confidence interval. Our approach to this is to run a prompt 30 times in a row before we report on its share of voice for that particular prompt, and we do this for 30 prompts every 7 days.
Paris: And the reason we do that is because we want to shrink that margin of error down below 2% so people can have trust and believe in those numbers that we're reporting. And most of the tools out there are not doing anything close to that. They're doing one run per prompt every day, and it's basically just a coin flip every day.
Paris: Not very reliable. Next, let's talk about the GEO market size. GEO market size estimates vary by more than 30 times. Cyana Marketing compared 10 published GEO, answer engine optimization and AI SEO market estimates. Across 8 estimates containing a 2026 value, the reported range was $160.9 million to $4.91 billion, a 30.5-fold difference.
Paris: The variation largely reflects different definitions of software services and related SEO markets. Some 2026 values were extrapolated from published growth rates, and the report is secondary compilation, not a new census of market revenue. So it's basically this is a huge unknown. How big is this market?
Paris: How big is the GEO market? Is it measured by how much people are just spending on software? What about services? So the estimates are just massive. I mean, from $161 million all the way up to $4.9 billion, we really just don't have any idea. But one thing that I remain very bullish on is this, the potential size of this market, because if we look at comparing SEO services to brand marketing budgets, brand marketing budgets, according to my research, are about 9 times bigger.
Paris: And I do think that GEO now has a seat at the table of brand marketing and product marketing in a way that SEO never did. So it would not surprise me if the size of this market does grow into the multiple billions of dollars over the next one or two years.
Paris: And onto our last story of today. HubSpot brings AI search education to mainstream marketing teams. This was announced on October 20. HubSpot scheduled a master class titled Found or Forgotten: Why AI Search Is Your Biggest Growth Channel. The program targets smaller marketing teams and covers buyer discovery, AI visibility diagnosis, and content strategy.
Paris: This is evidence of an incumbent's category education activity, not a new software launch or proof of adoption resulting from the event. AI engine optimization is increasingly being marketed to ordinary marketing teams rather than exclusively to technical SEO specialists, and this builds on the point that I just made.
Paris: It's a very good thing that GEO needs to be positioned as something bigger than SEO. This is not just an extension of SEO. It's not SEO 2.0 or a layer on top of SEO. It is something very, very different, and a lot of non-SEO specialists need to be aware and educated on GEO. So this is a great, great initiative by HubSpot.
Paris: Of course, they have their own tool. They have their own horse in the race, but, there are very few brands that have the ability to educate as many people in this new channel as HubSpot. So I wish them the best in this endeavor. All right, that's it for episode 66. Thank you all for listening, and see you on the next one.
