Fastly reports machine-generated traffic crossed 50% of its network in July and August, with AI traffic growing 6.5X faster than human traffic from January through May. Paris Childress covers Searchable Meridian Tech acquisition, rising GEO service demand on Fiverr, Semrush Claude Code citation outreach, ChatGPT vs Google AI Mode retrieve-to-cite funnels, Reddit engine breadth, Deca GEO recommendation benchmarks, Gmail Search AI Overviews, and Pronas AI visibility gap.
Paris Childress hosts episode 41 of The GEO Show on why machine-generated traffic has crossed half of Fastly's network, with AI traffic growing 6.5X faster than human traffic from January through May. The through-line: marketers should design for AI crawlers and answer engines as first-class audiences, not as an afterthought to human visits.
The episode also covers Searchable's seven-figure Meridian Tech acquisition, Fiverr freelancers reporting rising GEO demand ahead of Google ranking work, Semrush's Claude Code citation-outreach workflow versus in-platform CiteForge, ChatGPT versus Google AI Mode retrieve-to-cite funnels, Reddit's overall #1 rank with zero ChatGPT and Claude cites, Deca GEO recommendation rates versus vendor revenue, Gmail Search AI Overviews for Workspace, and Pronas quantifying the AI visibility gap.
This really is the top story. This is major.
Yes, on Fastly's reported network. Machine-generated traffic crossed 50% in July and August, and AI traffic grew 6.5 times faster than human traffic from January through May. Fastly also launched AI runtime control, an AI firewall, and expanded API security for models, applications, and agents at the edge. Paris argues content should be written for how AI crawlers understand pages as machines outnumber human visitors.
In Fiverr's survey of hundreds of top-rated freelancers, 67% reported rising client demand for GEO versus 56% for Google ranking optimization. At the same time, 61% still said traditional SEO is the strongest long-term investment. Buyers treat GEO as a complement and budget line, not a full SEO replacement.
No. Nicholas Sitter's hotel search dataset (616 prompts, 56 destinations, week of September 21) showed ChatGPT retrieving 2,679 sources and citing 2,599 (97% throughput), while Google AI Mode retrieved 12,930 sources (~6X) but cited only 2,103 (16%). Citation counts alone hide retrieval-layer differences.
Not in the Machine Relations Index window covered here. Reddit ranked #1 among 22,320 domains at 12.57% of monitored answer runs across Gemini, Google AI Mode, Google AI Overviews, and Perplexity, but ChatGPT and Claude showed zero Reddit citations. Paris argues citation strategy should scale the long tail, not only chase a few head domains.
Watch the full episode on YouTube: https://www.youtube.com/watch?v=HHOGLoZ5KEE. Audio: https://share.transistor.fm/s/265bcb58.
[00:00:00] **Paris Childress:** Hey everybody. Welcome back to episode 41 of The GEO Show, brought to you by GEOForge: full self-driving for AI visibility. Let's get right into some top stories. Today, we're gonna be talking about M&A activity in the GEO tools space. We'll be talking about GEO service demand versus SEO service demand, and also some product signal.
[00:00:34] So let's get right into the first story of today, which comes from Searchable. Searchable has acquired New York-based Meridian Tech in a seven-figure transaction. Meridian's customers and product will migrate into Searchable. Searchable describes this as one of the first transactions between two companies built specifically for AI visibility and answer engine optimization.
[00:01:05] So here we go. Consolidation has started a little bit early in this category. Um, well-capitalized players can now acquire customer bases, data sets, regional presence, and product capabilities rather than build everything organically. As far as I know, this is the first major M&A deal, uh, within two players in the space.
[00:01:33] We know that already Scrunch has been acquired by, uh, a CMS platform, but these are... This is a merger of two, I believe, roughly similar-sized GEO platforms, Searchable and Meridian. So this is going to be an interesting trend to, to see if it continues, and I would expect this to continue. And the reason is that this tool category has exploded and has gotten very fragmented.
[00:02:03] Lots of players are doing the basic, uh, AI visibility tracking and monitoring, but there are a lot of other players that are doing some very interesting niche things, and they could be acquired by larger players like Profound, Airops, Peak, and they could be acquired for a variety of things. Maybe their customer bases, maybe their technology, maybe a combination of both.
[00:02:31] All right, moving on to the next story. GEO service demand is now outpacing demand for Google ranking optimization. This is coming from a Fiverr survey. Fiverr surveyed hundreds of top-rated freelancers working across SEO, GEO, e-commerce, and user-generated content. 67% of surveyed search specialists reported rising client demand for GEO versus 56% reporting rising demand for Google ranking optimization.
[00:03:10] At the same time, 61% still said traditional SEO is the strongest long-term investment All right, this is interesting here. The thing that stands out most to me is that these are-- this is a legitimate survey. Hundreds of SEO and GEO specialists who are working on Fiverr, most of them freelancers, but this is a big enough sample set that it, it is, uh, I think pretty trustworthy.
[00:03:41] Um, pretty much two out of three of those people surveyed have reported that there is rising client demand for GEO. But at the same time, uh, sixty-one percent still said traditional SEO is the strongest long-term investment. So I think we're moving from GEO being, uh, a budget line, uh, for curiosity and becoming a real service budget line.
[00:04:09] And also buyers are not yet seeing this as a replacement for SEO, but more as a complement to SEO. So probably a lot of the SEO professionals on Fiverr are positioning themselves still as SEO experts who can offer the GEO layer on top. And I think this is a smart way to position both the SEO and GEO and to describe how they can coexist and, uh, frame it by saying that SEO is still the foundation and GEO is a value-added layer.
[00:04:49] I think with our product, with GEOforge, we're a little bit more aggressive because we are saying this is not an add-on to SEO, and we are not a platform for managing your SEO and your GEO, which is how many of the other platforms are positioning themselves, such as AirOps and the legacy platforms like Ahrefs and Semrush.
[00:05:12] But rather we are positioning ourselves as GEO first, not ignoring SEO, but putting GEO first and the workflows and priorities of GEO ahead of traditional SEO, because we believe that that will result in the biggest impact to AI search visibility, but also a strong and positive impact to SEO as well.
[00:05:40] All right. Next story From Semrush. Here, I think this is our leading story of the day and most important. Semrush Has published a new workflow that combines its prompt tracking data with Claude code to identify third-party pages cited by AI answers, analyzing hundreds of those pages, prioritizing targets, and running outreach intended to get a brand included on them.
[00:06:16] Semrush explicitly distinguishes AI citation outreach from conventional backlink building. So Semrush has made a major move in its product with this, uh, operationalization of off-site AI citation outreach. And this is new territory that I think is a natural extension for the GEO tool space. You start with AI visibility, then you move into the execution layer, starting with content production, and then after that, the next natural extension is citation management, citation outreach management.
[00:06:55] Because as we've reported here on this show, off-site citations really have the biggest impact in your AI visibility, more so than the content that you're publishing on your own website. So the, the two sides of this coin for GEO strategy really are high information gain content, which can be published on your owned media properties or earned media properties, but content that is published and generated and made available that contains high information gain.
[00:07:28] And number two is placement on third-party reputable sites, citation building, brand mentions. I checked out Semrush's article here, and really the whole workflow happens in Claude. It doesn't happen in Semrush. Semrush just basically allows you to download your citation opportunities. In fact, just-- you can just download all of the citations that they have scraped across all the prompts that they're tracking.
[00:07:58] And then you, you, you dump that CSV file into Claude, into a Claude project where they advise how to set up the Claude project with specific .md files and instructions. And you basically get the analysis, you get segmentation, and you get a clear outreach strategy that you still have to manually execute.
[00:08:17] Now, what makes our product different is that we have a module called CiteForge, C-I-T-E for citation building, and we actually handle all of the outreach within our platform. So we identify the opportunities and score them, and then you can trigger custom personalized messages, which will be sent over email to enriched email addresses that we look up, so we find the right contacts at those target sites.
[00:08:49] We write custom pitch emails, and we allow-- We have a full email client that allows you to send those emails and track those emails and the replies to those emails all inside of the platform. So not to brag here, but we are a step ahead already of Semrush on AI citation outreach
[00:09:10] Moving on to the next story, ChatGPT and Google AI mode use radically different retrieval to citation funnels. Nicholas Sitter's current hotel search dataset covers six hundred and sixteen prompts across fifty-six destinations. For the week of September twenty-first, ChatGPT retrieved two thousand six hundred and seventy-nine sources and cited nearly all of them, two thousand five hundred and ninety-nine, a ninety-seven percent throughput.
[00:09:43] Throughput here refers to the ratio of retrieved sources to cited sources. By contrast, Google AI mode retrieved twelve thousand nine hundred and thirty sources, so that's about six times more than ChatGPT retrieved, but yet it cited only two thousand one hundred and three, which is a sixteen percent throughput.
[00:10:11] So here we are seeing, uh, really radically different retrieval to citation funnels. These two engines, uh, Google AI mode and ChatGPT, can reach similar answers, but through fundamentally different retrieval architectures. So just counting the citations alone really hides a key behavior of the retrieval layer itself, and it looks like Google is able, just because of, of its foundation in search, Google is able to efficiently retrieve something like six times more sources than ChatGPT.
[00:10:47] But in the end, it's still only citing about the same number of citations Okay, next story. Reddit is the number one cited domain overall, but ChatGPT and Claude do not cite it. Reddit ranks number one among twenty-two thousand three hundred and twenty domains in the Machine Relations Index, appearing in twelve point five seven percent of monitored answer runs.
[00:11:16] Yet its measured engine breadth is only four out of six, and those four engines are Gemini, Google AI mode, Google AI Overviews, and Perplexity, which cite Reddit. But very strangely here, ChatGPT and Claude show zero citations over the measured window. This really flies in the face of the prevailing wisdom right now, which is that Reddit is the dominant-- one of the dominant sources of citations, especially for ChatGPT.
[00:11:49] This study is really saying the opposite. And what it is also showing here is that even though Reddit ranks number one in the whole study out of twenty-two thousand domains, it's still just showing up twelve percent of the time. So if we were to plot this as a curve, there's really just a very, very, very long tail here.
[00:12:13] And the way to do citation building, in my opinion, is not to focus on the one or two or three major, uh, domains at the head, but to try to figure out how you can scalably in, in an automated fashion as much as possible or in a scalable fashion, really try to go after the long tail and exploit the long tail
[00:12:39] Next up, AI recommendation visibility is separating from commercial vendor scale. This is similar to a study we reported on yesterday in episode forty, which had to do with what are the most recommended GEO tools, and does that correlate with their market share, with their revenue? And we know that Profound right now is the leader in terms of revenue and valuation and investment, but yesterday's report showed that Peak was the most surfaced in AI visibility in the category.
[00:13:13] Well, now we have something different. Deca GEO's current September twentieth weekly benchmark ranks Semrush number one in GEO with a sixty-five point one percent mention rate. Profound is number two with fifty point eight percent. Otterly number three at thirty-nine point seven percent. Air Ops is number ten on this list at four point eight percent.
[00:13:38] So we're still seeing a divergence between what are the most popular and most recommended tools versus which ones have the most market share and the most revenue. I would guess that Profound here has a very large sales team, and they're going out and building that business, and it is not-- their growth is not directly a reflection of inbound leads coming through AI search visibility
[00:14:08] Next story. Machine-generated traffic has crossed 50% on Fastly's network. Fastly says machine-generated traffic crossed 50% of network traffic in July and August, while AI traffic grew 6.5 times faster than human traffic from January through May. It launched AI runtime control, AI firewall, and expanded API security to govern models, applications, and agents at the edge.
[00:14:42] Uh, I had previously said earlier in this episode that, that there was a leading story, but this really is the top story. This is major, and this is something that, uh, I think a lot of marketers, we really need to accept this and wrap our heads around this. The web is increasingly becoming a web for machine-generated traffic, more so than human-generated traffic.
[00:15:06] And the mode that we should be in when we create content for the website is to think how are machines, how are AI boss, AI crawlers going to interact with and understand this content because they outnumber the humans that are gonna be visiting. With the, with the rise of this zero-click traffic trend, I think this is only gonna continue.
[00:15:30] We've seen a six and a half times faster traffic growth from January through May this year of machine-generated traffic versus humans, and I think that this 50% is just gonna keep climbing and growing more and more. And, uh, even later today, uh, Simon and I on our LinkedIn Live are gonna be debating whether or not, uh, the internet is even necessary for humans to interact with and visit directly anymore
[00:16:01] This next story is a fun one coming directly from Google Workspace. Gmail search AI overviews is being full-- is beginning a full scheduled release rollout Google's global expansion of Gmail search AI overviews now enters full rollout for scheduled release workplace-- Workspace domains. Paid English language users can ask natural language questions over their email and receive synthesized answers rather than manually search individual messages.
[00:16:38] Oh, this is such a fun story, and for those of you who have been frustrated for years and years, like I have, about just how bad and how weak Gmail search has been, um, you have to, you have to still use very specific search operators, and oftentimes you just can't find the emails that you're looking for.
[00:16:58] Hopefully, bringing AI overviews into Gmail, and really I, I think it's gonna flat out replace Gmail search, should be a really welcome addition. So I can't wait until this gets rolled out and see how well it works. And it is a little bit of a, a forecast to search overall because I do believe that AI overviews and Google AI mode is also a superior experience to classic Google search Now on to our last story of today.
[00:17:31] A twenty thousand eight hundred and thirty-nine prompt study quantifies the AI visibility gap. Pronas analyzed twenty thousand eight hundred and thirty-nine unique prompts producing Google AI overviews in August, recording two hundred and twenty-five thousand citations across nearly fourteen thousand root domains.
[00:17:57] Supporting studies separately analyzed travel and consumer electronic prompts. Pronas calls the mismatch between established web presence and AI representation the AI visibility gap. So the scale of the source layer here reinforces a very important point, which is that strong conventional digital presence does not automatically produce owned domain attribution in AI answers.
[00:18:27] And read another way, I would say strong conventional digital presence refers to search ranking. If you have dominated search ranking in your category, it does not necessarily mean that your domain will dominate citations in your category for AI search. And that is a really, really important lesson to wrap up this episode with.
[00:18:48] So that's all for today. Thank you all for listening, and see you all in the next one.
