Conductor, one of the longest-running enterprise SEO platforms, is making AI search the focus of its next phase after its AI product grew customers 12X year over year. On The GEO Show, Paris Childress explains why that pivot signals enterprise SEO becoming GEO, then covers ChatGPT ads, BrightEdge's finding that ChatGPT drives 95.1% of AI referrals, citation decay, multimodal reporting, and prompt universes.
The biggest signal in GEO this week came from an incumbent, not a startup. Conductor, an enterprise SEO platform with more than 15 years in the market, announced that AI search will be the focus of its next phase, promoted its Chief Product Officer Wei Zheng to CEO, and reported 12X year-over-year customer growth for its AI product. Paris Childress reads it as the moment established enterprise SEO officially becomes GEO.
The rest of the episode shows how fast the category around that shift is expanding: paid AI discovery, general AI marketing workspaces, citation lifecycle tracking, multimodal measurement, visual shopping agents, and a sharper debate about what share of voice should actually measure.
Conductor is pivoting because its AI product is where the growth is: 12X customer growth year over year, new logo growth tripling quarter over quarter, and adoption by hundreds of enterprises including Airbnb, Coca-Cola, Atlassian, and Charter. Its agent stack is also being extended through the Model Context Protocol and APIs into partners such as Optimizely, Acquia, Havas, and Capgemini.
"This is a major signal here that established enterprise SEO is really now becoming GEO." Paris Childress
Paris notes the pivot puts Conductor head-to-head with GEO platforms such as Profound, AirOps, and GEOforge. Its existing enterprise customer base is an advantage, but it does not guarantee a smooth transition.
Ads are growing fast in the answers AirOps monitors. AirOps research says ads went from appearing in 0% to more than 50% of answers in monitored industries within three months, a 113x increase in ad load. That is AirOps data, not OpenAI data and not a market-wide measurement.
AirOps launched an AI ads product on September 24 that:
Paris expects agentic ads to spread to Meta and to Google's AI Mode, with ChatGPT likely leading the space.
The two numbers measure different things. BrightEdge reports ChatGPT generated 95.1% of AI referral traffic in its data set, roughly 22 times Gemini, Claude, and Perplexity combined, after referrals grew 101% from January to August. Comscore's data measures prompting activity, where ChatGPT's share fell from 70% to 50% between January and June. Paris calls the gap "very surprising" and suggests ChatGPT may be adding more citations or hyperlinking brand mentions.
Treat citations as a lifecycle, not a ranking. Writesonic's new Citation Decay feature classifies cited pages as rising, holding, declining, or recovering across AI platforms and topics. Paris's playbook:
Mostly not your competitors. GrowByData tracked 44 prompts across ChatGPT, Google AI Mode, AI Overviews, and Perplexity (4,011 answer observations) and found 8 of the 12 most-present domains were outside the tracked competitor set. Paris separates share of voice against named competitors from share of voice across all cited sources; GEOforge lets users name up to 10 competitors as the denominator so third-party sources do not dilute the metric.
Paris closes on measurement: pick a representative prompt set, baseline it, watch the trend, and judge GEO by self-reported attribution rather than share of voice as a vanity metric.
Subscribe to The GEO Show wherever you get your podcasts, and watch the full episode on YouTube: https://youtu.be/D_oI51b-HWM
Paris Childress: Hey, everybody. Welcome back to an episode of The GEO Show, brought to you by GEOforge, full self-driving for AI visibility. We're on episode 45 on September 25th, and we've got some great stories to cover. We have a lot of signals coming from some of the major players in the GEO space, and that starts with Conductor.
Paris Childress: So let's get right into today's top story, which is about Conductor. Conductor, which is really a legacy enterprise SEO software platform, announced that answer engine optimization, AEO, will become the focus of its next phase, and it promoted five-year Chief Product Officer Wei Zheng to CEO. Conductor says its AI product has achieved 12X customer growth year over year.
Paris Childress: New logo growth has tripled quarter over quarter, and hundreds of enterprises, including Airbnb, Coca-Cola, Atlassian, and Charter have adopted it. Its agent stack is also being extended through Model Context Protocol, or MCP, and application programming interfaces into partners, including Optimizely, Acquia, Havas, and Capgemini All right.
Paris Childress: This is pretty major news. Conductor has been around probably for 15-plus years, and it's seen as really the SEO enterprise platform. I think the only one that I think of when I think of Conductor is, BrightEdge as another enterprise SEO platform. But Conductor's got all of the big brands, and what they have signaled with this announcement is really making it very official, I think, of the, the major shift from SEO into GEO.
Paris Childress: They are pivoting their entire focus into, of course, they're using AEO as the acronym, Answer Engine Optimization. They are promoting, the chief product officer of the AEO product to become CEO of the company. So this is a major, a major signal here that, established enterprise SEO is really now becoming GEO.
Paris Childress: And that means that Conductor has to compete now head-to-head with all of the GEO platform startups. That includes the big boys like Profound and AirOps, but also small players like us, like GEOforge. So, they are an incumbent. They have advantages because they have existing huge enterprise customers who are on their platform for SEO, but that does not guarantee that it will be a smooth transition into their GEO or their, their AEO platform.
Paris Childress: So we will watch this closely and see how it develops. The next story comes from AirOps. AirOps launches a full AI ads operating layer. AirOps launched AI ads for enterprise customers on September twenty-fourth. The product monitors live ChatGPT ads. It measures ad load and paid share of voice, analyzes competitors and creative, supports ad creation through AirOps Playbooks and Brand Kit, and connects campaign spend, impressions, clicks, click-through rate, cost per click, and cost per thousand impressions back to market activity.
Paris Childress: AirOps says ads went from appearing in zero percent to more than fifty percent of answers in monitored industries within three months, representing a one hundred and thirteen x increase in ad load. That figure is AirOps research. It's not coming directly from OpenAI. It's not market-wide measurement.
Paris Childress: So AirOps now has a pretty coherent three-pillar positioning strategy. There's on-site content that it helps people do, there's off-site authority, and now there's paid AI discovery. So this is a, a pretty major and very meaningful expansion of, for AirOps beyond organic AEO or GEO execution. So they really wanna become a holistic platform for managing, paid and organic GEO, and we've seen this in, in the SEO and PPC world.
Paris Childress: There have been many platforms out there that try to provide a value-add layer on top of, let's say, Google Ads and Bing Ads and, and others that try to aggregate those platforms. I would not be surprised if AirOps is positioning for this because today we have ChatGPT ads. I think in the near future, we will also have, or we will start to see agentic ads appearing within Meta.
Paris Childress: So basically, meta ads will also become a, a GEO paid surface. And who knows if other platforms are gonna follow suit. Google Ads, of course, is going to incorporate the agentic ad layer in its AI Mode, most likely. I presume that Bing and Microsoft Ads will also follow suit. I don't know if Claude will ever introduce advertising, but this is-- the ad space is gonna be very prevalent.
Paris Childress: And I do suspect that ChatGPT is gonna be the leader in this space. So AirOps here is foreseeing this, and they're trying to expand their platform beyond, just an organic platform. So they've got on-site content playbooks, they've got off-site authority for citation building, and now they wanna bring all these tools into the paid AI discovery space as well.
Paris Childress: And I think that's a really savvy play by them Moving on to Profound. Profound keeps expanding from GEO into general marketing execution. Profound published a September twenty-fourth product announcement for image generation inside its AI marketer. Its current change log shows AI marketer can generate campaign visuals from prompts and reference images.
Paris Childress: While the broader September release also includes dynamic bot rendering, page-level AEO keywords, search console context, Microsoft 365 grounding, and image ingestion into its knowledge base. That's interesting. Image ingestion into its knowledge base. So apparently, its knowledge base can understand images here.
Paris Childress: That's, that's interesting. But Profound is increasingly positioning itself as a general AI marketing workspace, not a narrow vertical AI search analytics vendor. And I do think that this is a strong signal for the entire category. We're starting with GEO/AEO. We're starting with discovery, with visibility, AI visibility and reporting, but it's getting much, much broader, much faster.
Paris Childress: And Profound and AirOps especially are leading the charge here with the expansion of their feature and product offerings. AirOps going into ChatGPT ads, analytics, and management. Profound really with this, with their AI marketer, is getting more and more into, really just a full AI marketing workspace now with, with image generation.
Paris Childress: So I think that's great for the category, and it's gonna mean that the GEO platform category is going to be exponentially bigger than SEO ever was Next story about another GEO platform, Writesonic. Writesonic launches Citation Decay. Writesonic introduced Citation Decay on September 24th. The feature classifies cited pages as rising, holding, declining, or recovering, and it compares movement across AI platforms and topics through its citations, overview, and momentum views.
Paris Childress: All right. A very interesting move here because as we have seen and reported many times on this show, citations have a short shelf life. They're not like rankings, which tend to be more permanent and deterministic. One day you've got a citation, the next day you've lost it. So they're fleeting. So this is a very, I think, a very savvy move by classifying citations as which, which are rising, holding, declining, or recovering, and that brings a level of nuance that I think is gonna be very helpful for marketers.
Paris Childress: Obviously, you wanna hold on to citations that you do have in place so that they don't drop off. You wanna monitor the ones that are rising and try to maximize their presence. And then if some are declining, then you may want to address that by either letting them go or maybe refreshing them to try to restore their presence.
Paris Childress: Or, maybe it's just a natural sign that there should be a new piece of content that takes its place as the best citation prospect. So I think that's gonna, that's gonna allow their users to, do a deeper level of analysis into these citations and then start to prioritize which pages they want to work on most for citations.
Paris Childress: Next story, Google makes multimodal AI discovery measurable. This is from Google Search Console, an announcement yesterday. Google added web multimodal search reporting to Search Console on September twenty-fourth, including both search results and generative AI performance reports. The new report, or the new filter covers Google Lens, Circle to Search, uploaded images, and Chrome's Search This Image flow.
Paris Childress: The rollout began globally on September twenty-fourth. So AI visibility is no longer primarily a text prompt measurement problem. We now have camera and image-based discovery that are entering the mix, and they're becoming measurable distribution surfaces in their own right. And Google's announcement here really legitimizes that because they have decided to commit resources to building, to building a Search This Image flow into Google Search Console
Paris Childress: Next story from BrightEdge. ChatGPT now generates ninety-five percent of AI referral traffic in BrightEdge's data set. BrightEdge reports ChatGPT referral traffic increased a hundred and one percent from January to August of this year and reached ninety-five point one percent of AI-generated referrals in its data set, roughly twenty-two times Gemini, Claude, and Perplexity combined.
Paris Childress: ChatGPT referrals rose another thirty-six percent month over month in August alone. They were careful to point out this does not conflict with Comscore's fresh finding that we reported on a couple of days ago that ChatGPT's share of AI prompt volume fell from seventy percent to fifty percent between January and June, while Gemini and Claude gained share.
Paris Childress: One data set measures outbound referrals, so the BrightEdge data is measuring the actual referrals, people clicking from ChatGPT into a, into a website, versus the other, which measures prompting activity overall, and that is a key distinction. So it looks like ChatGPT, according to Comscore, is losing market share of prompt activity, but they are still, sending the dominant share of referral traffic.
Paris Childress: That could be because they may be, changing the experience more. I haven't really noticed, but they could be putting more citations in. They, they could start to now hyperlink the mentions of brands and solutions within their responses. But to me, this is very surprising and, and almost hard to believe that as ChatGPT is losing market share of prompt volume, they are-- they're massively dominating the AI referral traffic space.
Paris Childress: But, I guess these two things, these two things are reported separately, both from BrightEdge and Comscore, and, and it is an interesting finding indeed. Next up, Meta turns Muse into a visual shopping agent. Meta Muse, which is the hottest AI product out there at the moment. At Meta Connect, Meta announced that Muse is coming to its AI glasses and can act on what a wearer is looking at.
Paris Childress: For example, a product on a shelf or a shopping list. Meta is expanding Muse's commerce connectors beyond Shopify to Walmart, Best Buy, American Eagle, Dick's Sporting Goods, Fanatics, Gap, Michael Kors, Sephora, Ulta, Wayfair, Instacart, Shop Pay and PayPal. What a list. Expedia is coming soon. Meta expects more than one hundred AI glasses styles by this year, by the end of this year.
Paris Childress: Wow. They're making a massive bet on the AI glasses, and I think I, I've heard that they are now shipping versions of these glasses without cameras so they can avoid the, all of the privacy complaints. but I wonder there still needs to be a camera here to detect what a wearer is looking at, but maybe there won't be a recording feature.
Paris Childress: I'm not really sure. But putting this into the GEO context, product discovery is, is moving very quickly beyond just text on a screen. product is-- discovery is, is moving from the, the, the LLM experience of today, which is a prompt or a query and getting an answer, towards actually wearing a device, seeing something, asking about what you're looking at, making a decision, probably with the help of your AI assistant, Muse, and then completing a transaction, also maybe even handing off or delegating that transaction to Muse or your AI agent.
Paris Childress: So, the personal agents are really starting to, clearly become the next phase of AI here, and visual agents should eventually be treated as a distinct discovery surface instead of just being shoehorned into conventional LLM visibility measurement Next up, Spangle launches a retail native AI discovery platform.
Paris Childress: Spangle launched AI discovery for enterprise retailers covering ChatGPT, Perplexity, Google, Alexa for Shopping, and Walmart Sparky. The platform identifies demand a retailer is positioned to win, uses governed AI agents to execute changes, and connects visibility to traffic conversion and revenue. It's also offering an eight-week AI discovery sprint.
Paris Childress: That's an interesting play here. The vertical GEO products have a structural advantage when they combine visibility data with proprietary assortment, inventory data, product, and revenue context. So this is a vertical data partnership here, and I think that this is an interesting spin-off of classic GEO because they can detect, demand basically from, on the, on the demand side.
Paris Childress: They can identify demand that a retailer is positioned to win, and then they can start to take action, and they can start to promote pages or make changes to certain pages to make them more visible to AI, specifically to take advantage of, real-time demand that they detect. So that is gonna be some interesting technology to follow.
Paris Childress: Next up, eight of the 12 most prominent AI answer sources were not tracked competitors. GrowByData tracked forty-four prompts across ChatGPT, Google AI Mode, Google AI Overviews, and Perplexity, capturing four thousand and eleven answer observations. Among the thirty-one domains it reviewed, eight of the twelve most present domains were outside its tracked competitor set, including platforms, publishers, and other third parties.
Paris Childress: So conventional competitor monitoring misses a large part of the ecosystem which is actually shaping AI answers. And, I find this interesting because the, the line is being blurred. If, if in traditional SEO, you al-- you would always measure your rankings versus your competitors, and it was very clear the, the competitive set was pretty much black and white.
Paris Childress: But here, the... It's getting a bit blurry. I mean, what's a competitor and what's not? I mean, clearly you have your business competitors, but the majority of the citations that you're competing with are actually not your real competitors. They're just other sources. And so it's in-- it's interesting, and I think it's important to distinguish between how you are doing against your competitors in share of voice, let's say, and how are you doing overall with share of voice.
Paris Childress: Because in many cases, your competitors may be just a fraction of the total citations that are appearing. And if you take your share of voice among all the citations and you treat all of those citation sources as competitors, you're probably gonna see a much, much lower share of voice than if you name specific competitors and then you track your share of voice only against those competitors and those domains in AI visibility reports.
Paris Childress: What we do at GEOforge is we allow the, the user to list specifically up to ten competitors, and those are the only competitors and that-that's the only competitive set that we're using in share of voice calculations. So effectively, that is the denominator in the share of voice calculation. And even if there are many, many other citations that are not in that competitive set, we don't count that.
Paris Childress: We don't use that to effectively dilute the share of voice Onto the last story of today. Searchable productizes the prompt universe Searchable is introducing Prompt Universe in a live launch session today, September twenty-fifth. Its workflow starts by generating the full question space around products, audiences, and buying situations, then narrows that universe into a representative monitoring set.
Paris Childress: It published Uber example... Its published Uber example contains twenty-two thousand four hundred and forty-two generated prompts. Registered users are promised free accounts to run their own Prompt Universe after the launch session. All right, this is an interesting approach to prompt discovery and prompt sampling and, and here this is an example of, of one of the players in this space, Searchable, really using prompt discovery and prompt sampling as, really as a standalone differentiator, a, a product capability.
Paris Childress: But it's an interesting approach nonetheless, which is to say we operate-- a brand operates in a certain category. We're going to try to define the entire prompt universe around that, that category or maybe a topic area, which to me seems like a Herculean task. I, I don't know how you do that, actually, because I consider that a prompt universe is infinite, actually.
Paris Childress: There is an infinite long tail of ways people can ask questions about your brand, and that's gonna change based on different personas, different situations, even just different ways that people have of phrasing a question. So I do quest-- I, I first of all would challenge how are they coming up with this prompt universe?
Paris Childress: But let's just say they're able to do that. They have some massive, universe of, of prompts that surround a particular topic, and then they're going to distill that down into a representative set. And, and I would like to know how large is that representative set and, and how representative is that set?
Paris Childress: Does it represent equally across the funnel, top of funnel, middle of funnel, and bottom of funnel? Because you have very different intent at different stages of the funnel. You have educational intent and informational intent at the top of the funnel, whereas you have more buying intent and stronger commercial intent at the bottom of the funnel.
Paris Childress: Also, what about personas? Is it representative across, a, a good set of personas that represent all the potential buyers? I'm not really sure about that either, but it's an ambitious task. It's an interesting approach to prompt discovery, and I still believe that prompt discovery is a very much unsolved thing in the GEO tool space.
Paris Childress: There have been some interesting approaches to it, persona-driven approaches, multi-funnel stage approaches, but ultimately, I think You, you need to have a set of prompts that you decide this is good enough, it's representative of where I wanna appear in AI search, and I'm gonna start tracking it, I'm gonna baseline these results, and I'm gonna observe the trend over time.
Paris Childress: But I'm not gonna obsess over share of voice purely as a vanity metric for my prompt set, my representative prompt set, because I wanna remind all of our listeners what really matters for ROI from GEO is asking your leads, asking your prospects that come through to fill out, answer the question, "How did you hear about us?"
Paris Childress: And it's self-reported attribution here. If they say AI search, AI chatbot, and that needs to be always one of the answers or one of the potential options for them to select, then downstream in your CRM or, or somewhere else, you wanna correlate that response data to your sales pipeline or to your, your lead volume, and that's the true way to measure success here. All right. That'll wrap it up for today's show. Thanks for listening, and we'll see you all in the next one.
