Sam Page of AirOps joins Paris Childress on what moves the needle in AI search once you have visibility data: owned vs earned execution, why tactics wear off in months, expertise and first party data as the moat, the AI search attribution problem, a ChatGPT ads thought experiment with sponsored agents, and why the goal is to be recommendable.
Sam Page, head of AEO and SEO services at AirOps, joins Paris Childress to talk about what actually moves the needle in AI search once you have visibility data. Sam explains why the right execution layer depends on the company, why tactics now wear off in months instead of years, and why your team's expertise and first party data are the real moat.
They also dig into Ramp style proprietary data content, the crowded AI visibility tool market, the AI search attribution problem, and a thought experiment on ChatGPT ads where sponsored agents join the conversation and close the sale.
There's no single answer. Established enterprise brands usually already have external mentions, so their wins come from on-site content and fixing bloated site architecture that crawlers give up on. Series B startups often have clean sites but lack external authority, so earned mentions come first.
Sam thinks incumbents have an even bigger edge in AI search, and going from zero to one is harder than it was in SEO's heyday. Clever tactics still work, but they wear off far faster: what once lasted about two years may now last about half a year.
Tools and features are easy to copy. The durable advantage is your experts' knowledge. At AirOps, the brand kit holds brand voice plus product knowledge, sales and customer service transcripts, and first party data, so content can be both on brand and grounded. Ramp shows how proprietary product data becomes content with a point of view.
Attribution is fragmented across brand search, direct, referral and ads. Sam expects before and after comparisons rather than SEO style last click clarity.
Paris floats a thought experiment: sponsored agents invited into a conversation at the moment of commercial intent, possibly paid on commission. Sam notes Amazon already lets ad spend influence organic ranking, and argues ads may force better attribution tools for AI search.
Hi everybody, welcome back to The GEO Show, and today I have the honor of being with Sam Page. Sam Page is the head of AEO and SEO services at AirOps, where he leads a hands-on team that is testing what works in AI search across client accounts. He's been doing SEO for 20 years.
He was Director of SEO at Slack, and Head of SEO, AEO and Content at InMotion Hosting. In October, 2026, he spoke at BrightonSEO on local AEO, where he called it the wild west of AI search. Sam, welcome to the show. Thanks, Paris. I appreciate you having me on. So really incredibly impressive resume here, here, in enterprise SEO. That's definitely something that jumps out, as I read through your bio. You've been doing this for a long time with some massive brands. You've seen and done SEO at a huge scale before. And now you're at AirOps. And having pivoted into AEO, I'll refer to it as AEO, because I know that's the one that AirOps goes with. In your opinion, what is the biggest shift from, let's say, the last 20 years, where you've been focused on enterprise SEO to now AEO? Yeah. Let me lay some foundational work here, too.
So we do actually call it AI search now at AirOps. We were AEO heavy and we pivoted to AI search. Because, yeah, there's1959 phrase now. What we call it seems to be pretty important. And then also, my enterprise, so my time at Slack, was actually relatively the H.I.R.T. I think it was only there for nine months. And this was because AEO AI search was really becoming apparent to me that there was something there. I was a little skeptical at first. But a few months into it, I'm like, OK, there's a lot of telltale signs of growth in this AI search industry. So we were seeing big erosion of our organic search and hockey stick growth in LLM referral visits. When you're at these enterprise levels,
and at a director level, too, you're really more like these alignment calls day to day reporting. You're not really super close to the work in order to be an expert. And I realized I needed testing in AI search. I think where we were earlier, maybe a year or two ago, it was like we just want visibility in AI, what's happening. We need to know the citations, the mentions, the prompts that we're tracking. And I think it's quickly gone from that to now what do we do about it.
And yeah, I think there's a lot of desire to to know what you take from these prompts you're tracking or the visibility you're tracking and how you actually can improve that and do something about it and impact the bottom line for the company.
Yeah. Well, let's dig deeper right into that. Because AirOps, you are one of the early movers on the execution layer on top of just visibility. When I look at the category now, I think it's still mostly saturated with visibility only tracking tools. And there are more and more that are moving into the execution layer. But their interpretation of what execution matters most differs. And so there's execution layers look different. A lot of them have to do with content production and owned media. A lot of them also have to do with offsite work and earned media. I do think that the complete AI search strategy needs to have a good balance of both. But what do you see as the most important execution layers?
And maybe you can frame it in owned versus earned, or there's something else. But how do you all think about the most impactful execution layers at AirOps?
You're going to hate this answer, I think, probably because it just depends on the customer. If you're a big brand who's already established coming to us, what really moves the needle for your company is a little different than a series B startup in a new industry or a fast-moving tech industry. If you're the big brand, you don't need so much of the external citation and dimension building. You pretty much have that already owned. And you really probably should focus more on your own on-site content creation and site architecture. These huge enterprise companies tend to have big, bloated sites. And crawler's kind of like Peter out a couple of layers deep. The site architecture is usually a problem for big sites. Whereas the series B startup, it's like they may have a really effective, efficient site. Content's good. Everything's really easy to navigate, move around in if you're a bot. But you don't have that external authority and branding.
And so that's where you would focus, in my opinion. Yeah, that does make sense. Does that imply that big brands have an unfair advantage already? Yeah. Yeah, yeah. I think for the last 15 years, they've kind of had an unfair advantage, if you ask me. And I think now it's even harder. And especially if you're going from zero to one in today's world, very, very difficult to do. Yeah, it's like you really need to throw a lot of money at these problems and get very specific in your positioning in order to get the same. You probably don't get the same growth that you would have had a few years ago. You're really like fighting over the scraps or the long tail. And this is just my opinion based on a few things I've seen and things I've tested on my own.
But yeah, being an incumbent in a startup in today's world is really difficult. Yeah, yeah. When maybe about a year ago, when I started to think really seriously about AI search, some of the early thought leader, the thought leadership that was coming out was implying that there was a way to leapfrog faster than you could with SEO. So if you were a challenger brand going up against an incumbent in, SEO, forget it. You would probably not get a top three, top five ranking. It might take you a year or two years plus with major efforts. And AEO, AI search was positioned back then as a faster way to leapfrog those competitors if you could outsmart, smart them, if you could do better content. Do you think that was true a year ago? Because I think it has probably changed now. Or was that just a little bit of wishful thinking back then?
I don't know. I think, yeah, you could pull together. You've probably seen this too, where it's like, side two are just churning and burning content. They go up, they go down. And it's not like a long term sustainable business model. So you could still get some business or some kind of growth through that. I think there is still some room for being very clever. And that's what is exciting to me about the AI search space.
What that looks like is probably quite a bit different than being clever in SEO 10 years ago, which was probably something along the lines of building a tool or something and marketing a tool for your business.
Today, you could probably still do a tool, they're very easy to create now with the help of AI. So yeah, I don't know. It's being clever, finding out a new way to do something
or a more efficient way to do something and tying that to your business is, I think there's still some opportunity there but I do think it's harder. I think my perspective is, and you probably have seen this over the years too, is you could do,
things just seem to last longer.
But you could have a longer connection from when you start a tool to, or a tactic to when you see maybe the effectiveness wear off the. And now I think that time gets shortened by quite a bit. So like tactics that worked for a while.
10 years ago, maybe let's say they lasted two years before they wore off. I think that's like exponentially faster now you're probably looking at like half a year.
It's not like you don't have the algorithm changes coming in to disrupt things. It's like a constant recursive improvement, . Or like a new model kind of disarms things too. Yeah, I think that's a good thing in general. What you said about building tools is interesting because really there is no more, I don't think there are any competitive advantages left in building a tool or really any kind of feature or underlying technology. Because everything can be replicated. So then the competitive advantage, in my opinion, it shifts even more towards, and you go to market strategy.
That was really complex and very hard to replicate.
If you had a great product, marketing usually becomes a lot easier. Just get it in the hands of people and then they'll naturally use it and spread it. But now if there's a lot of parity of technology, parity of features, across these products,
yeah, and I think actually the marketing game is now centered around expertise.
So like using your talent, using your expertise to leverage your marketing spend. Like here's this person on our team, who knows everything about this. This is the guy you wanna talk to, or you wanna talk to for X product. They know all the, they've been in it. They know everything about the products. Get that person's brain and put it onto your marketing team. I think that's where the advantages are. Yeah, there's no room anymore for just having like similar products and you wanna choose like, I don't know, you want Tide
or that they'll be less chemical, chemical erosion on. On your nice
on it a little bit.
the other thing that I would like to do earlier to extract their brain and put it into your marketing. We at AEROPs we have a brand the brand kit. All the context for your brand goes into this repository.
All of the answers to the questions like go talk to this person if you want this answer and those are the people that we wanna,
we wanna really get a hold of. Yeah, so the brand kit serves for, I guess both content generation that can conform to the brand guidelines, content that can be on brand, but also content that can be grounded in proprietary knowledge. Is that correct? It does both. Yeah, it does both. Yeah, it's, it's, yeah, here's your brand guidelines in the tone of voice and how you handle punctuation and grammar, but it's also and I think more importantly, it's all of the product knowledge, customer transcript, customer service, transcripts, sales transcripts, first party data, you know, what do you guys have? Every company has some insights that no other company has. It's just the, way it is right now. Are you tapping that? Are you tapping those insights to create content from? Yeah. Even I think especially product companies, they have a goldmine of proprietary data that lives inside of the product itself, the product data in the database. One example, example I think about is ramp. Yes. Ramp has been publishing some great content that's leveraging their proprietary data. And I think they're doing a really transparent way. That most other product companies, in particular software companies are not doing. And I think there's still an instinct to try to protect and even hoard your proprietary data. And in fact, this is the key that unlocks success in AI search. So that So that's a trade-off that every brand has to make, which is, okay, you are sitting on a goldmine. If you really want to win in AI search, which is probably going to help be the biggest
that's going to help you get that up. It's one thing to get r, a knowledge base, that's usually an easier lift, but the actual product data hooking into like a MySQL database, it's a whole other thing. But do you have clients that are willing to go that far? And oh yeah, absolutely. I think a lot of our customers and clients clients they come to us with help to do that. Like they have the data, but they want to be able to turn it into a story. You know, RAMP's sort of like that gold standard for that kind of content. And the funny thing about RAMP is it's actually like it's accurate, but it's also like has its' own bias. So it's not like, right, it's all for companies that use RAMP. Well, I mean, there's a lot of companies out there who don't. They have different buying habits.
So it's like, yeah, it's positioned as like, this is the way it is, but it's really just their POV. And so we, yeah, like we encourage that too with our customers, like, yeah, your POV based on the data that you have. And you can kind of like, you can make a narrative around like market trends. And it's not just products, it's services too.
I think, you know, here at AirOps, we have really cool first party insights too. And from our customers and from the data that we're collecting. So I think for us, you know, we could also pull in and make some really interesting content around our data data. And we've got some really talented folks on the staff that are working to make that happen as well. Like the research stuff is really important right now. John, for everybody.
Yeah, I mean, that's one thing that jumps out at me is that the amount of research coming out in AI search is way, way more than I've ever seen, especially in the early days of SEO. But there is a ton of research. And that's one of the reasons why I started the started this podcast. And I have committed so far to recording daily these stories that I'm getting in an intelligence brief from AI, which is just trying to find all over the internet, find all, all sorts of original research from studies that are being published around AI search. I mean, citation research over here or just all kind of stuff is emerging from all sorts of places, not just the dominant media, but long tail niche media as well. And that's exciting. I think that's gonna move this whole new forward a whole lot faster as learning gets shared more readily. Yeah, I think people will,
or companies will have a better, once they see other companies have success with sharing their internal first party data, it'll be much more, I mean, I think it already is more widely accepted. So that hurdle of like taking your snowflake data I to of making content won't be so high. Yeah, yeah. Coming back to ramp, that there was one particular ramp study that sticks out and I'm sure you saw it. You had the trend in market share for the top players over the last maybe 12 to 18 months. It was profound, AirOps, Peak. And I think there was a few others like Scrunch and Athena
that were lower down. Larger fortune 500, fortune 1000s, but among their customers that purchased software, I guess that's most of them purchased some kind of software, less than 1% had yet purchased anything in this new category, AI search category. I hadn't yet tried profound or AirOps or Peak, less than 1%. And that to me is amazing because first, I assume these are larger companies that might be more reluctant to, for whatever reason, like what we just hooked up snowflake or a product database.
But that to me just says forget about market share. That doesn't matter at all. It's more than 1% of the potential buyers even buying into the category yet. So just total greenfield, I think. Did you see the chart and did you have any? Yeah, I've seen this. I've seen it. I was a little bit skeptical on the data, this seems like it's like very ramped POV.
But yeah, I do think that I wrote a blog post on my personal site. I'm about like the evolution of how these tools have kind of gone. And especially looking at you since we were at Brighton, I was like wandering the halls there and it was interesting to see everybody was either like a link building, a citation building company or an AI visibility company company. And it reminded me of like all and different types building and it reminded me of like all the different types of vendors that were at SEO companies over the years and it's kind of changed from my rank tracking to link building to like all of those sort of interesting business models. Like I remember one that was really popular or as like reputation management and things like this in Google. And yeah, it's just, I think that the way the AI tool space is shaping out, I think like we talked about earlier, it's like tools are very easy to come by.
You and I could vibe code an AI visibility tool probably this afternoon and have it up and running and have some data from it. It's more about like embattled what now, like what do you do from this visibility data? And I think that's the hard part for companies to start actually implementing and getting results from. Yeah, yeah, this is where I was going to go next, was Brighton, that's where we met each other in the round table. And I also, it was clear to me,
she's great. Yeah, I had a nice interview with her and it was so clear to me that, yeah, I don't think think there was any link building tools or certainly SEO rank raking is long gone. I guess you could say there were were some legacy SEO tools like Ahrefs, a SEMrush, of course, and maybe some others. But even they are really pivoting pretty hard over AI search. Google seems to make it harder for them to. Right, their accuracy on the Google rankings is not what it used to be. Yeah, I've been reading about that. Google is actually spoofing them, showing them some false pages. I think I read I think I read the same thing you may have read about that. oraughedactactactactactactactactactact, or just blocking their crawlers or all kinds of really. Really tough. Really tough tactics, on them. Well, you know, I think that these incumbents
are going to have a tough time, but I don't think you can necessarily assume I think that that's going to translate right over to AI search. I think that for I think most of the people believe that this is a newB But SEO legacy software that AI
I think customers will leave and switch to the native AI search first look so to speak native platforms, which is good the news for air ops and others. But man, is it competitive. I think for most part customers are very curious about it and also overwhelmed by the choices that are out there. Yeah, I was always doing this research on the vendors from previous from the last 20 years of SEO conferences. A lot of the trends look to be like consolidation. So a lot of the search ranking trackers kind of consolidated into a handful couple, maybe a two or three big ones. And also agencies consolidating. We're kind of like back in that like whatever Cambrian explosion now for AI and things are going to be
are going to start from when the last generally from whenever the last time a company was an relent like an exhibitor at Brighton or some SEO conference to when they went out. Out of business and so there's going to be a lot of like churn and yeah, it's a lot of interesting insights from like from industry the industry changes, but I yeah, I anticipate like there's a lot of like a lot of differentiation. and so I. There to be some pain for a couple and then also like some like some pain for a couple and then also like some
of the other things that we're going to be doing. I mean, there's nothing . It feels like a gold rush or land. I really do wonder though what percentage if we can extrapolate that ramp statistic about that one percent?, less than one percent. But it's just hard. It's hard to hard to tell really there's like one or two. I think there's like one particular big tool like tool brand that gets a lot of folks into the space and then they turn off and go to the other like secondary third third level tools. Right. Yeah, they get in they get the visibility. They say this is interesting, but I'm not getting ROI from what I'm paying. Like I'm paying a lot of money from this.
There are industries. There's like one or two players that spend a lot of money, get a lot of like awareness to the space and then there's other brands that kind of like do better and and take those customers once they become aware of the market. Yeah, you just ride right in there awake and scoop up and scoop up all that churn. Good. Yeah, but they're doing a service for the for this category. They're blazing the trail. So that actually is good in a way. Yeah. Yeah. And I think that fun. You just mentioned ROI. So I think we should go there. Okay. Now, fundamentally, there was there was some understanding of ROI and SEO because there was a click usually. And I know that last click attribution is flawed. But for the most and no attribution is perfect. But for the most part, you spent money so that you could rank you could rank pages. You could track those clicks and the conversion rates off of those clicks and you could tie that to revenue. And
that's now really a blind spot for AI search. And that is, I believe, one of the biggest reasons why all that there's still a lot of money locked up in SEO, SEO budgets that still hasn't. I think that's the first part of money that's going to shift. a lot.
And it's is a lot of disciplines like product marketing, brand marketing, PR. And there are budgets. That live in those places, too. But the most natural places, the SEO budget is the one that gets that gets looted. No, I don,t know. I don't know. Transferred first. But there's no ROI. There's still no ROI story there. So how do we how do we get there? How do we solve for that? We didn't know how good we had it back when we had SEO attribution model. Yeah. We took it for granted. It was crap. But yeah, what I wouldn't give to have SEO level attribution now for AEO. You know, this is like this is the big problem for everybody. I think the AI search kind of comes in so fragmented. It's brand search traffic. It's ad traffic. It's unknown. It's direct. And it's referral visits. And yeah, it's hard to have a real like confident read on what your AI traffic is.
Like what kind of how is it actually getting to the site? You can have some confidence in your AI visibility. But like that's only a slice of the pie. We want to know how it's actually getting to the site. I think there's probably some education that needs to happen. I think some of the folks trying. I've seen. I like trying to equate your AI search budget to paid media or like advertising on TV TV or some. simultane the
There's a lot of like MMM talk, but I don't see anybody really like super excited about any of that that I don't anticipate attribution. Getting any clearer for AI search either. So I think we just do have to kind of adapt to to a less clear metric. There may be a way for us to sort of normalize a before and after the traffic for from those channels. I mentioned so the direct traffic before and after branded search terms. Even paid ads are going to be a little bit tricky because people click. Maybe they'll search your brand and they'll search your brand in Google after seeing it on AI search and then click an ad from that. So there's some of these things. Yeah, right. Right. Like a before and after and then you can kind of maybe make some inferences from that data.
But yeah, we're just we're going to be just a little unsatisfied. I think with with the attribution model at this point. Yeah, yeah, there's no there's no easy answer for this. I do have ads that I.
Whereby and a sponsored agent will be invited into the conversation by the by the. And the primary agent could be your Claude Duet your dots metamuse or whatever. I I think I think most people will have a primary personal agent R.
OpenAI has. Coz a lot of a lot more details about how ChatGPT ads works. It monitors the entire conversation. It's a multi threader conversation. Unlike triggering an ad off of a single keyword search in a moment. It's it's tracking the whole conversation and and even beyond the memory of the entire the entire account. And it is trying to detect a point of commercial intent. And at that point, it's triggering the ad auction. But instead of an ad auction that triggers a display ad at that point in time. What if that advertiser stood up their own bot their own sponsored agent. That was equipped with all that knowledge from that same knowledge base. We talked about and it was ready to sell and it was waiting to be introduced into conversations by the primary agent. And if and if that happens, then there's a sponsored agent that gets invited into it. Let's say now a three way conversation and that sponsored agent presents you the options captures the lead. Or even makes the sale inside of ChatGPT, which is I think that future will come because there's there's friction with leaving leaving
leaving your environment going to go to a and then just dealing with. A new design and a new experience I would, rather, probably complete a lot of purchases
and strides on attribution. If an agent could come in into a conversation and you'd go from kind of an organic conversation experience into a sponsored conversation in a seamless way. And there are the there's even a blurry line between what is organic search versus paid search. Yeah. But I'm surprised if maybe a year out two years, maybe from now. You're going to see these formats where in the ad platform you actually will just train a bot with your product catalog with your whatever. And it will there's not going to be audience selection. There's certainly not going to be any sort of prompt inputs that you put in. You just say whenever you think it's appropriate to be brought into a conversation, then we'll bid. Probably would be auction based meeting, but yeah, we'll bid to be to be introduced into a relevant conversation and then we'll we'll do all the selling. Yeah, I was like trigger moments or trigger keywords or something that they are listening for.
And they yeah, they jump in on that. Yeah, you know that in that case, last quick point, Sam, is that it wouldn't be a certainly not CPM or CPC. It would probably be just commission based. Yeah. And that point. So the advertiser that says I'm willing to I'm willing to give 20 percent of the sale to ChatGPT if it closes the sale. I don't care if they click through the website or not. But if that if that happened, then just imagine the cost of not advertising with ChatGPT when your competitors are. Yeah, it's an interesting thought experiment. It's kind of been ran through with Amazon and eBay.
Will it impact your visibility in ChatGPT organically to like, they'll have more data on your product. They know it converts. Well, they know customers don't return it or something, you know, like they may know these things even more. So they inform their recommendations a little more. This is what happens with Amazon. It's not fair. I mean, that but this is a firewall, just like Google has always said, there's a firewall between SEO and PPC. But Amazon doesn't. Amazon doesn't care. They like you can you can spend ads on Amazon for your products and it will rank better in Amazon because of that. Like it's a strat. Right. They're using that data. Oh, this converts really well. And then does there's the return rates really low and they ship on time. So will you rank them up higher because of that ad spend too. So the sales volume is part of the algorithm in Amazon. So yeah, it's not like unheard of. It's just Google's kind of had that stance of like, we don't impact organic, but other other companies. Absolutely. It'll impact organic. So why not?
Why not ChatGPT? Well, if that's true, everyone will have to advertise. How could you sit out of ChatGPT ads and they're already at a hundred thousand advertisers and a one billion run annualized run rate for revenue. Yeah. I think they're just barely scratching the surface. Yeah. Yeah. I guess I'm bullish on ChatGPT ads. This I guess we've talked ourselves into being bullish here. Yeah. I think there's a knock on effect or a second order effect of the ads with ChatGPT. And it's kind of contradictory to what I was saying earlier about attribution being more difficult. I think the ads makes it so that you have to have better attribution. Right. Otherwise, people are just going to stop investing. There may be and this is kind of what happened with Google, like where they were like, we're putting in app these ads. So people need to, we need to have some visibility for the advertisers on if they're getting ROI.
So there may be something like this where ChatGPT builds up some type of analytics or AdWords type of tool for us to have some data on. And if nothing else, you could imagine people run ChatGPT ads and then just collect the data and sort of make their own analytics or like their own keyword strategy. Yeah. Yeah. Hopefully that that will come some equivalent of a keyword research tool. Yeah. Whatever that would be like a prompt at least maybe with topics. Yeah. Yeah. Interesting stuff. All right. Well, Sam, I want to finish on this one thing that stuck out from our our pre call, which is your over your overarching strategy for AI search. In your words was to be recommendable, whether it's pursuing an own strategy, publishing on your website versus an earned strategy, trying to get mentioned around the web.
Can you describe what what you mean by. Being recommendable. Yeah, I think you could also think of it. Yeah, I think being recommendable is the key. It's the sounds very easy and it's actually really hard to do. But like if you or I were to recommend something to somebody like why would we recommend that thing to them? It's like we had this personal experience and we found it to be worth remarking. So I think the same thing applies here for your website, your company. Like have an experience that's worth remarking for humans to remark on. Like it's so good. It's this, you know, this. I had to go tell this Reddit this Reddit thread. How good my experience was with AirOps. And I think it's also a lot about being clear in your positioning, clear in your differentiation between, your competitors.
You have to make it so ChatGPT has a reason to recommend you your company over another company. And so. I
and so, yeah, we just need to if you, need to think about this as like, what, what would you recommend if you were the one who would recommend you or your business? Great, great advice. A great North Star, I think, for marketers to keep an eye on being recommendable. And it underscores the hardest work in AI search, I think, happens off site and convincing people why they should recommend you and what makes you stand out. And I think it just forces people to think very hard about how they're positioning themselves in the market. Yeah, I agree. No, and I think the other part of that too is it goes back to attribution too with the the external site investment. It's not easy to say we spent 10 dollars on this placement and we got 10 dollars back.
Right. It's like very much traditional ad spend money. But it's like you were saying like it's a requirement. So. Yeah. Yeah. Well, Sam, this has been fantastic. Thank you for for doing this interview with me and a lot more that I want to ask you. But maybe we'll save that for for another episode. Do a part two. That'd be great. Yeah. Yeah. As we wrap up, where is the best place for people to be able to find you and connect with you? Yeah, mostly LinkedIn. I'm rather non non social socialized on social media. I try to not absorb too much of my brain power with algorithms.
