Lily Ray + Max Willens

ROB: [00:00:00] Hi, Rach.

RACHEL: Hi, Rob.

ROB: How you doing?

RACHEL: I'm doing good. How are you?

ROB: I'm not bad, not bad. I've got AI on my mind at the moment, um, specifically about how people are using it in their personal lives, and you're my guinea pig for this question. So how are you actually using, like, ChatGPT or Perplexity in, like, your own life right now?

RACHEL: I am using AI basically as my unpaid assistant, um. I think at work it refines my emails, helps me sound smarter, and then-

ROB: It's put a lot of effort then already.

RACHEL: Exactly, so it's overworking. Um, and then in my personal life, I think, you know, it's my doctor. It's explaining my medical results, um, helping me overanalyze Zillow listings for new homes, um. And then also they are a beauty specialist, and they're helping me color match my makeup, so.

ROB: Oh, right. How's that going?

RACHEL: Not well. But at this point, AI is really just-

ROB: Augmenting your [00:01:00] adult life.

RACHEL: Yeah, exactly.

ROB: Welcome back to another episode of Awin-Win Marketing Podcast, and welcome back to the podcast too, to you, Rachel Tyrer.

Good to have you back. How you doing?

RACHEL: Hi, Rob. It's good to be back as your co-host. You missed that part.

Following on last week's episode, we're once again doing something a little different this week. While we were out in Chicago for Think Tank Americas, we sat down with Lily Ray and Max Willens off the back of their session.

And then for anyone who doesn't know, they are in fact returning guests. Lily, uh, joined us at last year's Think Tank in Portugal, and then Max appeared in our 2026 predictions episode in January, um, with you and I, Rob.

ROB: That's right. Um, and for those of you who didn't listen to those episodes and don't know these guys, they're really, really interesting, well-informed figures from across the industry.

So Max Willens is principal analyst at eMarketer, and he specializes with a focus on affiliate and creator. Um, and then Lily Ray is founder of Algorithmic, but also a kind [00:02:00] of SEO guru who I'm sure many listeners are familiar with from her regular posting on LinkedIn about the latest kind of algorithm changes and updates from Google, et cetera so. She also happened to be our keynote speaker for day two, where she shared a lot of her kind of findings and recent analysis, um, with the audience at Think Tank.

So among those things, I would say that kind of like one of the main takeaways was the fact that whilst there's been a lot of kind of analysis around the fact that GEO or AEO - however you choose to label it, this new kind of optimization of how you can show up in LLM citations - is, um, this new novel thing in the industry that people need to be aware of and educate themselves around.

Actually, Lily's analysis and even more of the recent kind of updates and guidelines from Google themselves, suggest that SEO principles are basically the foundation for a lot of this anyway. So if you're doing good SEO, the chances are that you're also gonna show up pretty well as far as LLM is concerned.

So kind of a bit of reassurance, I think, from, [00:03:00] from, in that regard, and I'm sure that many affiliates that are well-versed in these practices will be, um, ensuring that they shore up those SEO foundations.

RACHEL: We talked about the changing role of affiliates and SEO in the age of LLMs and AI search overviews.

ROB: Yeah, and we also discussed the big publisher companies choosing to block AI from scraping their content, companies like the New York Times, for instance. As well as how influencers and content creators can tailor their content to appear in AI citations.

Lily, Max, thanks so much for joining us at Think Tank. Uh, it was a great session we just had this morning. Lily, you gave your keynote presentation providing a breakdown of the latest trends in LLM search and AI visibility and its implications for affiliate marketers. And then you joined Max, uh, who moderated the panel alongside a whole bunch of senior industry figures, uh, from across different perspectives, publisher, advertiser, agency, [00:04:00] to provide their perspectives on that as well. So really insightful for the audience that were there. But we wanted to take this opportunity to also share some of the highlights from that conversation and that presentation with our- with our listeners as well.

And one of the first things that stood out to me, Lily, about your opening was just about talk about that speed of change that's happened. You talked about, like, two years ago, you know, was it, uh, ChatGPT or was it Google?

LILY: AI Overviews.

ROB: Yeah, AI Overviews were giving this recommendation about eating rocks, and two years later you're now able to build out, like, a CRM platform effectively on your mobile in, like, 10 minutes.

LILY: Yeah.

ROB: And, uh, is that unexpected for you about how rapid that acceleration has occurred?

LILY: Yes.

ROB: Is it an order of magnitude kind of quicker than we were all expecting?

LILY: Yeah, I mean, I think there's been a lot of debate about whether or not large language models are truly gonna be, like, the real AI technology-

ROB: Mm-hmm.

LILY: ...that gets us to artificial general intelligence. So there's been speculation about how quickly these things can truly move. Will they always have these hallucinations and everything built into them? So I was kind of skeptical because I thought-

ROB: Yeah.

LILY: [00:05:00] ...you know, if Google, for one, you know, the biggest player in the space, is launching AI Overviews, and even a year or two into AI Overviews we're still seeing these kind of, like, erroneous answers.

ROB: Yeah.

LILY: For us to now be where we are several months later where we're seeing tools like Cloud Cowork and Cloud Code and just agentic, you know, AI search just changing and evolving so rapidly-

ROB: Mm-hmm.

LILY: ...and being able to do so many things, I think we're in definitely that exponential part of the curve, and AI's learning from itself now, so it's- it's happening much more quickly than I think I and many others anticipated, and it's honestly a little scary where things are gonna go from here.

ROB: Yeah, yeah, yeah.

MAX: But I feel like the, uh... everyone should just kind of buckle their seatbelt because as fast as the growth has been, like, one of the things that you pointed out in one of your slides is that if you kind of look past the top lines, like Claude's coming in real hot because they focused and they said, "Right, we're gonna be the thing that helps you at work-"

LILY: Mm-hmm.

MAX: And just focus on that, and that's proven to be really valuable, and I feel like there's a, you know, good chance that someone will come along and say, "We are going to be the one that really helps you shop or that really helps you as a creative partner or [00:06:00] something." And so, like, as that happens, there's gonna be a whole cascade of changes and innovation too.

So, like, as dizzying as this all is, like, everybody just needs to kinda, like, buckle up, I think-

ROB: Yeah.

MAX: ...'cause it's all, like, really just starting.

ROB: Yeah, yeah. I think Chelsea from Acceleration Partners, she threw out the question effectively in the statement that is the industry actually prepared for this? Is it actually, you know, um, uh, well set to kinda deal with this level of revolution?

What was your kind of take or, you know, feeling around the sentiment from the panel?

MAX: The way she phrased it was, was super striking to me where she just said, "This is not an industry that's built for AI."

RACHEL: Yeah.

MAX: I would say that really few industries are, but the, but she also made the point that, like, it's really well situated to deliver real value and, you know, have it play a kind of pivotal role, and, like-

ROB: Mm

MAX: I don't wanna put words in Lily's mouth, but I feel like the other panelists, you know, seemed really excited by the prospect of what this could do for the industry, right? Because it, it has raised its profile in a way that a lot of publishers and [00:07:00] agencies have been clamoring for for a really long time, and, like, because of how early everything is, like, it's, it's really an open question, like, how much one can capitalize on it. You know-

ROB: mm

MAX: ... there's a big, big opportunity, I think, to raise the profile of affiliate publishers, of, of making affiliate agencies much more of a kind of strategic component in how brands think about their marketing spending. Even the brands that are investing in it, right? Like, the signal that you can get from an affiliate customer is, you know, unique and powerful in its own way, and so I feel like all the stakeholders, if they can, you know, not get too spun around by all the, you know, innovation that we all spend time talking about, like there's a lot of opportunity here.

And so I think that even though everyone understands there's a lot to figure out, like I think there's kind of a cautious optimism. Does that broadly sound right to you?

LILY: Yeah, I think, um, like I think it's more of a mixed bag for a lot of the clients and everything that we speak to because on one hand, yes, like there's so much new opportunity to use AI to create, you know, [00:08:00] automations and create efficiencies in your workflow and to vibe code new things and create all these types of new products and everything. So on that end, people are very excited. There's all kinds of like inspiration for building new things in a way that we couldn't do before.

And on the other side, you know, especially with publishers, there's just a lot of concern, a lot of fear because I think, you know, tools like Google's AI mode, which is potentially rolling out any day now as the default version of search, they just fundamentally cut into traffic that many companies have been used to receiving from organic search for a long time.

So I think there's hesitation and there's also excitement at the same time. But everything's changing so quickly that I think it's, it's just hard to say where we're all gonna land. But I imagine in the future we'll probably see a more kind of healthy ecosystem where we have all these AI answers, we have these AI solutions, but we also still have to be fair to the people that are basically fueling all the content that AI is using to answer questions.

ROB: Yeah, 'cause I feel like we, you know, we had this conversation 12 months ago at Think Tank in Portugal, and it still felt very, very early, very novel. Um, there was a lot of question [00:09:00] marks and I feel like there is, uh, a degree to which the platforms themselves have become pretty well established now. You talked a bit, a little bit about the, the growing user adoption, obviously, not just ChatGPT, but increasingly Google with its massive distribution across its other platforms.

But I also just wonder whether or not that question around monetization, we still feel like we haven't maybe progressed that far in that regard in terms of potential solutions. And I wondered if either of you had an opinion on where you think that direction of travel was going?

MAX: Th- this came up on the panel.

Like, there was a big question about, like, you know, what happens in a world where so much depends on CPCs, CPAs.

ROB: Mm-hmm.

MAX: Like, that's a mismatched, uh, metric, you know, in a world where, like, AI is potentially helping you convert a lot. But, like, it, it's, it's important to remember that we are still really early, right?

Like, one of the things that leapt out at me from the other, uh... I'm now gonna talk about other content that went on, uh, this week.

ROB: Mm-hmm.

MAX: But, like, there was a really good presentation that, uh, your head of, I think, insights and measurement put on.

ROB: Mm-hmm. Gareth.

MAX: Yeah, Gareth, and he, he had [00:10:00] a stat about, like, based on some measurement you guys have done, like, I think it's 0.24% of all-

ROB: It's, it's a fraction. It's tiny.

MAX: Yeah.

ROB: Yeah.

MAX: And frankly, like, until that number gets bigger, like, the, the guys that are gonna need the, the OpenAIs, the Googles, they're just not gonna take the time, I think-

ROB: Mm.

MAX: ...to sort of furnish the data that will allow everybody to be on the footing of understanding, like, okay, this is what the citation, you know, metric is. This is how effective this is.

ROB: Mm.

MAX: This is, um, these are all the people in the click stream that, you know, should be compensated. And so it's definitely important to get everybody thinking in a creative and novel way about it, and probably, like, get some brands out of the, like, "It's gotta be CPA. This is how we do it. This is how we've always done it." You gotta get out of that mindset.

ROB: Mm-hmm.

MAX: But, like, I do think we're still so early that the notion that some company is gonna be like, "I've got a new metric and a new d- you know, system," and, like, that person's gonna materialize in the next six months is, I think, a little misguided.

ROB: Mm-hmm.

LILY: Yeah, I think, um, it's a bit unclear right now. I [00:11:00] think one of the things I tried to get at in the presentation is the fact that, so according to some of the data that I looked at 1 in 4 of the most heavily cited or retrieved sites in ChatGPT's top one hundred most cited domains is technically an affiliate or a product review site.

And I think that speaks to the fact that so many consumers are going to places like OpenAI to- and ChatGPT to have, you know, kind of like transactional or research conversations where they're learning about different types of products. And realistically, they rely on all the great work that product review sites have done to review different types of products.

And, you know, we also talked about different ways that these large language models and AI tools are reducing traffic that these sites are receiving over time. So a lot of these businesses have for, you know, decades now have been in the business of driving page views and, you know, monetizing through page views.

And I think that that's an interesting conundrum, right, where it's like these language models need that content, but they're not necessarily sending out that traffic anymore. So I think we're at the early days of having these conversations, and I think that we probably are gonna need a lot more almost like existential conversations with the Googles and the [00:12:00] OpenAIs about if you need our content, we need to make sure that we're getting some value in return.

ROB: Mm.

MAX: Yeah, I mean, I think it's not an accident. This is something that is a couple of years old now. But like the first content partnerships that OpenAI had struck, like over a third of them were with publishers that have really robust affiliate content operations.

ROB: Mm.

MAX: You know? Like, they, they knew from day one, we have to keep these guys happy or connected or compensated in some way.

ROB: Mm.

MAX: And they, they didn't mess around. And, and you can see that in the, in the partnerships they forged.

ROB: Mm.

MAX: And as this grows and they-- I think they'll strike more partnerships and, you know, the, the sort of shape of those is gonna be really interesting to watch.

ROB: Do you think there's a different sentiment towards affiliate content from the LLM platforms than there has been historically with more traditional Google search?

Because Google's had a very ambivalent, is probably the best way of saying it, attitude, I think, towards affiliate content. Sometimes it's been penalized, sometimes there's been, uh, core updates obviously, that have specifically targeted affiliate content-

LILY: Yeah.

ROB: ...and said that this is not what we want in [00:13:00] our index to kind of rank highly.

LILY: Mm-hmm.

ROB: Is that changing, do you think?

LILY: Yeah. So Google has done all types of big updates in the last, especially in the last few years, to actually reduce the visibility of a lot of affiliate content, and I think that's because there were so many people that were able to just spin up an affiliate website-

ROB: Mm-hmm

LILY: ...you know, and not necessarily put in the hard work that Google recommends that you do when you review products. So they also introduced something called the reviews ranking system a couple years back, where they introduced product review updates, like specific algorithm updates to basically reward content creators that are putting in the hard work of doing the product reviews, but in a way where they're taking videos of themselves, they're doing images, pictures-

ROB: Mm-hmm.

LILY: ...original photography of the reviews, extreme, you know, methodologies behind everything. Really the real evidence that they're real people reviewing the products because so many affiliate companies were just not reviewing products and they don't have that actual evidence. So Google's done a lot of work to kind of clean that up.

Then I do think right now, you know, ChatGPT and the other LLMs do tend to pull in from [00:14:00] more kind of like unknown affiliate sites. So in a sense, there's a little bit more opportunity right now because they're not as strict with what they're looking for. But I think ultimately it's going to land exactly where it landed with Google with- which is that they want real evidence that the review content that they're looking at has been put together by somebody who's done the really hard work of creating-

ROB: Yeah.

LILY: ...really great review content. So I think there's only honestly like a small handful of players relative to the overall affiliate space that are really doing the hard work, and I think they'll be rewarded long term.

ROB: Yeah. So do you think it's those kind of first principles of SEO, EEAT is still basically gonna be the, the kind of governing model of what content is gonna be performing well?

LILY: Yeah. So EEAT is, uh, Experience, Expertise, Authority, and Trust. It's E-E-A-T. That's the fundamental way that I think a lot of this works, and I think that what Google and eventually OpenAI are really gonna be looking for.

ROB: Mm-hmm.

LILY: Um, but Google actually has its own criteria for writing good reviews, a lot of people don't know this.

So you can actually go into Google and say, "Google's guidelines for writing review content," and they have like 20 or 30 questions that they ask site owners to think about when they're producing review content. So check those first.

ROB: Yeah. Yeah.

RACHEL: One of [00:15:00] the things that kind of stood out from your session that I, I don't think maybe was talked about in, in Portugal or, or I've heard before, but - the decision for some of these publishers or brand sites that actually block AI bots.

I don't know if you can speak a little bit to that or maybe like what are the pros of that? What are the cons? Why would a site not wanna be cited by these LLMs?

LILY: For sure, yeah. So this is a big conversation and it's definitely one that's happening more so in the publishing space. Um, the thinking, and a lot of big companies, for example, New York Times are already doing this, where they're saying, "Our content is so valuable that we don't want it being scraped and repurposed for free. Even if we can get that potential citation-

RACHEL: Mm-hmm

LILY: ...most people don't click on it." That's how these models are designed, right? These experiences are designed to keep users directly in the interface in the AI chatbot. So they decided maybe, "You know, we'll get traffic and subscriptions through other methods. But we don't want a ChatGPT to be able to scrape our recent content." And when you have enough publishers that are banding together and doing that, now [00:16:00] OpenAI has a data problem or a content problem because they don't have access to all the best news publications. So I think we're already seeing that in the news space.

I don't know if we see it as much in the affiliate space, but you know, in theory, if all the big affiliate companies banded together and blocked, you know, OpenAI or Gemini or all the different LLMs from being able to access, whether it's some of their content, all of their content - the LLMs are gonna have no choice but to reach out and say, "We need that content back. So what do you guys need for it?" Right? So I think these kind of conversations around like licensing of proprietary data, I think needs to continue to happen because the ecosystem can't support itself if they're just able to take everybody's information for free.

ROB: Mm-hmm.

RACHEL: Do you think that's a feasible reality, like affiliates coming together and kind of taking that stance?

Because I don't know, at, at this point, maybe there's only a handful. I think you mentioned, you know, a larger publisher in your, in your speech, but I mean, that's a bold stance.

LILY: Yeah.

RACHEL: And do they stand to potentially lose out there or are there cons to that?

LILY: Yeah. I mean, you lose out on the immediate [00:17:00] visibility and traffic that you could potentially benefit from right now with being cited in the AI platform.

So this is a big debate. It's like, you have to look at what that looks like for you right now. I think I mentioned in my presentation, most companies are seeing zero to zero point five percent of their overall traffic coming from ChatGPT. So if you rely on traffic to make money, you know, you need to think about whether that traffic is actually valuable or that's enough traffic to sustain the business.

Um, but I don't know if it's realistic. You know, I think there's more and more, like I hear more and more events happening with the AI companies and the publishing industry to kind of have these conversations about what's fair. But I would recommend that at the very least, companies are putting pressure, and I know with the EU and everything, putting a lot of pressure on, on these AI companies because at the end of the day, they're just gonna continue to take everybody's content until they, they can't.

ROB: Mm-hmm.

LILY: So we have to do something.

MAX: Yeah. I think to the question of like whether they'll be able to do it - I'm intrigued to see what happens with the sort of rise of publisher content marketplaces. There's sort of a bunch of different flavors of this, but the, effectively it's, [00:18:00] you know, building infrastructure that allows for compensation when, you know, a scrape occurs.

And I think that it's not an accident that a lot of publishers that have affiliate businesses have at least sniffed around them and, you know, maybe signaled interest in them. And so in a kind of mature competitive industry, like not a long history of collaboration, but I think that there's more infrastructure in place for a potential banding together than there might have been in other circumstances, for whatever that's worth.

To Lily's point, I mean, maybe the LLMs take the lead and say, "We're gonna try to figure this out." Uh, but you know, these specific circumstances are, I think, more conducive to something like that than we've seen historically.

ROB: Mm-hmm.

LILY: It's another thing, too, just to add to that like, if an AI, if one of the LLM companies or AI companies decided that we are gonna introduce some type of system where you're, you're getting compensated if we're using your content, people might make a decision to block the crawlers that are not doing that and to only allow that crawler, and that's gonna result in them having much better content and much better answers.

ROB: Mm-hmm.

LILY: So there is some leverage that these companies have to [00:19:00] potentially attract more companies allowing their content to be available to them.

ROB: We interviewed a affiliate publisher in the UK called uFurnish, they're a, a kind of, um, interior furnishings comparison website. And they had, like, an interesting take, I think, on LLM citations and, and wanting to be present there, less so for the referral traffic and more so for the brand equity because there is this sense as a user that if that, your name is showing up, you get that sense that this is a legitimate source, that that is a branding exercise, that you're being featured alongside likely very prominent, um, fellow publishers.

And that, I think, is maybe gets sometimes lost in the anxiety that the industry has around referral traffic. And to your kind of research, the fact that there is so little, uh, click-throughs that we're seeing from these-

LILY: Mm-hmm.

ROB: ...these kind of overviews.

LILY: Yeah.

ROB: Does that come up at all in conversations or?

MAX: Well, it's... I mean, I think it's, the, the perception of the opportunity is gonna differ, right? If I'm, uh, you know, a publisher [00:20:00] that routinely just, like, wins in SERPs- Yeah ... you know, day in and day out, you're a top five result every single time. You're gonna perceive the rise of AI search differently from a company that has no shot to win that SERP placement.

ROB: Mm.

MAX: And it does matter, right? Because broadly, this is a, it's a moving target, but, like, broadly, when you look at consumer perception of, like, AI responses, like, they're broadly perceived as trustworthy, right? Like, it's not 100%, but it's well over 50. And so if you are, you know, kind of a challenger brand or challenger publisher and you're, you can't kind of break through in what have been legacy kind of distribution surfaces, like, you probably do look at this a little bit differently.

ROB: Mm.

MAX: Where it's a chance to sort of be on the same plane as your, you know, bigger, better funded competitors, so.

ROB: Yeah.

MAX: It's definitely worth considering. Yeah, so that, I think that's absolutely a valid thing to think about.

AD: This podcast is brought to you by Awin.

Whatever your marketing challenge, whether it's reaching new audiences, stretching limited resources, or driving [00:21:00] stronger results, Awin helps you find the partners who can solve it.

Our platform connects brands with trusted partners who enhance your campaigns and only get paid when they deliver real outcomes.

If you're ready to make every channel stronger and every dollar work harder, visit awin.com and see how we can make your marketing do more.

ROB: Lily, I wanted to ask you, um, a large part of your session not only looked at some of the latest research that you were analyzing in this field, but provided some examples of what's working well as a, as a, a bit of a guide for the audience. I wondered if you could just elaborate on a few of those?

LILY: Yeah, for sure.

So I think the first thing to keep in mind is that, and I mentioned this a bit in my talk, but how large language models work, and in many cases what the language model is doing is it's pulling from top-ranking results in search engines, which I think a lot of people don't necessarily make that connection.

So I argue that one of the best things you can do is to be as visible as possible in search for the questions [00:22:00] that you care about.

ROB: Mm-hmm.

LILY: Literally, like, try to rank number one in search.

ROB: Mm.

LILY: Um, 'cause that there's a very strong correlation between what the LLMs are pulling in and what's ranking well in search.

But beyond that, you know, we're seeing more and more over time that there's certain types of content structures that really tend to get cited more. So for one, you know, I mentioned having, like, a TLDR or, like, a very concise answer to the question, ideally at the top of the page or close to the top of the page that really answers the question.

Um, there's certain types of pages that are doing really well. So we've seen, uh, there's a lot of debate in this, in this space right now about, like, a listicle. So like, you know, 10 list items that kind of rank certain products or whatever. Comparison pages that might compare two different companies to one another, um, question and answer pages and FAQ pages. So really anything that's, like, really easy for an agent or a large language model to come to the page and say, "We see a really clear answer to this question in one place that we can pull in really easily." You know, before you might have had the answer kind of dispersed throughout the content, but now having, like, one block that really answers the question.

So we're kind of just, like, tightening [00:23:00] things up in the content-

ROB: Mm.

LILY: ...so it's easier for them when they come in, they don't have to scan the entire page to get the point. They find one block that really answers the question.

ROB: Yeah. Yeah. And you, you mentioned at the very start there about, like, the correlation between kind of core SEO principles and what's then earning AI visibility.

What are the differences there?

LILY: Mm-hmm.

ROB: Where are the, the kind of, um, the things that maybe don't matter as much as a signal for Google Search, for instance, that do matter more for LLMs?

LILY: What the LLMs are doing is they're not only pulling from the top-ranking results. That's one part of the process.

ROB: Mm.

LILY: But they're trying to corroborate what they found with third-party information.

ROB: Mm-hmm.

LILY: So they're looking at what do people on Reddit say about this? What do people on LinkedIn say about it? What's being said on, like, Wikipedia and other user-generated content websites to kind of corroborate the answer.

So back linking and off-site metrics have always mattered a lot for SEO. I think we've been very, very focused on building links.

ROB: Mm-hmm.

LILY: Now we're much more focused on what are people saying about our brand? How many times is our brand mentioned across relevant sources that the LLMs are [00:24:00] looking at? So it becomes much more of, like, a PR activity.

ROB: Mm. Yeah. Yeah.

RACHEL: Maybe a question for both of you, but, um, this idea of maybe all the different digital channels that we have access to, how would you advise brands to think about maybe SEO and affiliate as complementary channels rather than... Historically, I think we've always maybe seen them as either very separate or even competing.

MAX: One of the things that's so interesting, and, and Lily touched on this a moment ago, is that, like, the picture or the, the sort of, you know, disparate sources that all these LLMs draw on to, to get their answers does kind of put brands that invest in affiliate in an interesting position, especially with those that have diversified their partner mix in such a way.

Like, if you're investing in creators and user-generated content and, um, you know, creators who are really active on lots of different social networks, like that potentially provides a boost to your visibility that, you know, really was not [00:25:00] considered at all when you were purely thinking about ranking in Google.

And so-

LILY: Mm-hmm.

MAX: ... you know, there's a way in which if you kind of put affiliate at the center of your thinking about distribution or visibility, like it allows you to think about this in a way that's more holistic than like might have been the legacy standard or norm. But-

ROB: Mm-hmm.

LILY: I think one thing where they are pretty different right now, or something that's different for affiliate sites now compared to where we were before with organic search, we know we mentioned before that Google cracked down on lots of different types of affiliate content.

Another big example of that was in 2024, Google launched a new spam policy called Site Reputation Abuse.

ROB: Mm-hmm.

LILY: And many companies that had a lot of affiliate content, Google just basically said, "You can have that content, but we're not gonna allow it to be indexable in our organic search results 'cause it's basically pay-to-play content."

ROB: Mm.

LILY: Where things are different now is it turns out that large language models are using a lot of that pay-to-play content to answer organic questions. I don't know how long that's gonna be the case-

ROB: Mm.

LILY: ...but it seems to be a really big lever that, you know, people can pull to basically influence what [00:26:00] appears in AI search answers.

So I think right now there's actually more opportunity for affiliate sites and for brands that wanna get on those kind of sponsored articles to actually use that tactic to influence what appears in AI search. Um, the only difference is we don't know how long that might be the case. There might be a day when they- OpenAI has its own site reputation abuse policy, but that might be a long time away from now.

ROB: I wanted to ask you, pick you up on that actually, because, um, Adrian, on the panel from Linkby earlier talked about the fact that this feels very much like a wild west gold rush-

LILY: Mm-hmm.

ROB: ... era right now. Advantages to the first movers. When you compare this, what's happening right now in the LLM space, particularly to the early days of SEO, and you've been working, was it sixteen years?

LILY: Sixteen years, yeah.

ROB: Yeah. Are there parallels here, or do you think it, it feels quite different?

LILY: There's absolutely parallels.

ROB: Yeah.

LILY: Yeah. I talk a lot about what I call the cycle of SEO.

ROB: Mm.

LILY: So basically the idea is there's some type of new opportunity or set of opportunities that SEOs discover now with GEOs and AEOs, AI search people discover.

Uh, they all kind of, uh, [00:27:00] shout from the mountaintops, right? They go on YouTube, they go to conferences, they share all these approaches. It becomes highly popularized, and then suddenly you have a giant footprint of the internet that's engaging in these types of tactics. Google hates this. They don't, they don't want people trying to manipulate their search algorithms. And I'm sure OpenAI is starting to see the same thing-

ROB: Mm.

LILY: ...with a lot of the spam in, in ChatGPT. Um, so they patch it, they introduce algorithm updates and penalties and things like this, and that tactic stops working, and then we're right back to square one where we need to find new tactics, right?

And I think right now we're absolutely in a gold rush where there's all kinds of ways that people are trying to trick, you know, AI responses and manipulate what, what's, what appears in there.

And I think what the, the companies are a bit behind in cleaning that up, but I think they're not done yet. I, I foresee within the next year probably some very big changes to how they show content and rank content and everything.

ROB: Yeah.

LILY: Uh, but definitely it feels a lot like SEO before they had a web spam team at Google.

ROB: Yeah.

MAX: And it feels, too, like the cycle that you laid out is probably gonna play out over a shorter time horizon, right?

LILY: Mm-hmm.

MAX: Because, you know, all of these platforms, even though [00:28:00] they've posted really eye-popping growth, like you had a slide about this, ChatGPT's user base is kind of plateauing a little bit.

LILY: Mm-hmm.

MAX: And like they're not gonna be able to reverse that if people get the impression every time they ask ChatGPT a question, the answer is going to be shaped by a bunch of like shady-

LILY: Yeah.

MAX: ...you know, kind of twenty-first century black hat garbage. And so-

ROB: I mean, I think that's what Anthropic's, like, brand campaign is basically premised on, right?

MAX: Yeah.

ROB: Is this idea of, like, oh, they're, they're, they're injecting ads now. It's a commercial outcome that's being manipulated. It's not really something you can trust.

MAX: That's right.

ROB: Whereas, you know, Claude, this is something that is actually much more genuine.

MAX: That's right, and they have to just-- They have to be, I think, very vigilant from a, a branding and kind of, you know, future growth perspective.

If they want that stuff to continue, they're gonna have to, you know, crack down on this stuff a lot more quickly.

ROB: Yeah.

MAX: Yeah.

LILY: And it's important to remember - I think it's a bit naive to assume that these companies won't be able to figure it out. I think they'll actually be able to figure it out faster than Google ever could.

ROB: Mm.

LILY: Right? Because they're AI companies. They're LLM companies. What LLMs are really good at is pattern matching, right?

ROB: Mm.

LILY: So if you start to see a bunch of people doing something that kind of [00:29:00] skews from, like, whatever training data they're focused on, they could see-- they can see spam patterns more easily.

ROB: Yeah.

LILY: So I think it's just a question of when they figure it out, and you're gonna wanna be the type of brand that wasn't engaging in those tactics because what we've seen with SEO over the years is you would've been better off not doing those things in the first place, even if it worked for six months.

ROB: Mm-hmm.

LILY: Where it lands you after those six months, it's often very, very hard to recover.

ROB: Yeah. Yeah. Yeah.

RACHEL: Um, you mentioned this as kind of like through your career the last sixteen years, this has been, like, the most exciting thing that has come into the space as an update or, like, as all these things are changing.

I guess, could you maybe elaborate on that?

LILY: For sure.

RACHEL: I know we've talked about it a little bit, but what is so exciting about this?

LILY: I think for people that work in our industry, the exciting thing about it is that SEOs are finally getting a seat at the table.

ROB: Mm.

RACHEL: Yeah.

LILY: Even though they're calling it GEO and calling it different things-

RACHEL: Yeah.

LILY: ...the reality is the most qualified people to help with this are SEO professionals 'cause ultimately this is, this is search. We're optimizing for, for search.

Historically, it's been very, very hard to get companies and clients to [00:30:00] take SEOs seriously. We've always had attribution challenges. We've always had, uh, you know, takes a long time to see results and everything, but suddenly everyone's like: "How do I show up in large language models?"

And it's like: "Well, you wanna make sure your search is buttoned up." So I think, it's exciting for us not only because we are able to do so many SEO workflows and processes so much more quickly with AI.

RACHEL: Yeah.

LILY: That part's very exciting. We're also able to build things we could never build before because now suddenly everybody can be a developer with things like Claude Code.

But just the fact that clients are enthusiastic about SEO for-- more enthusiastic than they've ever been in the sixteen years I've been doing this.

ROB: Yeah.

MAX: Well, and also to your point, like about the time horizon of seeing results, like it's not three to six months anymore, you know? If you seed things in the right place, you can start burbling up in responses within weeks, right?

LILY: Mm-hmm.

ROB: Mm.

MAX: Yeah. I mean, no matter what the-

LILY: Even days, yeah.

MAX: Yeah. And that's, you know, it's a good way to get hired.

LILY: For sure.

ROB: Yeah, yeah, yeah..

MAX: Give me thirty days.

ROB: Um, I wanted to ask you, Max, about, like, creator is very much your space and your expertise, [00:31:00] and I'm not sure if that came up so much necessarily in the, in the conversation today. But how are you thinking about, like, the implications of this for the, the creator economy and how it affects them?

MAX: So there's a w- way in which it's similar to affiliate in that it is very well-positioned, but there are some sort of misalignments that need to be sort of popped into joint for it to really, really, you know, matter and, and deliver.

But creators as a kind of like publisher class are potentially gonna have huge, um, role to play in GEO and affiliate as well because as we talked about on stage, like LLMs really, really value feedback from regular people.

ROB: Mm.

MAX: Um, they, they really value creator content, user-generated content. You see that in the responses, like the top 10 sites that are cited most frequently, the domains, half of them are just basically like repositories of creator content.

ROB: Yeah.

LILY: Mm-hmm.

MAX: And that's, you know, those sites have built their businesses on it, so that's really not going anywhere. The difference and the, the kind of misaligned pieces relate mostly to like creators [00:32:00] understanding the role that they play in it, and also the kinds of content they create, right?

ROB: Mm-hmm.

MAX: So like, you know, Lily went into this in great depth on stage, but like the stuff that an LLM really desperately wants and needs, right? This sort of like comparison information, specific specs, uh, context-specific information about products or services. That's not necessarily always going to wind up in a creator post, right?

ROB: Mm-hmm.

MAX: Like, the things that creators are really great at is being, you know, direct, winsome, charming, maybe shocking, maybe funny. That's all fine for getting noticed and shared on social media. It's not necessarily going to work quite as well for LLMs.

ROB: Mm-hmm.

MAX: But what I think you will start to see among a certain class of creators is kind of a development of a parallel content operation-

ROB: Mm-hmm.

MAX: ... where it's like, "I've got this one post that, like, there's a portion of it that's really funny that I'll put on TikTok."

LILY: Mm-hmm.

MAX: But there's a more kind of in-depth, sober, like, "Now I'm gonna take this apart-

ROB: Mm.

MAX: ...and show you, like, why I actually think this is great. And maybe I put that on YouTube. Maybe I, like, you know, cut it up into things that I distribute on LinkedIn, whatever." And so [00:33:00] you have this ability to really give the LLMs and the brand partners a chance to, you know, appear in really important contexts.

But that's gonna take time.

ROB: Mm-hmm.

MAX: And so is the compensation piece of this, too, right? Like, most creators I know, you know, to the extent that they, they know what AI visibility is, they have very little kind of like means to sort of say - or gather up the data needed to say, "My content's actually really valuable to you partners and you p- prospective partners, so let's talk."

But again, we're so early that, like, I think in a couple, three years we're gonna ... The, the creators that get that and-

ROB: Mm-hmm.

MAX: ... kind of embrace the upside there are gonna really benefit.

ROB: Yeah, yeah.

One thing I was really interested in is, particularly on the creator side, is that that field has increasingly become - the medium by default is video and it's visual.

MAX: Yep.

ROB: Whereas the way that the LLM seem to give primacy to the content that they're ingesting is primarily text-based and copy-based. Is that changing? Uh, is video becoming a more valuable asset for them to digest, and does it require [00:34:00] transcripting, or is, is ... Can they take the visuals themselves?

LILY: Yeah, it depends on the LLM, but, like, Google, for one, prides itself on being what they call mult- multimodal.

ROB: Mm-hmm.

LILY: So it, it's able to ingest all types of different content formats, and I think this is increasingly becoming true with LLMs. So, you know, we see Google, I mean, obviously Google owns YouTube, so that's important to point out, but Google really seems to be prioritizing YouTube in AI-generated responses, and I think they just get so much of that great authentic kind of human conversation within the video that it often answers a lot of the questions people are asking about.

ROB: Mm-hmm.

LILY: Um, but I do think video, audio, images, all of this is a new playing field because LLMs are able to understand a lot of that content much better. But I would still recommend, you know, make sure if you're uploading videos and things to your site that you're looking at transcripts. I think YouTube does it automatically, but not-

ROB: Mm.

LILY: ...every video provider does, but that transcript is really gonna help kind of solidify what they know about the video.

ROB: Yeah, yeah.

MAX: Yeah, and I feel like that Lily's point, like the transcription doesn't get you all the way there, like, but it, it gets you pretty close.

ROB: Mm-hmm.

MAX: And so as long as you're [00:35:00] investing in that and there are tools at hand, whether it's- The native stuff on, on YouTube or elsewhere, like that's a reasonably solvable problem.

That's one that everyone should be looking at, right? Because I think that the LLMs, again, because they really are hyper-focused on sources of kind of authentic human feedback and, and information, are gonna prioritize those platforms, which are all, you know, trying to incorporate more video, whether they're video native like YouTube or Instagram, which is, you know, indexable in Google.

We're just gonna, you know, video is, is eating all of our media time, and so it's, it's gonna be a, a bigger component moving forward.

ROB: Yeah.

RACHEL: I was gonna ask a really quick question. You've both touched on like geotactics as well, and then Lily, I know you mentioned that there are actually tactics there that could very much have a knock-on or negative impact to SEO strategy and then ultimately AI visibility.

Can you expand on that a little bit just for people that weren't in the session?

LILY: So this is actually probably the place that I've put most of my focus over the last few months because again, there's so many people entering into the conversation right [00:36:00] now that might have very limited SEO experience, and if we think about that cycle, if you jump on the bandwagon really early without understanding what happens throughout that cycle where a lot of popular tactics become demoted or penalized over time, I think right now we're in this phase where...

So I give the example of what I call self-promotional listicles. So you can put a blog on your website that says, "Who are the best companies in, in our category?" And then you list your competitors at two through ten, but number one is your own company. You talk about how great you are, and then maybe you talk about your competitors being really bad.

Turns out this actually influences LLMs in a really big way, and a lot of people have discovered this. So a lot of people have made not only one of these articles, but five hundred or in some cases two thousand such articles on their website. And the reality is eventually Google usually catches onto this.

It's, it's a version of spam according to Google, and I think the AI providers are starting to figure this out. So I think right now we're in this space where a lot of people have found these exploits and kind of these hacks to try to influence what appears in LLMs, and then they've used AI to kind of accelerate it, create a lot of that [00:37:00] content.

Um, and I've been trying to like raise a flag because that's my space in, in the SEO space, I've always been the one to help with algorithm updates and everything. I anticipate based on what Google's done for 20-plus years, we will see a massive, massive algorithm update similar to the Panda, the Penguin, the Helpful Content. We've had so many of these in Google's history where they're gonna go against exploitative GEO tactics.

RACHEL: Mm-hmm.

LILY: And you're not gonna wanna be wiped out from Google search with all this stuff.

ROB: Mm.

LILY: So you just have to be very careful.

ROB: Yeah.

MAX: I'd be curious to see too, you, you brought this up in your presentation, but like whether there's a version of a future state where the FTC is also kind of a really-

LILY: For sure.

MAX: ...big piece of this too like, the potential now exists to really take kind of, you know, anti-competitor GEO to a whole new level.

ROB: Mm-hmm.

MAX: Like it would be very easy to sort of pay a legion of creators to go onto YouTube and talk about how crappy your competitor's products and services are and, you know, boost yourself in the process.

ROB: Mm-hmm.

MAX: Until these systems get [00:38:00] smarter, like I think there'll be a little bit of time where they figure it out. But like potentially if the, you know, FTC wants to take a more proactive stance, they could start just like bringing the hammer down and saying, "This is gonna cost you 200 grand. This is gonna cost you half a million dollars because, you know, today that is illegal."

And so, uh, I think there's a chance where they could also sort of provide a chilling effect on-

ROB: Yeah.

MAX: ... anything that, you know, might be perceived as not great.

ROB: I think it's as much about the speed with which this stuff can happen as it is the scale. And this is what these tools have kind of enabled now, is that anybody really can suddenly produce or adopt a tactic and it goes from day zero to a reality really quickly, but at a volume and a size that maybe historically it would've taken a lot longer to do.

And with regulatory or trade bodies like the FTC, do they have the speed to be able to anticipate or react to this stuff in time for it to, to prevent it becoming like a big, big issue?

MAX: Probably not today.

ROB: No.

MAX: But you know.

ROB: But they have the, they have maybe access to the same tools themselves, right? So.

LILY: Yeah.

MAX: That's [00:39:00] right.

RACHEL: Thanks for listening to this episode of Awin-Win Marketing Podcast.

ROB: And we will be back in two weeks time with tech partner, Upsellit, and US-based performance agency, eAccountable.

RACHEL: Until then, thanks again for listening to Awin-Win Marketing Podcast where we show you how partner marketing always offers a win-win.

ROB: Goodbye.

RACHEL: Bye.