Episode 65

Inside the Seismic Shift from SEO to Answer Engine Optimization

· 40:54 · Host: Todd Juneau & Kelly McGuire

Todd and Kelly sit down with Sami Akkawi, founder of Petra Labs, to explore the seismic shift from traditional SEO to Answer Engine Optimization (AEO).

Play this episode

Chapters

  1. 00:00:00 — The Trust Revolution in AI Search
  2. 00:01:00 — From Wall Street to AI: How a Morgan Stanley Banker Became an Answer Engine Expert
  3. 00:02:48 — From SEO to AEO: Defining Answer Engine Optimization
  4. 00:04:00 — Why AI Traffic Converts 7-15x Better Than Traditional Search
  5. 00:06:58 — The Fine Line Between Optimization and Manipulation
  6. 00:11:43 — How AI Models Assign Authority Differently Than Google
  7. 00:15:00 — Personalization at Scale: The Future of AI-Driven Recommendations
  8. 00:18:20 — Can Small Brands Compete Against Giants in AI Search?
  9. 00:22:43 — User Generated Content: Reddit, TikTok, and the 8-Month Rule
  10. 00:30:00 — Healthcare vs CPG: Industry-Specific AEO Strategies
  11. 00:35:42 — The Danger of Short-Term Hacks and Long-Term Consequences
  12. 00:41:18 — Content is Still King: Why Authentic Storytelling Matters More Than Ever
  13. 00:45:30 — The $399/Month Easy Button Myth: Why Real Work Still Wins

Read the full transcript ↓

Show notes

Todd and Kelly sit down with Sami Akkawi, founder of Petra Labs, to explore the seismic shift from traditional SEO to Answer Engine Optimization (AEO).

Sami, who made the leap from Morgan Stanley investment banking to AI-driven marketing, reveals why brands must fundamentally rethink their digital strategy as AI models like ChatGPT become the new gatekeepers of consumer trust.

The conversation dives deep into the tension between optimization and manipulation, exploring whether AI search truly delivers on its promise of truth-seeking or simply creates new opportunities for those with the deepest pockets.

From Reddit's hidden algorithms to the death of programmatic content, this episode challenges marketing executives and brand managers to confront the truths about authenticity, scalability, and the future of consumer engagement.

Sami shares compelling data on why AI traffic converts 7-15 times better than traditional sources, how smaller brands can compete against giants like Amazon, and why the SEO playbook of the last 20 years might be more relevant—and more obsolete—than anyone expected.

Chapters:

  1. (00:00:00) - The Trust Revolution in AI Search
  2. (00:01:00) - From Wall Street to AI: How a Morgan Stanley Banker Became an Answer Engine Expert
  3. (00:02:48) - From SEO to AEO: Defining Answer Engine Optimization
  4. (00:04:00) - Why AI Traffic Converts 7-15x Better Than Traditional Search
  5. (00:06:58) - The Fine Line Between Optimization and Manipulation
  6. (00:11:43) - How AI Models Assign Authority Differently Than Google
  7. (00:15:00) - Personalization at Scale: The Future of AI-Driven Recommendations
  8. (00:18:20) - Can Small Brands Compete Against Giants in AI Search?
  9. (00:22:43) - User Generated Content: Reddit, TikTok, and the 8-Month Rule
  10. (00:30:00) - Healthcare vs CPG: Industry-Specific AEO Strategies
  11. (00:35:42) - The Danger of Short-Term Hacks and Long-Term Consequences
  12. (00:41:18) - Content is Still King: Why Authentic Storytelling Matters More Than Ever
  13. (00:45:30) - The $399/Month Easy Button Myth: Why Real Work Still Wins

Links and Resources:

  1. Petra Labs Website
  2. Sami Akkawi on LinkedIn
  3. Kelly Maguire on LinkedIn
  4. Todd Juneau on LinkedIn
  5. Vuja Dé Digital

Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify, Apple Podcasts and YouTube to leave us a review!

Contrary to Popular Opinion is produced by Amplafy Media. Podcast production services · Client results

Full transcript

The complete conversation, in order. 8,281 words.

Teaser [00:00:00] People trust Cha Bt more than their coworkers. Like they'll fact check their friends and family by saying, wait, let's ask Cha Bt, right? Like, is this actually true? And you have such a trusted source of information that is being delivered at scale in a way that has never existed before. And so why does it matter? It matters because when you can deliver that trusted recommendation at scale, your own customers who you actually do care about will listen to it.

Intro Welcome to Contrary to Popular Opinion, a podcast hosted by Bo Jee Digital. In each episode, we bring you provocative topics and novel perspectives around leadership, business, and marketing that defy the norms, because exponential growth demands unconventional thinking and visionary action. So open your mind and let's get into this week's episode.

Kelly Hey everyone. Kelly McGuire with Bougie Day Digital. Excited to talk to you, today's guest about today's topic.

Todd Hey everybody, this is Todd Kelly's partner. Happy Monday. It's the week before the Christmas holiday, and I remember my early days, Kelly [00:01:00] working on the non-agency side and agency's all shut down for two or three weeks, especially the ones in New York where Sammy is.

Todd And I was like, man, that must be a great job to have. And now that I have an agency we can't shut down during those two or three weeks. So apparently that holiday is no longer for the agency business. today we've got Sam Akai, who is in New York, where it's colder than it is in San Diego, where Kelly and I are, sorry, Sammy.

Todd Um, Sammy's gonna talk to us about, well, I would say SEO, but it's no longer SEO. It might be a EO, it might be GEO. We're not really sure what to call it. Sammy's gonna help us understand that a little bit. Sammy comes to us from, well, I guess you were involved in investment banking with Morgan Stanley and private equity, and then somehow you decided to get into advertising and ai, which I think is a cool leap.

Todd so Sammy Akai from Petra labs.ai welcome. And, we're excited to talk.

Sami Thank you. Thank you Todd and Kelly. I'm very excited to be here. [00:02:00] talk to two of the smartest people I know in the marketing space in general, and hopefully share some opinions that are a big contrary. but I'll try my best.

Todd Yeah. Well, we like to, like I said, we like to pick on both sides of the topic and the debate, and not, acquiesce to either, but, make it, a challenge for, for people out there to hear both sides. so this is all new, this, this idea of, and I'm gonna have you explain this, a, e, o and GEO. So maybe kick us off by just defining what those two acronyms stand for and sort of where your journey is at going into this space and,

Todd Why does this matter to you now?

Sami Of course. Um, so A-E-O-G-E-O, ai, LLM, eeo, there's a ton floating around here and I think there's a lot of really smart people who are trying to figure out what the best word is to use. I think generally that the intent is that brands have a new custom. A new customer that is reading their content may be a bit differently than the customers they're used to.

Sami May be a bit differently than some of the machines that they're used [00:03:00] to reading it as well. And as with all new customers, there's a question of how do you make them happy and what's the impact to your top line? And so at the end of the day, answer. Engine optimization, generative engine optimization, you know?

Sami It is, how do you sell to that new customer, artificial intelligence that is now trying to build an idea and a perception of your brand, that is a bit more multidimensional than a traditional rank. Right. They have opinions, they have like strengths and weaknesses. They have reviews that they read.

Sami I'll pause there.

Todd I, I wonder where we're gonna land. do you have any sense of like, which acronym is gonna stick? I'm just curious what you're hearing in your field.

Sami I think GEO used to be more popular, but in the past two months, a EOI think is taking the lead. I think people have put a couple surveys on LinkedIn.

Sami A EO is taking the lead right now.

Todd All right, so we're gonna go with answer engine optimization as the new customer for brands. long forever. It used to be Google and SEO and a Google results page. But now, now we're talking about answer engine optimization. Why does this new customer [00:04:00] matter now?

Todd what's the implication of the impact in these very early stages of why does this matter?

Sami So I, I think the best way to describe it is to zoom back five or six years and try to think about what the most impactful channel was for a company that was trying to grow. And as business owners yourselves, you probably know, the best channel you can rely on for growth is warm referrals.

Sami Yeah, somebody with who really likes you, who tells a friend who trusts them, like, Hey, you guys should definitely go check out Todd. And because they trust that friend so much, so much of the barrier has already been removed. And it makes sense why that's the case today. For the first time ever, you have that same feeling of trust that you can deliver to a customer, but now at scale to millions of people running billions of queries a day on AI systems, I think it's funny because I look at how like on a really, really small scale, like my own dad.[00:05:00]

Sami He trusts Cha CBT more than his own doctor. He'll come back from the doctor and and say, Hey, I got prescribed X medication. Like what are the side effects? Should I get a second opinion? What do I do? People trust Cha Bt more than their coworkers. Like they'll fact check their friends and family by saying, wait, let's ask Cha Bt, right?

Sami Like, is this actually true? And you have such a trusted source of information that is being delivered at scale in a way that has never existed before. And so why does it matter? It matters because when you can deliver that trusted recommendation at scale, your own customers who you actually do care about will listen to it.

Sami And we can get into the details and the numbers and you know why this matters, but ultimately, that's why a lot of brands see AI traffic converting at seven to 10 to 15 times higher rates than any other traffic source they have. 'cause by the time they leave that chat window, they've compared options.

Sami They've received a recommendation that they can trust from a source that they trust more than known people in their lives, and they convert better. And so ultimately, that's why I think it's so critically important. It [00:06:00] might not be live or die for every brand in the world, right? You're selling to government, you have nine month RFP cycles.

Sami Maybe not where you put 20% of your budget, but generally there's a lot of companies for which this is a really exciting opportunity.

Todd Interesting. Kelly, do you, do you trust chat GPT to make a referral to you?

Kelly So I, I think it's really interesting that you use the word trust there because that, that really is where my thought process goes as I want to discuss this, right?

Kelly Is. What is trust and how has that evolved in the way that it is? And it makes me think of the timeline of SEO. It makes me think of the intent of SEO, and it really makes me think of the, what I've always struggled with in the nomenclature around search engine optimization and the idea that there's a pretty fine line between optimization and manipulation.

Kelly When you think about the opposing objectives that [00:07:00] exist in the environment, right? So in an altruistic way, if you go back to the beginning of the concept of a search engine, right? The idea was to the organization of information and the ability for people to find what they were looking for easier.

Kelly And Google comes to the table and they say, okay, you know, in all of this disparate information, all these sources in there, how do we find a way to create confidence or trust that if you are looking for a teddy bear, what is the best result for finding a teddy bear? And they tapped into this idea of well link authority, right?

Kelly If a lot of, if a lot of sources are linking to teddy bear and associating the word teddy bear with this. Source of information that conveys a lot of authority and trust that a lot of people have collectively as a group decided this is a great result for the term teddy bear. And then people quickly realize that you could manipulate that by saying, well, I can just create a bunch of links around the term teddy bear and say my [00:08:00] result is the best.

Kelly And you immediately get into this opposition of the altruistic intent of giving people what is most valuable to what they need, the moment versus the commercialization of. If you have an interest, I'm gonna point you in my direction 'cause I can make money or sell something or do something in that. And then that inherently starts to branch out this conflict that exists.

Kelly And I think it's expanded even further now that you've gone from sort of search and link authority to content authority to social signals, and now into all of the things that feed in answer. Because that's the interesting part when you go into into AI is at its core, the concept is AI is meant to be helpful, right?

Kelly That is generally the mandate of AI is to be helpful all the way to the point where it creates this whole other world that we won't go down that rabbit hole. But if you're thinking about, okay, it's meant to be helpful, I'm asking chat GBT or perplexity, or whatever it is I'm asking for, for an answer, and it is going and using the information available to it to form the best answer back to me to be as [00:09:00] helpful as possible to me.

Kelly So now you start to look at again, is the information backing that answer up agnostically looking at what is truly the best answer, or is it being manipulated or influenced into guiding me into an answer that is beneficial to someone else's sort of biased. I should be the best result. 'cause I want you to buy my product over it.

Kelly and I set the stage for that because that's the really interesting conversation to me as we've kind of gone through, is what are we trying to achieve? Are we trying to achieve the truly best answer? Are we trying to achieve the answer that best serves the people who are paying the most money to optimize to it?

Kelly And what is the line between optimization and manipulation? And then the last thing that I would say to that is. What I think AI has shifted from the old SEO is you used to be able to kind of, you had to build up that authority over time. Even if you were quote unquote manipulating it, you had to convince reputable sites to link to you.

Kelly They had to last for a long time. They had to be [00:10:00] in there. I would say that in this new world, things move really fast, and so the conversation shifts really quickly, right? So you can't even create that trust and authority over a long period of time. I would argue. You could be shifted out by, again, a PR crisis or something changing the narrative.

Kelly All of a sudden you have to combat and push back and, you know, offset misinformation. And I dunno, I'm creating a lot of problems in my head on this, but like, this is what is so compelling to, if we are trying to give people the best outcomes and results. How does the idea of optimization slash manipulation play into the achievement of that altruistic goal?

Todd The question you're asking Sammy, to be clear is we've known for years through SEO since it started, you could game Google and people have done it and still do it to this day. So does that same environment exist with chat, GBT or any, or perplexity or Gemini or any of these others are, are we already at a stage where trust is eroded because it's really smart people have figured out how to [00:11:00] game it.

Sami Okay. Amazing question, and I love this topic because. It is a real trade off that we have to make every day with some customers who say, Hey, I've heard of this tactic working. Like, what if we just inject this text into our HTML and have a page for agents, right? I think ultimately the goal of these models and the one that's going to win is gonna be the one that has the best truth seeking answer.

Sami And there are a lot of strategies that may be working in the short term today, but that we try and completely ignore and optimize for long-term success, agnostic of model changes. Right. What do I mean by that? When Chachi BT five came out, it was super interesting for us because there were so many websites before Chachi BT five, using AI generated content at scale to try and game the system.

Sami And a ton of that content looked like top 10 best, top 10, best X, Y, and Z. And like there were websites with a hundreds of top 10 best, you know, software that were [00:12:00] being just mass produced with AI at a huge scale each month. But when chat BT five came out, it deprioritizes some of that content and frankly, chat Ccbt five really deprioritized at scale content written by chat BT four and what got up prioritized, right human written content.

Sami And when chat BT six comes out, there may have been strategies that worked in the short run now that aren't gonna work then. And so ultimately. With Google, you had a constrained set of data because of just how the algorithm works to assign authority, right? Domain authorities assigned at the domain level attached to your actual brand domain and backlinks contribute to that authority of the domain.

Sami Google has also tried to incorporate things like organic search traffic to help diminish the success of some of those quick win tactics. They've also introduced legacy scores over time, right? If you ranked really well last year. It improves your chance to rank this year. What chat CBT tries to do is a similar game of assigning authority, but a bit differently and in a way that I think [00:13:00] will be more truth seeking in the long run Chat.

Sami CBT checks its sources in a way that Google doesn't. It looks at what you say about yourself on your own website. It then checks what other third party publications say about you, and it doesn't just pull from Wall Street Journal and Forbes. It finds really small, like. Two person owned blogs with 300 organic visitors a month that it deems high authority that might be buried on the 10th page of Google.

Sami And it also reads those, and then it checks what real users say on Reddit, even if there are absolutely no backlinks back to your domain. And then it checks YouTube, right? Like what are creators and real people who use your software saying? And it uses all of those different boxes to assign authority to your brand in a way that goes beyond just the actual URL that you're associated with.

Sami And I think there's really cool data that we have on this of like, what do we see to SEO and a EO and brands migrate the domains? And we can talk about that I think in a sec. I, I do wanna give you guys a chance to react like the major answer here, which is, you know, it's a big problem we see today. we [00:14:00] believe those strategies will die in the long run and at the winner of this model race, we'll be able to truth seek in OA that Google never could.

Sami And it's why we focus so much of our efforts on sub 15, sub 10 pieces of content each month that are human written case study based. And we try and focus them. We try and like work with our clients to be like, where are you truly differentiated? What's the segment that you should be winning in? Like, let's go tackle that one first.

Kelly Yeah. So it makes me think of two really. Probably deep question, so I'll ask them in order versus at the same time. Right. again, another word that you said that stands out and kind of brings my head down a path, right, is truth, right? But one of the interesting things that I think AI and the speed of AI and the power of it brings in is whose truth, right?

Kelly Is, you know, I think SEO used to have to be a more universal truth, right? Let you know. Let's take it of an example. If I was searching for what is the best dog food? The answer for you in New York with your type of dog in your [00:15:00] environment and what matters to you as a consumer might be a very different truth than me in Southern California with a different type of dog and a different type of priority on, on how things eat.

Kelly You know how my, my animal wants to eat? And so the ability to kind of assign a simple truth to a broad term, I think becomes really interesting. And then you take it further with the customization of ai, very much being a learning system about how I engage with chat. GBT is different. then, how you en engage with chat GBT and it's intended to learn.

Kelly My nuances and what matters to me in, in there. And so how customized and segmented and personalized do you think the idea of, I'm just gonna call it general engine optimization, can really get, I'll stop there because I feel like there, you could build on that answer quite a bit.

Sami Yeah. so Sam Ulman once did like a, a speech or a podcast or whatever.

Sami He says, when you look at our user base, if you look at [00:16:00] people that started using chat GT when they were like 30 or 40, they use it for tasks like that they would use Google for. They help it clean up some of their emails, and they'll maybe do like general search that's a bit layered deeper than Google with more, keywords in the, in the prompt.

Sami If you look at people who started using it when they were in their, you know, late twenties. Early thirties, they'll have it think a bit more, right? they'll give it a problem. They can assume that it can give them a, a series of steps to reason through the problem. And they won't just treat it like search.

Sami So it'll be it like a partner that helps 'em think. And then Sam said, if you ask people who started using chat ccbt when they were in high school, what they search on chat, CBT, they would be mortified to even tell you. and it goes to show that when people talk and we buy user search data, so we see this all the time, right?

Sami Like really long prompts. How do I text this girl back? Here's what she said to me, right? Like the most detailed and personal parts of people's lives, especially for younger generations who are just forced to use AI tools because they're in school and they don't wanna do their homework. And so it's the [00:17:00] center of their life.

Sami It's always open. It's the first app they go to. The personalization that we'll be able to see, I think is to an extent, and it honestly scares me, right? Like I already think consumerism. Like goes pretty far. Like, I don't do well when a really good Instagram ad targets me that like, knows that, that I would like it.

Sami Like I get that all the time, right? Like Ariix knows I'm a vulnerable customer and they'll sell a ski jacket to me and they'll, they'll keep pummeling me with ads, right? The fact that they can go a hundred times beyond that, I think is an insanely massive opportunity. Right? And it's why we see retailers like Amazon and Walmart shifting the algorithmic responsibility of surfacing products.

Sami From their own algorithms to the models. And the reason it's so interesting is because the Walmart and Amazon algorithms are not just like Google. At least they say we have some sort of desire to give a customer the right answer. Amazon and Walmart, they optimize for one thing, maximum order value at checkout.[00:18:00]

Sami They have spent decades perfecting an algorithm that maximizes order value at checkout, and they are now saying we can't compete. In order to maximize order value at checkout, we are gonna outsource that algorithm responsibility to a model that has so much more context outside of the bounds of our system than we can ever have.

Sami And it's a, it's, tricky for large brands who have relied on general authority across the long tail. It's a huge opportunity. For small brands that might win individual verticals. And it's really funny for me because we work with huge customers with hundreds of millions in revenue, and one of our value props to them is, Hey, for every single like segment you wanna track, we will find all of the little startups that are like eating away your market share.

Sami Like long tail by long tail. And then for the smaller companies you work with, it's like. What is the one segment that you can go in and dominate and show up above Nike, above Brooks, above Hoka. Like what's that one segment for you? And so it's two, you know, [00:19:00] opposite sides of the pendulum, but really interesting.

Kelly I'm so glad you said that at the end 'cause it's exactly the next question that I wanted to lead, which is, given how much customization, how much personalization, how much differentiation, how quickly things can shift, How do we avoid, or is it avoidable to not have this massive disparity gap in this giant homogenization of all of the answers going towards the top ended players who can afford to flood the market with and cover every vertical and every segment.

Kelly And how do you not get the Amazons and the Googles basically saying. I have the resources and the means to control the narrative because I can flood the market. I can pay for the Sammy's of the world to go manipulate the engine optimizations and flood the market and all that. Whereas you have mom and pop brand who just does not have that resource and that ability to pay for someone to constantly be looking at it, constantly updating and constantly monitoring and can't do.

Kelly 30 [00:20:00] different versions of the same piece of content that's personalized to a different, you know, gender and age and, and, and all of that. Like, it gets a little bit worrisome that it is so favorable to the giants that it kind of drowns out the ability for sort of that startup and that really quality element to break through and not just get sort of buried even more so than historically has always happened.

Sami It's a good question and I think at the core, a company that has more dollars to throw at something is gonna have an edge. Whether that's with paid ads, whether that's with, you know, backlink purchases, whether it's with, just better design and a better website and better conversion edge checkout because they can pour money into a design agency.

Sami That being said, I think there are a few reasons that make Chacha PT slightly more favorable to smaller companies than the previous channels that we've seen. The first is that these are probabilistic models. Which means that even if Amazon is the number one answer, it will never be the number one answer in a hundred [00:21:00] percent of questions.

Sami And what that means is chat, CBT over time, especially as it moves more into shopping, is going to look at user feedback as a core input into what brands it should surface. And because it's a non probabilistic model, some of those companies that may be stuck on page two or page three of Google are going to have in two to three to 4% of answers.

Sami They're shocked at being the number one, spot. And if those brands convert well, when they do show up, it'll improve the way that they can then show up later over time. And to think the nature of these models just being probabilistic in general and never giving the same answer is one way that smaller brands will have an edge in an AI search future relative to the future that came before this.

Sami Right? Um, that's one thing. The other thing is that there are genuinely too many permutations. I think the prompts are too long. Like programmatic, SEO was already a challenge for really big companies to [00:22:00] do at scale. they were already building out thousands and thousands and thousands of pages, and I believe that AI content's gonna lose in the long run.

Sami And we can talk about why that's the case. And so it has to be human written, but generally there are too many permutations to optimize for. And I believe that smaller companies can and should focus on the niche where they believe they will convert customers better than the bigger players ever can. And they will build out a vertical for themselves that they can dominate in a way that was more difficult to do before now.

Sami Will they ever have a level playing field with somebody who can afford to pay $60,000 a month to optimize? Probably not, but I do think there's a little bit more of an edge in something we may have seen before.

Kelly So I think that becomes really interesting to me. 'cause I think now that dovetails into a conversation that Todd and I have had a lot, which is this idea that.

Kelly Brand really matters, and building a quality company really matters. You know, we went through a, a long period where it felt like things like fast fashion and fast brands and sort of, you could build up something and sell something really [00:23:00] fast when meta was really amazing and matching things up and, and there was just this kind of golden age of like.

Kelly E-commerce and direct to consumer and things where a lot of that fell by the wayside. And we've definitely felt a pulling back from that. But the way that you described the idea that if AI is truly gonna prioritize the reflection of, human reality and experience and, authenticity that exists and the sources of your Reddits and your YouTubes and the things that people go through, it almost feels to me like the idea of grassroots becomes incredibly.

Kelly Important for these smaller brands to be able to break through and be seen, because they're gonna have to build momentum off of that kind of grassroots, what they've cultivated, what they've created of their own audience, and turning that into as much digital option and, and activity as you possibly can.

Kelly And then building for that, it almost feels you can't sort of synthetically or arbitrarily create momentum just because you're serving ads or just because you're [00:24:00] doing something. It has to be. Ads has to be what it always should have been, which is amplifying an authentic experience, an authentic audience that then says, Hey, this group really loves it and they're endorsing it, and now we just need more people to know it exists.

Kelly And to sort of hear that testimonial and that experience. And so that's really interesting to me that almost shifts back to the only way for a, a smaller player to compete is by creating more authentic and genuine. Elements and, and thinking about how that becomes the strategy of blowing that up and amplifying that strategically

Sami a thousand percent.

Sami And I think one of the coolest parts about how AI search is different than SEO is the huge prevalence of user generated content that never links back to your brand. Right. models are now being trained on reels, on TikTok, on LinkedIn posts, right, on Reddit. And one of my favorite questions to get like from customers because it's like a, it's a really, this is maybe how we [00:25:00] get into misconception land.

Sami Like what's a common belief? It's like, oh, we should just go post on Reddit. Like everyone says, Reddit's important. Let's just go flood red. most people actually don't know how the models treat Reddit and how it works. And while it may not be perfect today, I wanna share an example of why I think they're moving in the direction of truth seeking.

Sami Right. The first really interesting data point that we see is that up votes aren't correlated at all with the rich Reddit threads get cited. And so you might have just a happy customer who's posting a glowing review, an account that doesn't have a lot of history, but that's talked about the same topic a few times.

Sami It is getting picked up over a post that a really large company has paid to get like a thousand up votes. Right. Like that's one really cool example. Another really cool example here is the age of the Reddit threads that we're seeing in our dataset. When we look at Reddit threads that get pulled for our customers, the vast majority are eight to 12 months old.

Sami The models are intentionally deprioritizing threads in the short term. They're waiting for moderators to filter them out. They're waiting for users to comment real people and be like, oh, another spam, [00:26:00] another ad, right? So that they can deprioritize it and they are letting the market stabilize before taking in that Reddit content.

Sami And it's a bit harder to then find real sentiment in real time, but that's why I look at reels and that's why I look at TikTok and that's why I look at LinkedIn. But user generated content overall is so important and I think difficult to gain because the previous way of trying, and this is maybe like one of those examples of.

Sami You can still do better if you have more money, but it's harder to do better. Previously, you could buy, you know, likes and farm accounts to manufacture views on TikTok and manufacture views on LinkedIn, right? Like there are services that provide that, as a service. But in a world, reviews are uncorrelated and the models are relying on more unknown factors to assign authority.

Sami Right? It's like quite interesting. And one of the coolest products that we have for our customers today is giving them a way to measure, the revenue impact of an individual influencer. They'll have YouTube campaigns, even have an enterprise customer that has YouTube campaigns launching for 10 different product lines.

Sami And what we're able to say is, [00:27:00] oh, this campaign by this user got picked up in higher detail. Like they're outperforming the other influencers in your channel, despite having way less followers, despite having way more views now, one that gives them pricing information that they can take to negotiations, right?

Sami And it's now like a more one-sided pricing negotiation. but two, it's quite interesting because it's just more and more reassuring evidence that we believe these models are moving to a world where you can still gain the system with enough money, but. They are generally already at a place where it feels more truthful, and I think that's why people trust it, than Google in traditional search.

Kelly Do you think that that inherently carries risk? That if you're too focused on a short term solution now that you know, you said it's prioritizing things that are maybe eight or 10 months old or, or things like that, that you can see maybe an, an uplift now, but you may get punished or knocked down for it in six months or a year because.

Kelly It starts to recognize an inauthentic pattern or something that [00:28:00] definitely comes through as a paid strategy versus a, genuine strategy. It's, it's interesting to think with how, fast everything's moving. Decisions you make now could really have kind of long lasting effects as the models continue to get smarter.

Kelly Can. Figure out sentiment, intent, authenticity, credibility, there's almost no place to hide element to it if you're not careful and thinking both short and now, a lot more long term of what could this look like, you know, a year from now.

Sami 10000%. Right? And and that's a huge thing that we try to avoid.

Sami The other consideration there, Kelly, is that right now track CBT is dominant. That may not be the answer in the future. And companies are okay taking risks today to improve their search presence in track CBT, even if that might downgrade them in Gemini, even if that might downgrade them in rock or plexity.

Sami 'cause each one thinks differently. But good authentic content, strong UGC presence. This performs well across all of those models, and it's one of the reason that we try and focus on building a base of strong human [00:29:00] written, high quality content. Case studies are our favorite thing. We interview our customers customers around case studies that in the long run are gonna help you across all of these models.

Sami And in a world where there's three or four dominant model providers that all have good market share, some of these short-term hacks that work for chat, CBT might downgrade you in other arenas. And so it's just another reason why we believe in avoiding some of these tricks that people might see in the short term.

Sami The other reason is you might be in an industry where those tricks don't apply to you. We didn't really talk about this in a ton of detail, but one of the other big differences of AI search versus Google search is how unique it assigns authority signals based on industry. And if you're selling any sort of consumer healthcare, any sort of like supplement, any sort of, you know, wearable device for your health.

Sami And you start taking some of these strategies that you see online, which is like, make a top 10 article, right? Like, add an FAQ to every single page. Go and flood Reddit. add, you know, all these things that feel like they are one size fits all. You might actually risk damaging your brand in a way that.

Sami [00:30:00] You wouldn't if you were in CPG. Right? Healthcare queries are intentionally treated differently by the model makers, and optimizing for them looks differently than optimizing for SaaS. And so there's a, you know, a third level of consideration of if you are chasing short term hacks that feel like they're working for other people, you might really hurt yourself in the short term.

Sami And then the final, like just cherry on. Is the customers we talk to who get recommendations for AI search and then they see their Google traffic go from 50,000 a month to zero 'cause they get blacklisted. And as cool as I think AI is in the future, like it is not fun. It is not fun to get blacklisted from Google.

Sami And we have talked to brands that fuel that pain, right? And it is not worth it. Like if genuinely, if a customer asks me, you have to pick one. I would say maybe in two years, come back, give up, chat, GBT, and go for Google. Right now it's still a bigger market.

Kelly It makes me think of almost a variation of the saying of like you know, you don't have to remember a lie if you tell the truth.

Kelly This is one where like you don't have to worry about the long-term ramifications. If you focus [00:31:00] on your genuine wins, your genuine customers, your genuine content, and focus your attention on maybe less volume, higher quality, because again, it may not move the needle as fast as you want. But it's more, it's a stronger foundation, right?

Kelly It's, I'm use another cliche, you know, you're not building your house on, on a foundation of sand, right? So

Todd a hundred

Kelly percent.

Todd well that's a lot for anyone listening to think about, including me. the one thing I'm taking away from what you guys were discussing and Sammy, I don't know, what do you think about this?

Todd In the last 20 years, anything I've, I've been involved in around SEO or marketing just in general, but certainly SEO has been content is king. And all I heard you talk about, for the most part in this was a lot about content, UGC content, brand content, case study content. Is that still one of the number one, if not the number one ingredient for answer engine optimization?

Sami And the recommendations might be different, like Right. We might uncover for our clients', you know, [00:32:00] publishers they never knew about that they should be on, we might uncover specific red threads that they should go and put a branded transparent comment on. Generally a lot of the same skills that were deployed in SEO, even if the recommendations are different, like the topic you should write about is different, are going to apply here.

Sami And that's why like for us, we believe that the SEO skillset of the last 20 years is one of the most well positioned rules to actually optimizing four generative engines in the future. Right? There may be like. Other things you have to care about and other considerations, and the recommendation might be different, and you might have to build, a perspective that's unique to your industry and your brand, but ultimately it's those skills that are gonna do well in the next five years.

Todd Well, the interesting thing about this for me is, and I'm old and I've been doing this since before any of this existed, so my perspective's a little skewed, but brands and the identity of a brand is, A promise. And a [00:33:00] promise delivered. And a lot of times that is fulfilled through proper storytelling.

Todd that's a very commonly used phraseology. But that's still really, really important, is that a brand promise must be conveyed through really authentic, truthful, unique storytelling content, whether that's through a pr. Piece or a user generated piece, or just a brand case study and, and sort of statement of fact that's in there.

Todd It sounds like the brand promise and this content storytelling is still the most important thing, which I'm happy to hear because I, I don't know how people think about brands anymore and consumerism and how much of it is AI and manipulated, or real or fake anymore. It, it just seems very, very difficult, universe to navigate.

Todd So, but I'm happy to hear that it sounds like. A brand and their promise and their story still matters. I'm gonna leave my happy note on that. Kelly, what is your, uh, final thoughts on this?

Kelly Yeah, so I've, kind of one tough final [00:34:00] question for you, Sammy, to think about and we did a episode recently, recently, and you know, there was a rapid fire question session, and one of the questions was, is ai, hype or breakthrough?

Kelly And. I find myself, the more and more I go down this rabbit hole with AI and learning and exploring all these things. I leaned my answer was around hype. it's, you were the only one that's said hype. Just for the record, I was the only one. But even though at its core, I definitely believe there's unbelievable breakthrough in there.

Kelly And I set this up in the question to you, which is in your space, in what you're dealing with every day, what do you believe is the most dangerous sort of. Type or over promise or kind of misguiding thing that you're getting these marketing people who want to believe there's an easy button, want to believe there's pay a little money and triple my sales type things.

Kelly Like what are you seeing is sort of that over promise or that over. Emphasis on what's possible versus the reality that we just talked [00:35:00] about, which is there's still authenticity, there's still truth, there's still long-term planning and foundational work. Like where are you seeing that big gap come up and what are the the common themes that you are the most worried about people believing and getting burned by?

Sami I honestly want to go back to answer that to what Todd just said about storytelling, and we onboarded one customer who switched over to us from another. G-E-O-S-E-O provider who that they had been working with. And the first thing that we did was we were like, Hey, like we wanna schedule time with your team to understand how your business works.

Sami And they're like, okay. And then we get on the phone, we start asking questions like, what do your customers like about you? What's negative feedback that you guys receive? What are you working on improving? how do you want your customers to feel when they engage with your brand? And we ask all these questions.

Sami They're like, wait, you know, what's this for? Like, I thought you guys were gonna go push some buttons and then like. Have, a new Reddit thread on a webpage and get us, mentioned on chat Ccbt. I think people are surprised that it just like takes really hard work and you have to [00:36:00] truly understand the business and that the person who's best positioned to help you with GEO, like outside of somebody who you really trusted a third party is probably you.

Sami Right. Like it's probably somebody who has the most context on your brand, who can communicate your story most authentically. And you should probably be thinking about it yourself internally before you try and outsource a lot of this stuff. And I'm saying this as somebody who, I want people to outsource it to me.

Sami Right. And I think, you know, we're a bit better than the average for a lot of reasons, because we sit down and we try to understand this story before we ever go into our software. But it takes real hard human work. And I think most of the venture dollars in this space have gone to companies that are promising AI workflows at scale, right?

Sami Like you can push some buttons and you will get, answers at scale. That's where most of the VC money has gone. It's where most of the eyeballs have gone. And I think customers want to feel like it's true. Customers are surprised when they hear our price point versus some of the software tools.

Sami And I, I say like, you know, I'm sorry to break it to you, but like. You're working with a real person who's gonna speak to [00:37:00] your customers, who's gonna understand how they engage with your brand, who's gonna ask them what their story was, working with your product or your service, and who's gonna sit down and spend three hours writing real content about that?

Sami And they're like, well, why don't I just go make, you know, they're charging me 399 bucks a month. Why don't I do that? And I say, you know, honestly, you can try. but I think in my space, that's our contrary opinion to where all the investing dollars have gone and where all the eyeballs have gone and.

Sami We get investors that say to us, well, you know, how are you gonna do that at scale for 20,000 companies? And my answer is like, does that matter if it's just the right answer? Like does it matter if you can do it at scale? What if you can't? What if it's just something that needs. A lot of hard work. I don't think they like that answer that much, but

Kelly it's funny that, yeah, I mean, your answer is literally exactly what ours is in our space.

Kelly So it's really refreshing to kind of hear and yeah. Yeah, I, I agree with it completely. It's people, the thing that's bothering me the most is people want to believe too much, too [00:38:00] fast and too easy, and it's, and I keep going back to what it's really powerful at doing is augmenting good people to do really, really great work.

Kelly And you're right. People don't want that. People want, no, I want easy button. I want cheap, fast and done. And I'm like, yeah, well you can have that, but it's not the real value that's there. So yeah, it's really interesting. That's the exact same thing in your space and and ours, and I think a lot of others.

Todd Yeah. Well, everybody wants the easy button, Kelly. That's what everybody wants. lots to talk about, on this chat and a lot more. So I'm, I'm gonna propose that you come back maybe in January and we talk about the advertising models that are starting to sprout up. open AI's ad launch.

Todd What does that mean? Personally, I think as soon as you open things up to paid advertising, that's when a lot of the. Sort of manipulation and, bad things happen. but we'll see how that evolves. So I'd love to talk to you more about that, maybe how some of the retail partnerships are starting to find their way into [00:39:00] this.

Todd I, I would love to even dig into any of your knowledge about sort of the agentic, agents that are going on, are starting to be developed. In other words, can I just go create one and tell it to shop for me? How does it know what to shop for if I don't tell it? There's probably some manipulation going on in there, so be thinking about that.

Todd 'cause I'd love to have you on again to talk about that. 'cause I'm sure brands are wondering how they're going to navigate that space. on the consumerism side of things, and I'm as guilty as you are as. Or North Face or Patagonia or any of those brands show up in my feet, I'm, I'm more than likely to buy.

Todd So they know that. and they stick it to me as they are right now during the holiday shopping season. anyway, thanks for, coming on today. We've got a lot more to discuss and hopefully we can get you back in here in January and talk.

Sami Love it. Thank you guys both so much for having me.

Todd what's the website? URL You want people to go check out? We'll put it in the notes. And, um, anything else you have to plug while you're at [00:40:00] it?

Sami Petrol labs.ai, I'll, put the URL so you can put it in the notes. And no, other than that, hope everyone enjoys, a really great Christmas and holiday break and gets a chance to unplug.

Todd Very good. we'll talk to you soon. Thanks Sammy. Thank you.

Intro Thanks for listening to contrary to Popular Opinion where we challenge you to think independently, push the limits, and take uncommon action so you can achieve game changing results in your business and your life. To stay up to date with our latest episodes, be sure to click the follow or subscribe button now.

Intro Thanks again for listening. See you next time.

About Contrary to Popular Opinion

Contrary to Popular Opinion is the show where marketing's conventional wisdom gets cross-examined. Hosts Todd Juneau and Kelly McGuire take the advice every marketer repeats — and test whether it actually builds brands or just follows the crowd — drawing on decades inside the advertising business to give leaders permission to do the thing nobody else is doing.

See all 81 episodes of Contrary to Popular Opinion →

Book a 30-minute strategy call