GEO, AEO, LLMO:
Separating Fact from Fiction
& How to Win in AI Search

Catch the key takeaways

Over the past two years, SEO has evolved faster than ever before. With the rise of AI-driven search tools, new terminology like GEO, AEO, and LLMO has flooded the industry. But as tempting as it may be to lean into every acronym and trend, many marketers are missing the bigger truth: today’s AI Search success still comes down to understanding people, not just platforms.

AI search is expanding traditional search, not replacing it. In fact, data shows that 95% of ChatGPT users still rely on Google, and AI search currently drives less than 1% of total site traffic. While large language models are influencing discovery, they’re built on the same foundation traditional SEO supports: accessible, trustworthy, well-structured information.

The way people search is evolving, but the way we earn trust, visibility, and authority is still rooted in the same principles that have always defined great SEO: useful content, technical precision, and a strong brand reputation.

It’s worth noting that while Amsive prefers the term AEO, the terminology describing evolving search tactics overlap. Lily references GEO throughout the presentation, which is similar to AEO, and is sometimes used alongside related terms like LLMO and AI search optimization.

Dive into the transcript

00:00 Hi everyone, and welcome to GEO AEO LLMO, Separating Fact from Fiction and How to Win an AI Search. This is an online adapted version of the presentation that I gave at MozCon New York City last week, and I just wanted to share it online as well.

00:17 So, for anybody who doesn’t know me, my name is Lily Ray, and I’m the VP of SEO Strategy and Research at an agency called Amsive, based here in New York City.

00:27 And this was my fifth annual year presenting at MozCon, for anybody in the SEO and AI search space, this is a bit of what it’s felt like to stay on top of all the latest trends and developments this year.

00:50 We have all types of new vocabulary and terminology that we need to learn, like chunking content, prompt engineering, token efficiency, cosine similarity, model context protocol, agentic workflows, so much new, you know, new concepts and new things to learn this year.

01:07 It’s been very overwhelming. I feel like every day we’ve had to stay on top of learning all the latest developments in the AI space.

01:13 And I would say the most important question on the top of everybody’s mind is, is this GEO, AEO, LLMO, or really just SEO?

01:22 And in order to present this today, I wanted to keep things simple with a little bit of a story time.

01:29 And I was thinking about, well, how can I convey this in a way that would make sense? And I thought back to this timeless article describing our beloved SEO industry as one giant alligator party.

01:39 So if anybody remembers the people that ruined the internet, uh, it was written by an author at the Verge. Um, they talked about how our industry is an alligator party.

01:47 So I thought I would stick with that motif throughout today’s presentation. You might be wondering, well, I’m based in New York City, why on earth would I be talking about alligators in New York?

01:57 Well, believe it or not, we have had our share of alligators before. Appearing in our subway system of New York, in the gutters of the subway.

02:04 So it’s not impossible to encounter an alligator once in a while. Um, but you’re much more likely to encounter a rat on the New York City subway.

02:10 So don’t be too alarmed when you come to visit our city. So anyway, let’s take a journey into 2025, the year of AI search, the allegory of the alligator.

02:24 Our story begins in mid 2023 in the SEO swamp, deep in the gutters of the subway beneath New York City, where the SEO industry’s biggest evolutions are handled by a party of alligators.

02:38 Once a year, the SEO swamp alligators gather for a their most sacred tradition, the SEO hype cycle monopoly tournament, where each alligator insists that they’re winning, but no one ever actually lands on page one.

02:50 Here we meet mother SEO, caring, strong, and a bit tired of explaining canonical tags. She’s pregnant and expecting another little alligator, a shiny new evolution.

03:00 The swamp is already hyping up. Could this one truly be the next big thing in search? Father PPC is, well, always reallocating the family budget, preferably toward anything that can be tracked, attributed, and justified in a quarterly report.

03:18 Each alligator represents an important chapter in SEO history. Each child still has a faint trace of the bright colors they were born with during their own hype cycle, a reminder of when they, too, were the next big thing in search.

03:33 Here we have mobilegeddon, that 2015 update where Google essentially told every website, if your site isn’t mobile-friendly, it might as well not exist.

03:41 He’s still forever flexing about that one heroic moment where he forced the entire internet to go responsive, yelling, go responsive or go home at anyone who dared to load a desktop experience on a mobile device.

03:55 Next up is Voice Search, the darling of 2016 SEO conference stages. She was absolutely phenomenal. She absolutely convinced that she would replace all typing by 2020.

04:03 For a brief, glorious moment, she had us shouting, okay, Google, into our phones and believing this was indeed the future.

04:10 While she still insists it is, if you would just speak clearly into the microphone and stop having so much background noise.

04:18 And here comes AMP, otherwise known as Accelerated Mobile Pages, who once ran faster than anyone at the SEO swamp. In 2016, he was promised a life of fame, top stories, Google Discover, lightning bolts, instant mobile glory.

04:31 But after peaking far too early and losing his special treatment, no one seems to call anymore. He still keeps his running shoes on, though, just in case Google chooses not to give him a second chance.

04:45 Now we have featured snippets. The overachieving middle child who always has the right answer. Back in 2018, she was the golden girl of search, answering questions directly at the top of Google, complete with a shiny position zero crown.

04:58 In her opinion, she’s the one who’s truly started zero-click search. But did she ever get the credit? Of course not.

05:04 Someone else reads her work out loud, takes the clicks, and she’s less sitting there like, I literally wrote that. And here’s Core Web Vitals, the gym bro of the family, forever flexing about his gains in page speed, responsiveness, and visual stability.

05:19 He arrived in 2021 with one goal mission, whip the entire web into performance shape. He’s always shouting, always be optimizing, while checking his Lighthouse scores like they’re his biceps in the mirror.

05:30 97% of the SEO swamp still doesn’t understand what he does, and they know deep down he’s still just a tiebreaker ranking signal, but they applaud anyway.

05:42 The SEO swamp alligator family lives in perfect harmony, each alligator adding its own tint of past hype to the ecosystem.

05:49 They’ve conquered every SEO is dead headline together. But tonight, deep beneath the roar of the New York City subway, something is stirring.

05:58 Because tonight, a brand new revolution. Is about to hatch, one that promises to change everything, just like the last eight did.

06:07 And then, Mother SEO’s long-anticipated egg finally arrives. But not just any egg, this one is covered entirely in gold. The SEO swamp has never seen anything like it.

06:17 There’s awe, confusion, and a tiny bit of fear. If you listen closely, you’ll hear a faint whisper from within. Feed me data.

06:27 The elder siblings peer suspiciously at their new brother. They’ve seen this before. The shimmer. The promises. The venture capital circling overhead like hungry vultures.

06:37 AMP and Core Web Vitals exchange a look that says, give it six months. A star is coming, but what will his name be?

06:46 The naming ceremony commences. Everyone has an opinion about what to name the new baby brother. Featured snippets suggest AEO, Answer Engine Optimization.

06:58 Voice search mentions that the name is already held by American Eagle Outfitters. The mother proposes LLMO, Large Language Model Optimization.

07:06 But Mobilegeddon says, Mom, that’s a mouthful. All eyes turn to grandma SEO who’s been quietly observing the spectacle from the back.

07:13 She’s ancient, wise, and permanently unimpressed. She adjusts her reading glasses, glares at the chaos, and grumbles. This is all just SEO.

07:24 Father PPC proudly holds the golden baby above his head. He proclaims, We shall call him Geo. Amp tilts his head, blinking.

07:33 But dad, doesn’t Geo already mean, like, ten different things? Father PPC waves it off. Geo it is, he says. Generative Engine Optimization.

07:42 It sounds sufficiently complicated. Baby Geo grins, glowing, glimmering, and already convinced that he’s the main character. No sooner has Baby Geo hatched than the headlines start to explode.

07:56 There’s a full-blown media frenzy. They claim, SEO is dead. All hail David Geo. Everything is chunking now. Without Geo, you’ll be invisible in AI search.

08:06 There’s all this new, significant pressure from the media and the C-suite to think beyond SEO because Geo has arrived. The rumors about Geo begin to fly around the SEO swamp, and Baby Geo is thought to revolutionize the entire search landscape.

08:22 So what do we begin to hear about Geo? What are those rumors? First, we hear that Google and SEO are dead because everyone starts their search now on ChatGPT.

08:31 We hear that you should optimize for conversations and prompts rather than focusing on keywords. We hear that content must be properly chunked, converted to vector embeddings, and measured with accuracy.

08:40 With cosine similarity, or else you’ll be invisible in AI search. We hear that you must structure your content for AI ingestion, like creating FAQs, questions and answers, structured data, bullets, and tables.

08:52 We hear that you need to earn third-party mentions across highly cited domains to be visible in AI search. And that you need to own the conversation on Reddit, YouTube, and Facebook.

09:01 We hear that AI search is multimodal, which means that it can ingest different types of content like video, audio, images, and more.

09:08 We hear that you need to think about the evolved SEO metrics with new KPIs and new large-language model tracking tools.

09:12 We also hear that query fan-out is extremely important, as these AI tools can conduct multiple queries. Multiple searches simultaneously, so we need to think about all the different questions that users might be asking and optimize for all those different intents.

09:29 Baby Geo becomes a celebrity overnight, dripping gold, wearing sunglasses, posing for the paparazzi, just like every prompt was written just for him.

09:37 He bursts into the scene with a set list of new tricks, like converting content into embeddings, dominating AI overviews, eating clicks like they’re candy, and almost always refusing to cite his sources.

09:52 The AI press, ChatGPT the Owl, Gemini the Dragonfly, Claude the Pelican, all praise him as revolutionary and disruptive. Investors swoon, CMOs panic by dashboards, LinkedIn influencers update their job titles to GEO experts before lunch, the SEO world cheers and trembles.

10:13 The SEO swamp erupts in chaos. Young Geo, now sporting designer sunglasses and a boa, performs his hit single, Chunk It Like It’s Hot, to a crowd of mesmerized tech reporters and confused aunties trying to understand what their nephew does for a living.

10:30 So, budgets and staffing begins to change to capture all this new AI search demand, and we see that Google’s AI overviews are indeed leading to declining organic clicks.

10:40 The reality is that we see with a lot of sites that AI overviews are cutting into traffic, and we see this phenomenon that’s commonly referred to as the Google Search Console alligator mouth effect, where impressions are rising up while clicks are declining.

10:54 We hear that SEO KPIs, like clicks and rankings, are beginning to subside in importance, as we need to think about new SEO KPIs, which we’ll talk about a bit later.

11:03 And as a result of this, new large language model tracking tools begin to arise. Behold, King Geo, perched atop his velvet throne labeled AI overviews, basking in the spotlight.

11:14 Here in his swamp palace, he devours your organic cliques like they’re Michelin-starred grapes. And what does he offer to the humble wealthy?

11:20 Tiny scraps of paper currency, each one reading click-through rate, 0.8%. Publishers, shown here as the swamps, frogs, turtles, even the occasional possum, leap desperately for the crumbs of visibility.

11:36 For many, it’s become feast or famine. Meanwhile, GEO’s loyal servants, ChatGPT, Claude, Gemini, and Perplexity, present silver platters overflowing with queries, snippets, and answers.

11:48 GEO smiles and takes all the credit. But the real spectacle lies just beyond the throne, in GEO’s gift shop. Because remember, visibility is temporary, but service swag is forever.

12:02 A new line of AI visibility tracking tools has emerged. Profound plushies, peak AI puzzles, scrunch board games, Ahrefs brand radar binoculars, and even end-to-end automations and make.com puzzle kits for the true power user.

12:16 Collect them all before your CFO notices. Finally, we’ve arrived at GEO’s glamorous digital PR penthouse, where influencer peacocks, eagle-eyed executives, and brand-obsessed otters gather.

12:30 Here, Geo teaches one golden rule. If you want to show up in AI Search, you’d better show up everywhere else too, and with a good reputation as well.

12:40 Meanwhile, back at the SEO swamp. As the swamp settles into evening, the forgotten siblings gather around the TV. Voice Search clutches a dusty MozCon trophy.

12:52 Voice will take over by 2020, it reads. Remember when Cindy Crumb said I’d change everything, she whispers, hopeful yet hoarse.

13:00 Mobile get in, pats her gently. Hey, we had one incredible year. Mother SEO smiles proudly, flipping channels, but on every channel, it’s Geo.

13:11 Late night talk shows, sports highlights, a cooking segment titled chunking content for breakfast, the siblings groan, it’s the circle of hype.

13:20 The TV flickers again, and Geo’s golden face fills the screen for the 47th channel in a row. Mother SEO wraps her tail around her children and murmurs, don’t worry little ones, remember that everyone gets their moment in the spotlight here in the SEO swamp.

13:38 Let’s take a trip back to a simpler time, when every SEO conference keynote promised a revolution. We start with the glamorous alligator Cindy Crumb.

13:47 Charismatic, stylish glasses, perfect hair. On the MozCon 2018 stage proclaiming, voice search is the future. 50% of searches will be voiced by 2020, and mobile-first indexing will change.

14:00 In the audience, a young voice search and mobile get-in, glowing with their original bright colors, leap up and down, convinced that this is their forever moment.

14:12 Now, a roaring stadium in the swamp, AMP and Core Web Vitals are racing the 100-meter dash. AMP bursts forward with unbelievable speed.

14:20 The crowd loses its mind, up until Core Web Vitals jogs past with perfect form. The scoreboard reads, need for speed, as fans chant like they’re trying to pass Core Web Vitals before month-end reporting.

14:35 Now flashback to Ross, Ross Simmons. His alligator takes the MozCon 2019 stage. Stylish jacket, swagger of a creature who knows distribution is destiny.

14:45 Create once, he says. Distribute forever. In his talk, Ross shared the importance of having your brand mentioned across all platforms, like Reddit, X, and YouTube, and ensuring content was produced in different formats.

14:59 For multimodal search, such as audio, video, images, presentations, and more. The link, the viral linkbait alligator, still rainbow-colored, sheds tears of pride.

15:09 This was his era. The swamp cheered for cross-platform amplification. Does all of this sound a bit familiar? We also start to hear Google’s response to the rise of Geo.

15:20 A few months back, Danny Sullivan from Google gave a presentation where he confirmed that good SEO is really the same thing as good GEO.

15:31 We start to hear some other sites and other companies debunking GEO myths. First, we hear Google again saying that having good SEO, and therefore good Geo, is largely just having great content for people.

15:42 We hear that Google saying that you don’t really need anything new or special to appear in Google’s AI overviews, like AEO or GEO.

15:50 We hear companies like Ahrefs saying GEO, LLMO, AEO, it’s really all just SEO. And John Mueller from Google talks about optimizing sites for embeddings, which he refers to literally as keyword stuffing, which, is a form of spam in Google search.

16:04 We also start to see throughout the year, an interesting phenomenon whereby new articles that people are publishing, like these articles that I published on Ampsiv and on my personal blog and on social media, without doing anything new or special, you know, just provide some good content and use SEO

16:21 fundamentals and best practices, we’re seeing that these articles are quickly cited within a few hours within all the different AI search platforms.

16:29 And when I share this with many people, they say, well, yeah, Lily, you have a big platform. You know, you have lots of people reading your content.

16:35 So maybe it’s because of that. Well, my teammate, Will, who’s been publishing a lot of great content this year. He sees the same thing as well.

16:42 When he publishes on the Amsive blog, we see his article cited within a few hours. And if you search for a lot of different queries around the blog that he wrote this year, you’ll see his content being cited across different LLMs as well.

16:53 I’ve spoken to many people publishing content throughout the year without doing anything revolutionary, anything new or different. They’re seeing their content cited across all the different LLMs within a few days.

17:04 So suddenly, cracks are starting to appear in our alligator Geo’s golden skin. Because nature, as always, has other plans. Under the harsh light of questions about what tactics have actually changed with AI search, Geo’s golden coating begins to flake.

17:20 Publish your in the audience gasp. Beneath the shimmer lies green scales, the very same green as mother SEO. Geo panics, frantically Googling how to maintain hype cycle momentum.

17:34 But it’s too late. The swamp creatures whisper the forbidden truth. Chunking is just good formatting with optimized headlines. Schema markup?

17:44 But we’ve been doing that since 2011. Featured snippets, Q&As, FAQs, tables, and bulleted lists. All of that is so 2019.

17:51 Perhaps the golden child isn’t so golden after all, they say. In the soft glow of the SEO swamps’ fireflies, Mother SEO’s embraces her golden child.

18:03 Listen, my precious data goblin, she says. Every one of your siblings thought they too were the chosen one. But here’s the family secret.

18:11 We’re all just trying to answer searcher’s questions better. Sweetie, that golden shine in your scales? We’ve seen it before. Mobile first, voice search, even your brother AMP had his moment in the sun.

18:25 She gestures to the SEO swamp ecosystem around them. Reddit frogs are croaking in the distance. YouTube herons are fishing for engagement.

18:31 Wikipedia turtles are slowly but steadily providing the source of truth. Baby Geo proclaims, but they said I would I’m supposed to be the future.

18:43 Oh, my darling boy, Mother SEO proclaims. The future doesn’t erase the foundation. You still need us to retrieve, to index, and to ground the truth.

18:52 If Google and Bing are the brainstem, I’m the cardiovascular system. Without me, SEO, you’d hallucinate yourself into a lawsuit. Geo murmurs, huh, so I rely on you?

19:04 And Mother SEO says, Oh, sweetheart, even GPT-5 relies on search more than ever. The reality is that large language models use what’s called retrieval-augmented generation, otherwise known as RAG, which powers traditional SEO power systems.

19:21 Our AI search with retrieval-augmented generation. And the reason for this is because AI tools are using this technique that enhances the accuracy and relevance of large language models by allowing them to access and retrieve information from external knowledge sources, like search engines, before generating

19:36 a response so they can provide the most clear, factual, up-to-date information, especially when the user conducts prompts that are recent in nature.

19:46 And the reality is that GPT-5, OpenAI’s latest model, is even more reliant on search. This is a great article by Dan Petrovic, who’s been digging deep into how large language models work.

19:56 And the reality of GPT-5 is that without grounding, or retrieval augmented generation, this model is virtually useless. It’s designed to be the brain on top of tools and information that it’s provided with, which really implies that SEO has never been more important than it is now, because in order to

20:12 appear across these different LLMs, you need to appear in search engines first. This is Brittany Muller, former MozCon MC and multiple-year MozCon speaker as well.

20:24 And she’s been working in the AI space for a long time. And what she shared on LinkedIn, and you can look up her post on this, is that every single URL that you see in an LLM output actually comes from a search engine API, Google or Bing.

20:38 And the reality is that many of the supposedly new GEO tags are actually rooted in traditional SEO best practices. We hear a lot about content chunking, so breaking up your content for, so it’s more digestible for large language models.

20:51 The reality is that Google’s been doing a version of this for a long time. BERT in 2019 was a deep learning language model that helped Google to understand queries and passages better, which helped them to, uh, improve disambiguation and to improve the relevance of featured snippets.

21:07 So chunk optimization isn’t really an SEO tactic. it’s an AI engineering function. Large language models don’t see your pretty formatting.

21:14 They see tokens or tiny units of meaning. They, not you, decide how to slice your content. There’s not really a cheat code that you can use to chunk your content for AI overviews because every model chunks differently.

21:27 They’re constantly changing this for speed, accuracy, and cost considerations. So what really is in our console, in our control, it’s to create clear, self-contained, atomic answers that stand on their own, whether a human being reads them or an AI extracts them.

21:44 We hear a lot about writing clear, concise answers that are formatted for LLMs, but the reality is we’ve been doing this for a long time using things like frequently asked questions, how-to, Q&A, and the schema that supports all those different content types.

21:58 We’ve been doing this for a long time. All these things can be helpful for AI solutions. We hear about multimodal content optimization, which is this notion of producing content in different formats because large language models can indeed ingest different types of content like video, audio, and images.

22:14 But the reality is we’ve been doing this in the SEO space for a while. We’ve been producing video content for YouTube and Instagram and TikTok.

22:22 We’ve been producing optimized podcasts. You know, Google’s been supporting podcast optimization for a number of years. We’ve been producing audio transcripts for web accessibility and to make sure that when we produce something in audio or video that we have the transcripts to support it.

22:33 We’ve been doing online reputation management to make sure that our brand is mentioned across all the different places. This is where it matters.

22:43 In the AI search space, we hear a lot about earning brand mentions and citations, but in the SEO space, we’ve been doing digital PR, integrating with social media teams, PR, and influencer teams for many years to make sure that our brands and our products are cited across all the different sites and

22:59 all the different assets where it matters to be mentioned. We hear a lot about query fan-outs, so this notion of you know LLMs conducting multiple search queries simultaneously that basically anticipate all the different questions that the user might be asking as part of one prompt, but the reality is

23:14 we’ve been doing a version of optimizing for this for a long time in the SEO space. We have tools like People Also Ask, Long-Term Keyword Optimization, and Question Research across all the different popular, uh, SEO keyword research tools that basically do a similar version of what we’ve been optimizing

23:30 for with query fan-out. It’s also important to understand that AI search is absolutely exploding, but let’s not forget Google. This is some data that was shared with me from SimilarWeb that looks at how frequently people are using different LLMs compared to Google search.

23:43 So if you look at the bottom here, you’ll see usage for different large language models around the world, with ChatGPT being the green line that’s growing the most, and Gemini’s also growing quite a bit as well, that’s Google search.

24:02 While people are heavily adopting large language models this year, it has not cut into Google’s market share lead across the whole entire search ecosystem.

24:12 If we also look at the number of large language models that are being downloaded worldwide, uh, the green bar here is ChatGPT.

24:18 People absolutely have, you know, adopted ChatGPT all throughout the course of 2025, but if you look at the last few months beginning in May, we’re also seeing that Gemini is being widely adopted and download it around the world.

24:31 This is data that comes from SimilarWeb as well. And if you think about the actual impact that AI Search is having across our traffic for many different types of sites, Glenn gave you wrote a great article about this earlier in the year.

24:43 And it’s something that we’ve seen with our clients at AmpServe as well, which is that if you group all the different large language model referrers together in a tool like Google analytics and you look at how much traffic they are driving compared to something like organic search, it’s still pretty

24:57 minuscule. Uh, Glenn Gabe saw with his clients, it’s about 1% on average with AmpServe. We’ve seen about one to maybe now 2% of traffic coming from LLMs compared to organic search, which is usually, for many sites, good to be 30, 40, 50% of their overall traffic.

25:09 We also see that the majority of ChatGPT users are still using Google. most of them have not abandoned Google. If anything, they’re actually using Google more.

25:26 And that tiny blue sliver here is the percentage of users that are only using ChatGPT. So the reality is most people that are using ChatGPT are still heavily using Google.

25:36 And there’s some new data that came from Rand Fishkin and Spark Toro. They’ve also been working with Dados as well as SEMrush to collect some new data around search engine usage and LLM usage in 2025.

25:48 And what they’re seeing is that traditional search usage, which is reflected here in these blue bars on the left side, has basically remained the same throughout 2025.

25:57 While people are indeed adopting ChatGPT and other LLMs, which you can see here as the pink bar. The light pink bars are heavy usage of large language models throughout the year.

26:06 This hasn’t exactly cut into their use of search engines like Google. If anything, it’s actually expanded had they used Google.

26:13 So the table on the right shows a study that SEMrush put together, which basically looks at the number of searches per week on Google.

26:20 Before using ChatGPT, the average user was seeing 10.5 sessions per week on Google. After using ChatGPT, that number actually increased to 12.6, which means that ChatGPT is actually encouraging users to go to Google more than it was before.

26:37 So the reality is, while there’s significant overlap between LLMs and search engines, questions. Thank you. They do use these tools in different ways, and Google used some great language for this recently.

26:46 It was an interview with Robbie Stein, who’s the VP of product at Google search. And what he says is that AI search is actually expansionary.

26:53 It’s a new additive tool that people are using on top of Google search. And I thought that this was an interesting way of thinking about this, because the reality is there’s different ways that people are using LLMs compared to search engines.

27:05 If you think about this, while there’s a lot of overlap, there’s many unique use cases for LLMs, such as creative writing, homework help, brainstorming, learning a language, generating code, acting as a personal assistant or a therapist, summarizing documents, creating personalized recommendations.

27:21 Or generating images. These are all net new ways that people are using LLMs, but they still are using search engines because there’s a lot of unique and important use cases for search as well, like fact checking and research, looking up images or videos, maybe navigational search, like navigating directly

27:36 to a website, shopping online, finding copyrighted music and lyrics, looking at maps and directions, all kinds of different use cases that are still to this day distinct from LLMs.

27:47 Even though LLMs are starting to adopt some of these behaviors, remember that most people around the world start their searches on Google and it’s going to be very hard for them to really kind of change those habits as we’ve seen over the last few years.

28:00 So after a little long day of shedding expectations and, well, actual gold skin, GEO looks in the mirror. What remains is a strong, confident mix of green and gold.

28:09 He finally realizes something important. He isn’t here to replace the SEO family. He’s really just their newest evolution. Mother SEO reminds him how special he is.

28:20 You help the world discover answers in brand new ways, she says, with conversations, multimodal searches, personalized results, paths that we never had before.

28:30 Geo beams, because he knows he has a job that nobody else in the SEO swamp can do. He celebrates his new role with his siblings, high-fives for featured snippets, a supportive nod from Quora website.

28:41 vitals, even a hug from his sister voice search, because the reality is with AI, the people search, the way that people search has absolutely changed.

28:51 But the tactics to earn AI search visibility are really just evolutions of existing SEO and marketing best practices. So what are those tactics?

29:01 What’s so new about GEO, which, by the way, it would be a good time to mention that my team calls it AEO, answer engine optimization, also known as AI search.

29:10 There’s all kinds of different acronyms for it, but it largely all means the same thing. So what’s so new about GEO or AEO and its role in the SEO family?

29:19 Well, there’s a lot of new metrics. This is actually absolutely true. The way that people search is changing and the metrics are evolving just as fast.

29:27 So before we used to focus on things like clicks and rankings, now we need to think about some new KPIs for AI search.

29:34 For example, branded impressions, how frequently people are actually looking for and seeing our brand name, share a voice in AI search, direct and organic clicks to the homepage.

29:44 As large language models will often recommend your brand, people might be navigating directly to your company website or searching for your website in search engines.

29:52 We need to think about AI search visibility by topic which you can use with any of the large language model tracking tools to kind of break out visibility by topic.

30:00 We need to think about how we can We traffic engagement and conversions from AI search traffic. And we need to think about how visible we are across different large language models.

30:09 And are we more visible on one large language model than another? We need to think about setting up analytics to reflect all of this.

30:17 So Dana DiTomaso is an expert in Google analytics who shared some great articles, which you can look up around how to track and report on traffic from AI tools in GA4.

30:26 You can group them all together into a custom segment like AI search to understand the true impact that this is having on your traffic, on your engagement, and on your conversions as well.

30:36 I also think it’s important to answer all the different branded questions to try to get ahead. And of how people might be searching for your brand, your products, your services, your competitors, and answer all these questions directly on your website.

30:48 So you can use regular expressions and search console to generate all the different question keywords, plus your brand name, and to kind of filter Google search console to just see that data.

30:57 And you can also use SEO keyword research tools like Moz, like Ahrefs, like SEMrush, and many other tools to basically isolate question keywords around your brand and to make sure you’re answering all those questions directly on your website.

31:09 because the way I see it, AI search is really just an evolved form of experience, expertise, authoritativeness, and trustworthiness, which is a popular acronym in the SEO space, and reputation management as well.

31:21 So think about proactively answering all the possible questions that a user might have about your brand using clear and concise language that would make a lot of sense for a large language model, as well as for humans.

31:33 Here’s a few different resources that you can use, uh, that we’ve built, uh, from AmpServe. So, number one, you can visit our branded question checklist by scanning this QR code.

31:43 And what this does is it gives you a checklist that you can use, and obviously, use it as you see fit.

31:47 Maybe not all questions make sense for your brand, But you can think about answering all these different questions throughout your site to really provide clarity about what your brand does, what your leadership team does.

31:58 And so when people are asking large language models, they have a very clear answer. They have place that they can look to on your website that answers all these questions.

32:05 You can also build and expand pages for EEAT and AI search visibility. So this is another sheet that we put together where you can look at some inspiration around how can we talk about our company, our mission statement, our leadership team, our corporate structure, all our different products and use

32:21 cases, uh, comparison pages, integration pages, legal and investor relations, all kinds of different things that many companies haven’t actually thought of adding and clear language to their website, but we think all of this can be really, really helpful so that when people ask these questions of large

32:36 language models, they have a place that they can go on your website to clearly answer those questions. With AI search, you need to think about personalization because as LLMs become increasingly personalized, your content should as well.

32:48 You want to think about having translated and regionally appropriate versions of your content if you have an international website. You need to think about having your content targeted to specific audiences and speaking to the specific pain points of different demographics.

33:02 So for example, if you’re a car company, you might want to talk about having different content for new parents than you might for new drivers that just got their driver’s license.

33:11 You need to think about having FAQs, Q&As, and user-generated content that’s really rich with authentic, honest conversations to comment questions, because large language models do tend to cite.

33:21 Information that they’ve found in real human user-generated content. You can use LLM tracking tools like Profound, Peak. There’s all kinds of new tools that are out there that help you monitor the top sided domains and pages in your space.

33:35 This is a great way to go into your specific niche and understand which pages, which domains, uh, which of them are being cited the most, which brands are they mentioning the most, and how can you ensure that your brand is mentioned in these conversations, which is often a digital PR play.

33:52 Remember that LLM citations are highly volatile. This is by design. So if you have a piece of content that you have cited in AI Search or you want it to be cited in AI Search, recency doesn’t matter.

34:01 It definitely matters. Now, I definitely don’t recommend what Google calls artificial refreshening, which is to update content or say that content’s been updated when you haven’t really made substantial changes to the page content.

34:14 I believe Google has ways of demoting this type of content, which can affect your visibility both in search as well as large language models.

34:21 It’s also just not that great for users. So if you are going to be updating your content, you need to make sure that it is truly updated and that it reflects the most recent and up-to-date information.

34:32 The reality is that large language models are heavily influenced by Wikipedia, Reddit, and YouTube as well, so to the extent that you’re going to have a lot of content, you need to make sure that it reflects the most recent and-to-date information, so to the extent that you in to that, you’re going have

34:53 a lot problems. can think about producing informational content because the reality is, especially Google, you know, Google loves to cite YouTube, but all the different LLMs will cite YouTube content.

35:21 That’s something that you can own, and you can get ahead of those conversations by creating short videos, um, and there’s all kinds of new AI tools, new-new AI tools that you can now use to splice up those videos into things like YouTube Shorts and Instagram Reels as well.

35:35 Keep tabs on agentic browsing and commerce. You know, this is absolutely a new area, a new focus area for those of us that are interested in AI search.

35:44 We now have ways of transacting on ChatGPT. We have what’s called the agentic commerce protocol. So like new feeds that you can create to get your e-commerce site and your e-commerce products visible in ChatGPT.

35:54 This is absolutely a space that you should pay attention to. It’s something that is absolutely changing Optimizing for AI search.

36:04 So we’re back at the annual Alligators Who Ruined the Internet Gala. The whole family dances, each glowing faintly with their expired hype colors.

36:12 Gio jumps in for a guest DJ set with me, your resident DJ, Lily Ray, and he’s actually quite good at it.

36:18 He mixes Regex Romance into Chunk It Baby one more time. The banner reads, See You in the SEO Swamp, and honestly, it’s iconic.

36:28 The whole SEO family is on the dance floor, each glowing in their signature SEO color like some kind of Power Ranger.

36:35 Featured Snippets is doing the Googlebot, obviously. Voice Search is karaoke-ing, and Mother SEO is pregnant yet again. Plot twist, this time with twins named Quantum Search and Brain Chips.

36:45 The Alligators have ruined the internet yet again, but this time as a family. Nature is healing. And with all that, thank you all for joining me at the SEO Swamp.

36:56 Hope to see you next year.

The SEO landscape is evolving at breakneck speed. Explore our complete guide to visibility in AI search, or let’s talk about how Amsive can help you future-proof your SEO strategy.