---
title: 'AI SEO Course for Beginners: Complete AEO Tutorial'
source: 'https://youtube.com/watch?v=uza9GX0E2mw'
video_id: 'uza9GX0E2mw'
date: 2026-07-14
duration_sec: 5216
---

# AI SEO Course for Beginners: Complete AEO Tutorial

> Source: [AI SEO Course for Beginners: Complete AEO Tutorial](https://youtube.com/watch?v=uza9GX0E2mw)

## Summary

This course by Samo from HRS teaches Answer Engine Optimization (AEO), the practice of making content visible to AI systems like Google AI Overviews, ChatGPT, and Perplexity. It covers how AI search works, how to optimize content for citations, and how to measure results, emphasizing that AEO builds on traditional SEO.

### Key Points

- **AI Overviews Reduce Click-Through Rates** [00:28] — As of December 2025, AI overviews on Google reduce the click-through rate for the #1 ranking page by 58%.
- **AI Traffic Converts 23x Higher** [00:53] — In June 2025, AI search accounted for 0.5% of traffic but drove 12.1% of signups, a 23x higher conversion rate than organic search.
- **Definition of AEO** [01:40] — AEO (Answer Engine Optimization) is the practice of making content visible and useful to AI systems that deliver direct answers, also called GEO or LLMO.
- **AEO Builds on SEO** [02:50] — AEO does not replace SEO; it builds on top of it. SEO is the foundation, and AEO future-proofs online presence.
- **Two Sources of AI Information** [07:13] — AI search engines use training data (static, updated every 6 months) and real-time retrieval (RAG) for fresh information.
- **Query Fan-Out Explained** [09:28] — AI expands one query into many subqueries (fan-out). The average prompt triggers 9-11 fan-out queries, some up to 28.
- **AI Visibility is Probabilistic** [12:40] — AI citations are probabilistic, not fixed. Consensus, freshness, and authority matter. 76% of AI overview citations come from pages already in Google's top 10.
- **Low Overlap Between AI Platforms** [15:57] — Only 7 out of the top 50 most cited domains appear on Google AI Overviews, ChatGPT, and Perplexity (14% overlap).
- **Only 28% of AI Mentions Include Links** [23:07] — On average, only 28% of AI mentions include a link. Perplexity links 51.6%, AI Overviews only 10.7%.
- **Branded Web Mentions Strongest Correlation** [25:06] — Branded web mentions have a 0.664 correlation with AI visibility, stronger than backlinks or domain rating.
- **Content Length Doesn't Matter for AI** [44:34] — Correlation between word count and AI citation is 0.04. Over half of cited pages are under 1,000 words.
- **Freshness is Critical** [45:07] — AI-cited content is 25.7% fresher than traditional results. 76% of ChatGPT's top cited pages were updated within the last 30 days.
- **Listicles and Data-Driven Content Perform Best** [46:04] — 43.8% of ChatGPT citations are listicles. Data-driven content with original stats also gets cited heavily.
- **BLUF Principle for Content** [47:00] — Bottom Line Up Front: start every section with the answer, not the backstory, because AI weighs beginnings and ends more heavily.
- **Three Tiers of Mention Sources** [52:31] — Tier 1: third-party editorial (listicles, reviews). Tier 2: UGC (Reddit, Quora). Tier 3: own properties (YouTube, podcast).
- **YouTube is Most Cited Domain in AI Overviews** [58:32] — YouTube mentions have a 0.737 correlation with ChatGPT visibility. GPT-4 was trained on over a million hours of YouTube transcripts.
- **5.9% of Sites Block GPTBot** [01:03:54] — Around 5.9% of 140 million websites block OpenAI's crawler, making them invisible to ChatGPT.
- **AI Sends Visitors to 404 Pages 2.87x More Often** [01:10:26] — AI assistants send visitors to 404 pages 2.87 times more often than Google. ChatGPT is the biggest offender (1% of clicked URLs lead to 404).
- **AI Traffic Converts at 23x Rate** [01:20:01] — At HRS, AI visitors convert at 23x the rate of organic search visitors. AI traffic has grown 9.7x in the past year.
- **AI Vulnerable to Misinformation** [01:23:44] — After planting fake sources, Gemini and Perplexity repeated misinformation in 37-39% of answers. ChatGPT stayed under 7%.

### Conclusion

AEO is a new layer on top of SEO that focuses on getting mentioned in AI responses. Starting early with a systematic approach—technical checks, content optimization, mention earning, and measurement—can provide a significant competitive advantage as AI search grows.

## Transcript

Hey everyone, my name is Samo and welcome to the  answer engine optimization or AEO course by HRS.   In this course, I'll be teaching you how to get  traffic from AI search and to get more brand   visibility in it. This course will have a heavy  focus on execution. And while the topic might  
be new to many of you, I don't want you to feel  intimidated, especially if you have a background   in SEO because SEO is the foundation of AEO. Now,  here's why you need to pay attention to this right  
now. As of December 2025, the presence of an AI  overview on Google reduces the click-through rate   for the number one ranking page by 58%. That means  for every 100 clicks you used to get, Google now  
keeps 58 of them. Chat GPT alone has 900 million  weekly users. It handles roughly 12% of Google   search volume. And AI traffic to websites has  grown 9.7 times in the past year. So, you can  
either ignore the shift and watch your traffic  shrink while raising your fist at the clouds,   or you can learn how it works and capitalize on  it. Because here's the thing most people miss.   In June 2025, AI search accounted for just 0.5%  of our traffic, but it drove 12.1% of our signups.  
That's a 23 times higher conversion rate than  organic search. Search has changed and it's no   longer just about clicks because AI traffic is  one of the highest converting channels we've   ever seen. So, let's start with the basics. In  this first lesson, I'm going to explain what  
AEO actually is, why it matters, and give you  a road map for the entire course so you know   exactly what's coming. So, what is answer engine  optimization? AEO or answer engine optimization  
is the practice of making your content visible and  useful to AI systems that deliver direct answers.   That's Google AI overviews, ChatGBT, Perplexity,  Gemini, Copilot, basically any platform where  
AI is generating a response and choosing who to  site. You'll also hear this called GEO, generative   engine optimization, or LLMO, large language model  optimization. They all mean essentially the same  
thing. So, we'll use AEO throughout this course.  Now, you might be wondering, how is this different   from traditional SEO? Well, in traditional  SEO, you optimize web pages to rank higher   in a list of search results. You're competing  for positions. The user sees your blue link,  
clicks it, and lands on your site. In AEO, there  is no list. The AI reads from dozens of sources,   synthesizes an answer, and decides who to mention  or site. You're not competing for a position,  
you're competing for a mention. And the rules  for getting mentioned are a bit different   from the rules for ranking in organic blue link  results. Now, we'll break down exactly how that   works in module one. But for now, the key thing  to understand is this. AEO doesn't replace SEO.  
It builds on top of it. Think of SEO as your  foundation and AEO as futureproofing your   online presence. The fundamentals still matter.  quality content, authority, technical health,   but the strategy needs to evolve because  the way people find information is changing.  
Now, I know some of you might be thinking, "Is AI  just going to kill SEO completely?" And I get it,   but this is not the end of SEO. Ethan Smith from  Graphite talked about the zero sum bias on HR's  
podcast, and this is a cognitive bias based on  the assumption that if something new goes up,   something old must go down. And we saw the exact  same thing in 2010 when the app store blew up.  
Everyone said, "The web is going away. Mobile apps  are taking over. Only old people use websites."   And yes, mobile apps did blow up. That promise  was real. But the web didn't die. It grew  
alongside these apps. AI search is likely going  to be the same story. It is growing fast. Those   numbers I just showed you are real. But Google  still processes billions of searches per day.  
Traditional organic traffic still drives massive  business value. So, the smart move isn't to panic,   it's to play both games. And let me be real  with you, most businesses haven't even started  
thinking about this. We looked at HF's own data  and found that our content and product pages have   been mentioned 7,470 times across 2,39 pages in  AI search without any effort to optimize for it.  
That's all happening organically because we've  been doing solid SEO and creating content that   people find useful. Now, imagine what happens when  you actually start intentionally optimizing for   this. When you know which platforms site  what, when you understand how AI decides  
who to recommend when you're actively building the  signals that make AI want to mention your brand.   That's exactly what this course will teach you.  The course is broken down into four modules plus   this introduction. Module one is about how AI  search actually works. We'll cover the mechanics,  
how AI finds content, how it decides what to site,  how different platforms like Chat GBT, Google AI   overviews, and Perplexity actually differ from  each other, and what AI visibility actually looks  
like for different types of businesses. This is  the foundation that makes everything else make   sense. Module two is all about strategy. I'll show  you the data behind why brand mentions are the   single strongest lever for AI visibility, stronger  than backlinks, domain rating, or anything else.  
You'll learn keyword research for AEO, prompt  research, and how to figure out where your brand   stands versus competitors. Module three is about  execution. This is where the actual work happens.  
How to create content that gets cited, how to earn  mentions and citations for other sites, YouTube   optimization for AI visibility, which no other  course covers, and the technical stuff to make   sure AI can actually find and crawl your content.  In module four is about measurement. You'll learn  
how to set up your analytics for AI traffic, how  to run competitive audits to find your brand gaps,   and how to evaluate whether AEO is actually worth  the investment for your business. By the end of   the course, you should be able to put together  a concrete action plan that's bespoke for your  
business with workflows to actually do this stuff  immediately. All right, so that's the lay of the   land. In the next lesson, we're diving straight  into how AI search engines actually find and site  
content, including the concept of query fan out,  which is probably the single most important thing   to understand about why AI search works the way  that it does. I'll see you in the next lesson.   Hey, it's Samo and welcome to the first module  which is on how AI search actually works. In this  
lesson, I'm going to walk you through the three  things you need to understand about how AI search   engines find, evaluate, and site content. Because  here's the thing, if you don't understand how AI   search works under the hood, every optimization  tip you hear is just going to feel like a random  
list of tactics. And I don't want that for  you. I want you to understand the why behind   every strategy we cover in this course. Let's  get started. So where does AI actually get its   information? This is probably the most important  distinction to understand. AI search engines have  
two sources of information and they work very  differently. The first is training data. This   is the massive collection of text that the AI was  originally trained on. So books, websites, PDFs,  
social media, YouTube transcripts, basically a  snapshot of the internet. So when you ask ChatGpt,   "Who is the CEO of Apple?" and it instantly  says Tim Cook without searching anything,  
that's coming from training data. It already  learned that pattern. But here's the problem   with training data. It's static. It gets updated  maybe every 6 months or so. So if you launched   your product last week, the AI doesn't know  about it yet. not from training data at least.  
And that's where the second source comes in, real  time retrieval. This is where rag or retrieval   augmented generation comes into play. It sounds  complicated, but Patrick Stock said it best.   Uh then you've got the the retrieved pages, which  is like a secondary process. So you've got the  
the trained LLM data and then you've got the data  where it goes out and fetches a bunch of relevant   pages and those come with other probabilities.  For example, when ChatGpt or Google's AI mode   needs fresh information, or when the question  is too specific for training data alone,  
it goes out, it searches the web using APIs, it  pulls back a bunch of pages, reads through them,   and then generates a response based on what  it found. Now, why does this matter for you?   Because it means there are two ways to influence  what AI says about your brand. The first is to be  
mentioned so widely across the web that you're  baked into the training data itself. And second   is to make sure your content shows up when the  AI searches the web in real time. And guess what?  
We already know how to do that. That's SEO. The  skills you already have from traditional SEO,   ranking in Google, earning backlinks, creating  quality content, those directly influence   whether AI picks up your pages during real-time  retrieval. Now, here's where it gets interesting.  
The AI doesn't just search for the exact thing you  typed in. Let me explain. Search engines used to   work one to one. One query, one set of results.  Then they evolved to many to one where different   queries like Sydney plumber and plumbing service  in Sydney could return the same results. But AI  
search has flipped the model to one to many. One  search gets expanded into many. And this technique   is called query fo. For example, when someone  enters a prompt like, "Plan me a 5-day trip to  
Japan in November." The AI fans it out into dozens  of smaller longtail subqueries. Things like, "Best   neighborhoods to stay in Tokyo, November weather  in Kyoto, Japan Rail Pass worth it?" All running  
simultaneously behind the scenes. It then pulls  information from multiple sources across the web   and combines it into one complete answer. In  fact, research from Seir Interactive in Nective  
found that the average prompt triggers 9 to 11  fan out queries with some going as high as 28.   And ChatGpt's deep research mode ran 420 searches  for a single query about buying a red phone case.  
So, if your content ranks for those niche specific  queries, your brand has a much better chance   of being included in the AI's final response.  And this is a huge shift from traditional SEO   where you could optimize one page for one target  keyword and call it a day. In AI search, you need  
to be relevant across an entire topic. And I could  even argue across an entire niche. Because if your   page about how to start a podcast only covers the  basics, but doesn't mention equipment, hosting, or  
promotion, the AI is going to find someone else's  page that does. Now you might be wondering, can I   see these fanout queries? You sure can. In the AI  responses report in HFS brand radar, you can see  
the fan out queries for chat GPT and perplexity  prompts. But there's an important caveat. Despina   from HFS wrote in her guide on query fan out that  these aren't like traditional longtail queries.  
They're synthetic, generated by AI in the moment.  They're inconsistent. The same prompt can trigger   different fanouts every time. And over 95% of  them have zero search volume because real humans  
would never type them. So don't think of fano  queries as a new keyword list to optimize for.   Think of them as a window into what topics the  AI considers important for a given question.  
We'll get into exactly how to use this  strategically in module 2 because for now   we need to talk about how AI decides who to  actually site. In traditional search rankings   are relatively stable. Like if you're number three  for a keyword today, you're probably going to be  
somewhere around there tomorrow. But AI citations  are probabilistic. Patrick Stocks explained this   really well. He said, "AI outputs are built on  probabilities on top of probabilities on top of   probabilities. The training data creates patterns.  The retrieve pages add their own signals and then  
there's a temperature setting that introduces  randomness so the AI doesn't generate the exact   same answer every time." Now, what that means in  practice is that if you ask the same question five  
times, you might get cited three out of five. Or  the AI might mention your competitor twice and   you twice and someone else once. There's no fixed  position to rank for. This is why we talk about AI  
visibility rather than AI rankings. It's more like  a probability distribution than a leaderboard.   Now, that said, there are patterns in what gets  cited more often. Based on the data we've studied   at HRES, consensus matters. If multiple sources on  the web say the same thing about your brand, AI is  
more likely to repeat it. And the more places your  brand is mentioned in a consistent way, the higher   the probability the AI picks it up. Freshness  matters. AI cited content tends to be about 25%  
fresher than what you'd see in a traditional SER.  The AI is actively looking for recent information,   especially for topics that change. Authority  still matters. Pages that rank well in traditional  
search have a major head start. Our data shows  that 76% of AI overview citations come from pages   already in the top 10 of Google. So, I'll hit that  gong one more time. SEO is the foundation of AEO,  
but it's not only about Google rankings. 14% of  pages cited in AI overviews don't rank in Google's   top 100 at all. And for platforms like Chad GBT,  the overlap with Google's results is even lower.  
So there's real opportunity for brands that aren't  dominant in traditional search to still show up   in AI. So let's tie all of this together.  AI search engines pull from training data   and real time web search. They expand your one  query into dozens of subqueries through fan out.  
They merge and score results from all of  those searches. and then they generate a   response that's probabilistic, not fixed, based  on patterns, consensus, freshness, and authority.   Understanding this process is what makes every  other lesson in this course make sense. So when I  
tell you to earn more brand mentions, it's because  of how consensus drives probability. When I talk   about topic coverage, it's because of query fan  out. When I say that SEO is the foundation again,  
it's because 76% of citations come from pages  already ranking well. Now that you understand the   mechanics, the obvious next question is, do all  AI platforms work the same way? And the answer is  
not even close. Not even close. In the next  lesson, we're going to compare AI overviews,   chatbt, Perplexity, and Google's AI mode side  by side. And the data on how different they   are is pretty surprising. I'll see you in the  next lesson. Hey, it's Samo and welcome to the  
second lesson which is on the differences between  AI search platforms. Now, in the last lesson,   I covered how AI search engines find and site  content. But here's what I didn't tell you.   They don't all do it the same way. And this  matters a lot more than most people realize.  
Because if you're treating AI search as one  thing and optimizing for it the same way across   the board, you're likely leaving visibility on  the table. So in this lesson, I'm going to break   down the major AI search platforms like Google's  [clears throat] AI overviews, Google's AI mode,  
ChatGpt, and Perplexity, and show you the data  on just how different optimization for these   platforms are. Let me start with a stat that  might surprise you. We looked at the top 50   most cited domains across Google AI overviews,  ChatgBT, and Perplexity. And out of those 50,  
only seven appeared on all three platforms. That's  only a 14% overlap. Think about that for a second.   If you're only optimizing for one platform,  you're potentially invisible on the others.  
And it gets even more interesting when you look  at what each platform actually prefers. Google AI   overviews favor authoritative established sites.  Think health, finance, encyclopedic content,   and Google owned properties like YouTube. In fact,  YouTube accounts for about 5.6% of all AI overview  
citations. And just look at how fast they're  adding YouTube videos to their AI overviews.   And obviously, Reddit's big here, too. ChatGpt  leans heavily toward publishers and media.   For example, authoritative sites like Reddit,  Wikipedia, Amazon, Forbes, Business Insider,  
and Wired are some of the most cited domains  here. In fact, the median domain rating of   Chat GBPT's top-sighted pages is 90. So, it's  pulling from high authority publishers, and  
that's partly because of licensing deals OpenAI  has with some of these outlets. Perplexity is   actually the most aligned with traditional Google  search. About 28.6% 6% of perplexity citations  
come from pages that rank in Google's top 10.  Compare that to Chat GBT, which only overlaps   with Google's top 10 about 8 to 10% of the time.  So, if you're already ranking well in Google,   Perplexity is probably where you'll see the most  immediate AI visibility. And then there's Google's  
AI mode. Now, you might assume that since AI  mode and AI overviews are both Google products,   they'd site similar sources. Well, they don't.  The citation overlap between AI overviews and  
AI mode is only 13.7%. Despite the fact that  their answers are 86% semantically similar,   meaning they're saying similar things but  pulling from completely different sources.  
AI mode's top-sighted domain is YouTube by a  wide margin, followed by Google and Wikipedia.   We also found that it cites Quora 3.5 times more  than AI overviews and pulls from social platforms  
like Facebook and Instagram much more heavily. So  the takeaway here is clear. AI search is not one   thing. Each platform has its own index, its own  biases, and its own preferences for what kind of  
sources it likes to site. So which platform should  you focus on? This is the practical question and I   think it's important to be strategic here rather  than trying to optimize for everything at once.  
There are really two things to consider.  The first is market share. As of right now,   Google's AI features and chat GBT account  for the vast majority of AI search traffic.   Proplexity is growing, but it's still a fraction  of the volume. So, if you have to prioritize,  
Google and chat GBT are where most of the eyeballs  are, at least for now. The second thing to   consider is how much overlap there is between what  you're already doing in SEO and what each platform   rewards. If you're already ranking well in Google,  you have a natural head start with AI overviews  
and Plexity. So, why not lean into it? 76% of AI  overview citations used to come from pages already   in Google's top 10. And I say used to because that  number has dropped. A more recent study shows it's  
now closer to 38%. Which is still a lot. AI  overviews are increasingly pulling from pages   outside the top 10, including YouTube and Reddit.  So, the connection between Google rankings and AI   citations is weakening. Chat GBT is much more  interested in publisher authority and editorial  
content. So, if your brand gets mentioned  on Forbes, in a Reddit thread, or on a niche   review site, that might matter more for ChatGpt  visibility than your own pages Google ranking.  
And here's something a lot of people miss. Many of  the top sighted domains in AI search don't get any   traditional search traffic at all. AI visibility  is its own game. Now, if you want to see the top  
domains for a specific platform, you can check  that in HR brand radar. Just run a blank search   and go to the cited domains report. Then you  can filter using your desired AI platform.   What you'll typically see is that the websites  citing your brand are different on each platform.  
A domain that shows up heavily in AI overviews  might barely appear in ChatGBT and vice versa.   This is why a one-sizefits-all approach to AEO  doesn't work. Now, the good news is you don't   need to build completely separate strategies  for each platform. A lot of the fundamentals,  
creating quality content, earning mentions,  building topical authority help across the   board. But knowing where each platform pulls from  helps you prioritize your efforts. For example,  
if chat GBT is a big traffic driver for your  niche, you'd want to focus on getting mentioned   on high DR editorial sites and publishers. If AI  overviews matter more, YouTube and Reddit should   be on your radar. We'll get into these specific  strategies in modules 2 and three. But for now,  
the key thing to understand is that each platform  is its own ecosystem with its own rules. Now,   before we actually start optimizing for AI search,  there's something else you need to understand.  
What does winning an AI search even look like?  Because it's not just about getting a link or   a mention. So, in the next lesson, I'm going to  break down the different types of AI visibility   and why some of them might be more valuable to  you than others. See you in the next lesson.  
Hey, it's Samo and welcome to the third lesson  in module one. Now, in the last two lessons,   I covered how AI search engines find and site  content and how each platform works differently.   But there's a question we haven't answered yet.  What does winning in AI search actually look like?  
Because most people hear AI visibility,  and they think it means one thing,   getting a link in an AI response. But that's  just one piece of the puzzle. And honestly,   it might not even be the most valuable  piece for your business. So, in this lesson,  
I'm going to break down the different types  of AI visibility, show you the data on how   often each one actually happens, and help you  figure out which type matters most for you.   Let me walk you through the three types of AI  visibility, and I want you to think about which  
one applies to your business as I go through each.  The first is cited and linked. This is the one   everyone thinks about. AI includes a link to your  page and its sources. The user can click through   and land on your site. This is the best case for  direct traffic and it's the easiest to measure.  
The second is mentioned but not linked.  AI says your brand in its response but it   doesn't give the user a link to click. You don't  necessarily get the direct traffic or any traffic   but the user now knows your name. And if  they're interested, what are they going to do?  
They're going to search for you. This is  basically a word of mouth recommendation at scale.   And the third is not visible at all. AI doesn't  mention you. It doesn't link to you. you're just   not in the conversation. And this is actually  important to know because you can't fix what you  
don't know is broken. Now, here's the thing. Most  people assume that when AI mentions your brand,   it's probably including a link. Well, not exactly.  On average, only about 28% of AI mentions include  
a link. That means roughly seven out of 10 times  your brand comes up in an AI response, there's no   link attached. And it varies a lot by platform.  Perplexity is the most generous, about 51.6 6% of  
mentions get a link. AI mode is around 36.8%.  Chat GBT is at 26.9% and AI overviews only 10.7%.   So if you're mostly tracking AI overviews, almost  nine out of 10 times your brand shows up, there's  
no link. Now, you might think that means most AI  visibility is low value since there's no click.   But here's something that might surprise you.  When we weighted these mentions by search volume,   basically how many people are likely asking  similar questions, the ones that do include a link  
tend to appear on much higher traffic queries. For  example, on Perplexity, links show up in about 78%   of total impressions, even though they're only in  51% of individual mentions. On Gemini, it's even  
more dramatic. Links appear in 71% of impressions  despite being in only 16.8% 8% of mentions.   So, the takeaway here is this. Citations  are relatively rare, but when they happen,  
they tend to happen on the ones with the most  eyeballs. But that doesn't mean unlin mentions   are worthless. Far from it. Think about how LLMs  actually work. They learn by reading the web.   And every time your brand name appears on a  credible site connected to a specific topic,  
that becomes another training example. The more  the model sees your brand associated with a topic,   the more confidently it mentions you when someone  asks about that topic. It's like when you hear   peanut butter, you think jelly. Or when you hear  Tesla, you think electric cars. LMS build these  
same associations. And unlin mentions are what  feed them. And there's real data to back this up.   In a study of 75,000 brands, branded web mentions  had the strongest correlation with AI visibility,  
a 0.664 664 correlation with showing up in  AI overviews that's stronger than backlinks,   domain rating, referring domains,  or any other traditional SEO metric.   So even when AI doesn't give you a link, just  being mentioned is building your brand's presence  
in the model. It's the long game and it compounds  over time. Now, there's another layer to this.   Not all AI responses look the same and the  type of response matters for your business.  
For example, some queries trigger step-by-step  guides like how to fix a leaky faucet,   how to set up Google Analytics, and if you're  a service-based business or you create how-to   content, these can be opportunities if you can  get AI to recommend you as an expert. Then you  
have direct answers like what's the capital of  France? If you're a publisher orformational brand,   being the source AI sites for factual answers  can build massive authority, but you probably  
won't get a click from these since AI gives the  answer directly. And then there's video citations.   We covered this in the last lesson. YouTube is  one of the most cited domains across AI platforms.   And if you're a creator, your videos can show up  directly in AI responses. And some people prefer  
watching videos over reading text. So here's  what I want you to take away from this lesson.   AI visibility isn't binary. You're not  just visible or invisible. It's a spectrum.  
And the type of visibility that matters most  depends entirely on your business. If you sell   a product, you want your brand showing up when  people ask about your category. For example,   a query like best golf balls surfaces tons  of different brands and models and someone  
can literally click through from the AI response,  land on a shopping result or the manufacturer site   and buy something right away. That's a citation  doing real work. If you're a publisher or   content creator, training data visibility  is your long gate. Every article AI cites,  
every video it surfaces reinforces your authority  in the model. And if you're not yet visible, well,   that's why you're taking this course. Now, if you  want to see where you stand in AI search, go to  
HS brand radar and search for your website. And  right away, you'll see how you stack up against   competitors across different AI platforms. And you  can filter by mentions, which are when your brand   is named but not linked. citations, which is when  AI actually links to your site, impressions, and  
share a voice. The gap between your impressions  and your mentions is your opportunity. If AI is   responding to queries about your topic, but not  naming you, that's exactly where you need to   focus. And that's exactly what we'll be covering  in module 2, where we'll cover AEO, strategy,  
research, targeting, and gaps. I'll see you  there. Hey, it's Ammo and welcome to the second   module in our AI search course. Now, in module  one, we covered how AI search actually works,   the mechanics, the platforms, and what winning AI  visibility looks like. In this module, I'm going  
to walk you through an AEO strategy that works  in two phases. First, we'll assess where your   brand stands today, figuring out exactly where  the gaps are. Then, we'll move on to discovery,   where we'll find the keywords and prompts people  use to get information related to your business.  
So, let's start with step one, assess. In  this lesson, I'm going to walk you through   how to run a brand gap analysis. A brand gap  analysis measures the difference between where   your brand should be showing up and where it  actually is. So, this is in Google search,  
AI results, and basically across the web. It  examines everything shaping your discoverability   and reputation from how AI describes you to  which competitors are cited instead of you. So,  
the first step in this analysis is to map out your  branded entities. Before you can find your gaps,   you need to get clear on what you're actually  measuring. And here's the thing, your brand gets   referred to differently by different people.  So, you want to map out your main brand name,  
any subbrands, product names, proprietary  features, proprietary metrics, and even personal   brands associated with your company. For example,  at Hrefs, we'd map out the main brand, Hrefs,  
product brands like Site Explorer, Brand Radar,  and AI content helper, proprietary metrics like   domain rating and traffic potential, and personal  brands like Tim Solo, Patrick Stocks, Ryan Law,  
and Glenn Alaw. Each of these has its own  visibility profile in search and AI results.   And you can repeat the analysis I'm about to show  you for each one. Once you have your list, connect   each entity to the topics and attributes people  should associate with it. Search engines and LLMs  
don't understand brand names on their own. They  infer meaning from how your brand is described and   discussed. So, you need to clarify what problems  you solve, what qualities you're known for,   and what context you belong in. A quick way to  do this is keyword research. Look for recurring  
adjectives, modifiers, and descriptive  phrases people use alongside your brand   or category. things like affordable, AI powered,  enterprisegrain, whatever applies to your space.   This gives you your benchmark for what your brand  should be known and found for. Now, let's move on  
to the second step, which is to run your audit in  hres. Start by entering your brand's website into   site explorer. You'll get a dashboard of baseline  metrics, domain rating, referring domains,   organic keywords, organic traffic, and traffic  value. This gives you your traditional SEO  
snapshot. But what we really care about in the  context of AEO are the AI citation metrics which   give you a snapshot of your brand's visibility  in AI search across different platforms.   Clicking into any of those takes you to the brand  radar report for that platform which is where the  
real audit happens and there are four key things  to pay attention to. First are mentions which are   the number of times your brand is mentioned  in AI responses. Then we have citations which   are the number of times your website is actually  cited as a source. impressions, an estimation of  
exposure based on how often responses containing  your brand are shown, and AI share a voice,   how often your brand is mentioned compared to your  competitors. Now, the real power of Brand Radar is  
in its filters. This is where you can isolate the  exact scenarios that matter. For example, you can   filter for prompts that mention your brand name  in the AI platform of choice. You can filter for   responses that mention your brand but don't site  your website. Those are miscitation opportunities.  
You can also add a competitor and then filter for  queries where they're mentioned but you're not.   And you can look at specific topics to see  whether AI associates them with your brand   or someone else's. Now take note of these numbers  and let's move on to the third step which is to  
identify your gaps. When you're looking at this  data, it helps to think about your gaps across six   dimensions. And this is a framework Despina from  Atro's laid out in her brand gap analysis guide.   The first is your visibility gap. This is where  your brand appears less often than competitors in  
search or AI results. The second is your narrative  gap. This is how AI or media describes your brand   versus how you actually want to be positioned.  For example, maybe you're a premium tool, but AI  
keeps calling you a budget alternative or your key  differentiator just isn't coming through. Third   is your topic gap. These are topics you should  be associated with but aren't. Like if you're a   project management tool, but AI never mentions you  when people ask about remote team collaboration.  
That would be a topic gap. Fourth is your format  gap. AI tends to site certain types of content,   guides, videos, reviews, comparison pages, and  if you're not producing them, that's a gap. For  
example, in module one, we talked about YouTube  being one of the most cited domains. And if your   competitors have YouTube content getting cited and  you don't, then you've got a content format gap.   Fifth is your web mentions gap. These are  external sources, listicles, review sites,  
forums, publications that mention competitors  but not you. And as we covered in module one,   those third party mentions are one of the  strongest signals for AI visibility. And sixth  
is your demand gap. These are branded queries or  searches that signal awareness opportunities you   haven't captured. People are searching for things  in your space, but your brand name never comes up   alongside those searches. Now, when you look at  all six together, you get a complete picture of  
your brand's visibility in AI. Not just how often  you show up, but whether you're showing up for the   right things in the right way in the right places.  So, as you do your brand audit in step two, it's  
worth adding your prompts to these buckets. And  it'll set you up perfectly for the fourth step,   which is to prioritize. You're going to find a lot  of gaps, and you can't fix everything at once. So,   you're going to have to prioritize your efforts.  You're generally going to do one of three things.  
Fix, which is improving or optimizing something  that already exists. you're going to build, which   is creating new content or pages for opportunities  you're not covering at all, or influence, which  
is strengthening your off-site visibility through  outreach and brand mentions. And you want to weigh   each opportunity by asking, "How much demand  could this drive? Does it support your brand   credibility? And does it improve your chances of  being cited in AI?" Start with the quick ones.  
If you have a page that's already ranking but  just needs a content update to close a topic gap,   that's low effort and potentially high impact. If  you're missing from a major listical that all your   competitors are on, that's a web meions gap you  can close with outreach. Now, as you go through  
this process, you'll probably find that most of  your gaps fall into one or two of the buckets I   mentioned before. And that's a good thing because  it tells you exactly where to focus so you can   create systems around them. and I'll walk you  through how to close each type of gap in module 3.  
From creating content that gets cited to earning  mentions from thirdparty sites to optimizing your   YouTube presence. So the audit you do right now,  it becomes your personalized road map for the rest   of this course. Now, one more thing, you can do  this exact same audit for your competitors to get  
a wealth of insights on new topics and gaps  that will be relevant to your business. Look   at their branded queries to see what users and AI  connect to them more strongly than they do to you.  
Sometimes those are quick wins. You might already  have the features, you just haven't created the   content that connects them to your brand, or you  don't have enough pages that are talking about   that feature. We've put together a template  you can use to organize all of this and share  
it with stakeholders. So, I'll link that in the  description. Now, you have your baseline saved and   a clear picture of where the gaps are. But knowing  your gaps is only the first step. You also need to   know what opportunities are out there. And keyword  and prompt research for AEO is how you do it,  
but it's a bit different from what you might be  used to. There are queries where AI dominates,   queries where traditional SEO still wins,  and a whole set of platforms most people   aren't even thinking about. That's what we'll  cover in the next lesson. I'll see you there.  
Hey, it's Ammo and welcome to the second lesson,  which is on keyword and prompt research for AEO.   Now, in the last lesson, I walked you through how  to run a brand gap analysis, so you should have a   clear picture of where your brand is showing up  and where it isn't. Now, it's time to find the  
keywords and prompts to actually go after. And  in this lesson, I'm going to walk you through   the full keyword research process for both SEO and  AEO. Why? Because there's a ton of overlap between   the two. And half the battle is really identifying  which is which and knowing how to approach them  
differently. So, the first step is to build  your keyword list. And this part honestly hasn't   changed that much. You still need two things. Seed  keywords, which are broad terms related to your   niche, and modifiers, which are add-ons like best  or howto that turn those seeds into real searches.  
Now, a quick way to come up with these is to just  ask your AI assistant of choice something like,   I'm doing keyword research for my type of site,  which makes money through my revenue model.   My target audience is this group. Give me 10  seed keywords that are one to two words max  
and five plus modifiers that will help  me surface appropriate content formats   I can use in my keyword research. The seeds  and modifiers should not share the same words.   And just like that, you've got a solid  list of seeds and modifiers to work with,  
but these are just the starting point. Take your  seeds and paste them into keywords explorer.   Then go to the matching terms report and add  your modifiers using the include filter. And just   like that, you should have hundreds, maybe even  thousands of real keyword ideas your audience is  
actually typing into Google. Now, the second step  is to vet those keywords because some of them are   a trap. Some of these keywords are going to look  incredibly enticing. High volume, high traffic   potential, low difficulty scores, all the right  metrics, but not all of them are worth targeting.  
Before you commit to any keyword, it needs to  pass three tests. I call this the bid formula.   B is for business potential. Ask yourself, if  I rank number one for this keyword, does it   actually help my business? A keyword like what  is espresso has solid volume and low difficulty,  
but someone searching that isn't looking to  buy anything, maybe ever. Compare that to best   espresso machine under $500, where the searcher  is showing intent and has a budget. Always choose  
keywords that move the needle. I is for intent.  Google the keyword and look at what's actually   ranking. If every top result is an e-commerce  page and you're trying to rank a blog post,   it's not going to happen. The SER tells you what  searchers want. Match the intent or move on. And  
D is for difficulty. You need to choose keywords  you actually have a chance at ranking for. Check   the referring domains and domain rating of the top  ranking pages. In general, the more links and the   higher the DR, the tougher the competition. If  you see a few low DR sites in the top 10, that's  
usually a good sign. If a keyword passes all three  tests, you should consider targeting it. Unless,   and this is where step three comes in, the  AI filter. Before you commit to any keyword,  
there's one more question you need to ask.  Can AI fully satisfy the user for this query?   Because even if a keyword passes bid, if the  AI overview is so good, there's no reason for  
anyone to click through, that keyword might  be a trap. And there's hard data behind this.   AI overviews appear on about 21% of all keywords,  but for informational queries, it's much higher.  
Nearly 58% for question queries, 46% for queries  with seven or more words, and 99.9% of keywords   that trigger AI overviews areformational in  intent. So before you commit to a keyword,  
Google it. Look at what shows up. Put yourself in  the searcher shoe and ask, "Am I satisfied with   this answer or do I need to click somewhere  to learn more?" If the AI overview nails it,  
that keyword might not be worth targeting the  traditional way. But here's the good news. Even   though AI is eating clicks for a ton offormational  keywords, there's a whole category of searches   AI hasn't touched. Free tools, search backlink  checker, no AI overview, mortgage calculator,  
nothing. Word counter, nothing. Why? Because when  someone searches for a tool, they need to actually   use something. AI can't replace that yet. To find  these opportunities, go back to keywords explorer  
with your broad seeds. Head to the matching terms  report and add modifiers like calculator, checker,   generator, tool, template, finder, planner, and  maker. These are all actionoriented queries where  
someone needs to do something. AI can't satisfy  that yet. So the organic click is still there for   the taking. You can also filter for transactional  intent directly. Just choose transactional in the   intent filter and you'll get queries where people  are looking to buy, sign up or take some kind of  
action. Now the fourth step is to find your  AI mention opportunities. A minute ago I said   if the AI overview nails the answer, that keyword  might not be worth targeting the traditional way.  
But that doesn't mean you ignore it. These  keywords just need to be targeted differently.   And that's where AEO comes in. Instead of trying  to rank and earn a click, your goal is to get your   brand mentioned in the AI response itself. And to  do that, you need to know which queries to focus  
on and which pages AI is pulling from. Now,  it helps to know what AI is actually citing.   We studied this across AI overviews and chat  GPT and found that 43.8% of all cited pages  
are listicles. And in my opinion, these pages show  up so often because they help AI build consensus.   So if your brand is mentioned across multiple  lists, that's multiple sources recommending you  
and AI picks up on that. So to find  the queries that matter for your brand,   go to brand radar and enter your website. Find  your brand, hover over the AI platform you   want to research and click others only. This  shows you that mention gaps where competitors  
are showing up but you're not. Then filter for  queries containing best, top, versus, review,   or alternative. And you'll see the queries where  AI is mentioning your competitors but not you,  
and the ones AI is most likely pulling from when  looking for brands in your space. This is your   short list. Now, one thing to keep in mind, over  45% of citations change when AI overviews refresh,  
and that happens on average every 2 days. So, this  isn't a one-time exercise. It's worth revisiting   this report regularly to catch new queries as they  come in. And that brings us to the fifth step,  
prompt research. When someone opens  Chat GBT, Google AI mode, or Perplexity,   they're not typing keywords. They're having  conversations. They're saying things like,   "I'm a small agency owner looking for a marketing  platform. Which one should I choose?" Or,  
"What's the best way to track rankings if I'm  just getting started?" They use natural language   with full context and so every person phrases  it differently. So you can't approach this the   way you would keyword research where you find the  exact terms and go after them one by one. With AI,  
the same question asked 10 different ways can  get 10 different answers with 10 different brands   mentioned. And on top of that, AI fans each prompt  out into many sub queries behind the scenes,  
most of which have zero search volume and will  never repeat. If you're invisible for a topic,   it's not about optimizing for one specific prompt.  It's about building visibility across that entire  
topic, which is exactly what we'll cover in  module three. So, with your topic list and   your prompt gaps mapped out, let's go and close  those gaps. I'll see you in the next lesson.   Hey, it's Ammo and welcome to the third module  in our AI search course. In modules one and two,  
we covered how AI search works and how to  find the keywords and prompts to go after.   Now, it's time to actually execute. And that's  what this module is all about. And in this first   lesson, we're going to be talking about creating  content that gets cited. And I've got some data  
on what AI actually sites to back our strategy.  First, content length doesn't matter. We analyze   over 174,000 pages cited in AI overviews and the  correlation between word count and getting cited  
is just 0.04. That's basically zero. Over half  of all cited pages, 53.4% are under a,000 words.   So, if you've been writing 3,000word articles  because you think longer content ranks better,  
that doesn't apply here with AEO. AI  doesn't care how long your page is.   It cares whether your page answers the  question. Second, freshness matters a lot.   Content cited by AI is on average 25.7% fresher  than what ranks in traditional organic results.  
And when you look at chat GBT specifically, 89.7%  of its top cited pages were updated in 2025.   And get this, 76% were refreshed within the  last 30 days. Think about that for a second.  
If your content hasn't been touched in 6 months,  you're already at a disadvantage for AI citations.   Now, this doesn't mean you should just change the  publish date and call it a day. Google can detect  
that. You need to make meaningful updates to the  actual content. We'll talk about how to do that   later in this lesson. And the third data point is  that format matters. If you remember from the last   module, we talked about how 43.8% of all cited  pages by chatbt are listicles. Best X, top X,  
comparisons, reviews. That's not a coincidence.  These formats are built for AI because they give   clear, structured recommendations that AI can pull  from. But lists aren't the only format that work.  
Datadriven content with original stats also get  cited heavily because AI loves citing specific   numbers. And comparison content like X versus  Y pages perform well because they map directly  
to how people ask AI questions. So, the key  takeaway here is simple. Keep your content fresh,   keep it focused, and lean into formats that AI  can actually use. Now, let's talk about how to  
actually structure your content so AI can use it.  And here's something a lot of people get wrong.   They think they need to write differently for  AI. Like, there's some special AI optimized   format they need to follow. There isn't. You're  writing for humans. AI is trained on what humans  
find valuable. So content that serves human  readers well is content AI wants to site.   That said, there are a few principles that help  both humans and AI get more out of your content.  
The first one is bluff, which stands for bottom  line upfront. This one's borrowed from military   communication, and it's simple. Start every  section with the answer, not the backstory.  
For example, instead of opening a section with,  "Over the past few years, link building strategies   have evolved significantly due to changes in  how search engines evaluate link quality,"   you could write, "The most effective way to build  backlinks in 2026 is to create original research."  
Now, here's why this matters. When humans scan  a page, they follow what's called an F pattern.   They read the beginning closely, skim the middle,  and maybe pick up again at the end. LLM show a  
similar pattern. They weigh the beginning and  end of a passage more heavily than the middle.   So if your key point is buried three paragraphs  into a section, both humans and AI might miss it.  
So put the answer first, then support it with  context and examples. The second principle to   apply is atomic content. This means every section  on your page should be able to stand on its own.  
So, when editing your content, ask yourself,  if AI pulls just one section from your article,   does it still make sense, or does it depend  on context from three paragraphs earlier?   This matters because AI systems chunk your  content into pieces when they process it.  
Different AI models chunk content differently,  and you can't control how they do it. But if every   section is self-contained, it doesn't matter where  the chunks fall because the meaning survives. A  
good test you can do is to take any H2 section on  your page and read it completely out of context.   If it doesn't make sense without the rest of the  page, rewrite it so it does. The third principle   is entity rich writing. AI understands text by  looking at entities and the relationships between  
them. Entities are things like brands, products,  people, places, and specific concepts. So instead   of writing this tool helps with SEO, write HF's  keywords explorer helps you find keywords with  
low difficulty and high traffic potential.  By providing more entities, relationships,   and context, AI has more to work with. And  this also ties into how AI picks what to site.  
The more specific and concrete your content  is, the more useful it is when AI is trying   to answer a specific question. And the fourth  principle is to keep it simple and declarative.   Use short sentences and clear subject verb object  structure. Basically, try to keep things to one  
idea per sentence. This isn't about dumbing down  your content. It's about making it easy to parse   for both people and AI. If a sentence takes two  reads to understand, it's probably too complex.  
Simplify it. Your readers will thank you and  AI will have an easier time extracting the key   points. Now, these four principles will help you  create content that's more likely to get cited.   But there are a couple more things worth  knowing that can push your chances even further.  
And the first is something most people don't think  about. LMS have a tendency to flatten originality.   Like if you come up with an original concept  or framework and you're the only one talking   about it, AI will often absorb the idea without  crediting you. It becomes part of its general  
knowledge. The way around this is to label your  ideas with your brand name. For example, instead   of calling something a content scoring matrix,  call it the HF's content scoring matrix or your   brand content scoring matrix. Define it explicitly  and then distribute it widely across your blog,  
social media channels, podcasts, Reddit, etc. The  more places it shows up with your name attached,   the harder it is for AI to flatten it into  generic knowledge. Now, another thing you  
can do is refresh what I call sleeper pages. These  are pages on your site that used to rank well but   have slowly declined over time. They already have  the back links. They already have the authority.  
They just need a refresh. And because freshness  is such a strong signal for AI citations,   updating these pages can be one of the fastest  ways to gain AI visibility. Here's how to find   them. Go to site explorer and enter your domain  and head to the top pages report. Then sort by  
traffic change and look for pages with significant  declines. What you're looking for are pages that   have two things. A decent number of backlinks,  which means they already have authority,   and a clear traffic decline, which usually means  the content has gone stale. Now, before you update  
anything, make sure it's actually a content issue  and not a backlinks issue. If the page never had   many referring domains to begin with, updating  the content probably won't help. But if it has   links and the content is just outdated, that's a  high potential opportunity. So now you know how to  
create and update content that AI actually wants  to site. But creating great content is only half   the equation. You also need to get mentioned on  other people's content, the pages AI is already   pulling from. And that's exactly what we'll  cover in the next lesson. I'll see you there.  
Hey, it's Samo and welcome to the second lesson  in this module, which is on earning mentions   and citations. Now, in the last lesson, we talked  about creating content that gets cited. But here's   the thing. A lot of AEO isn't about what's on  your site. It's about getting mentioned on other  
people's pages. And not just any pages, the ones  AI wants to pull from. Branded web mentions are so   important that in our study of 75,000 brands, they  had the strongest correlation with visibility in  
Google's AI overviews, stronger than backlinks,  referring domains, and even domain rating. So,   in this lesson, I'm going to show you the three  types of pages you should be getting mentioned on   and how to find the ones that matter most. Let's  get started. So, I like to think about mention  
sources in three tiers. And the reason I use tiers  is because not all mentions are created equal.   Where you get mentioned matters just as much as  how often you get mentioned. According to our   data, we found that getting mentioned on highly  linked pages has a 0.7 correlation with appearing  
in Google's AI overviews. That's even higher  than the general branded mentions correlation.   So, the quality of the page you're on makes a  huge difference. Now, let's talk about tier one   mentions, which are found in third-party editorial  content. These are the hardest mentions to earn,  
but they're also the most valuable. Think  industry publications, review sites like   Wire Cutter or Techraar, listicles and comparison  posts on authoritative blogs, and YouTube reviews   from creators in your space. Why are these so  powerful? because they're exactly the kind of  
pages AI loves the site. Remember, 43.8% of Chat  GPT citations are listicles and comparison pages.   So, if you're mentioned on those pages, you're in  the pool of content AI is already pulling from.  
Now, to find the right pages to go after,  go to brand radar and enter your domain.   Then, open the cited domains report. This shows  you the top websites that AI is citing for topics   related to your brand. You'll see the number of  AI responses, the number of pages, and how often  
each brand is being mentioned. From here, you can  identify which sites matter most for your niche.   Maybe it's kbb.com for cars or cnet for tech  or specific niche blogs in your industry. These  
are the sites where getting a mention can directly  lead to AI visibility. Now, it's important to note   that you don't have to wait for AI to start citing  a page before you try to get on it. Instead, look  
for pages that already have a lot of links and  cover your topic. Even if they're not cited now,   there's a good chance that they'll be cited at  some point. You can use content explorer for this.   Search for your title and add title colon best  or title colon versus along with a minus operator  
before your brand name. This will surface  listicles and comparison posts in your niche   where your brand isn't mentioned. Then filter for  pages with a good number of referring domains.   Next up are tier 2 mentions which can be found on  user generated content and community platforms.  
This includes Reddit, Quora,  niche forums and community sites,   and these matter more than you might think. Reddit  is one of the most frequently cited sources by   Chat GBT. It's also one of the foundational  training sources for large language models.  
The play here isn't to spam Reddit threads with  your brand name. That'll backfire fast. Instead,   find threads where people are asking questions  your product or expertise can genuinely answer.  
If someone's asking what's the best tool for  X and your product is actually a good fit,   contribute a real answer. You can find these  threads through brand radar as well. Look at   the cited domains report and see if reddit.com  shows up for your space. If it does, dig into  
which specific threads AI is pulling from and  join the conversations that are already happening.   Now, if you want to find niche specific Reddit  pages that are already ranking well in Google,   you can go to site explorer, enter reddit.com,  and open the organic keywords report.  
Filter for keywords ranking in the top five,  and add your niche terms to the include filter.   And now you've got a list of Reddit pages that are  already ranking well in Google and are related to   your space. And you can do the exact same thing  for niche forums and Q&A sites. Wherever your  
target audience is having conversations about  problems that you solve, that's where you want   to be present. Finally, we have tier three, which  are your own properties. Some brands own multiple   domains. At for example, has detail.com  and blogger genen. If these properties  
are authoritative in their own right, they can  serve as additional citation sources for AI.   Now, even though you don't have the resources to  run multiple sites, the principle still applies.   Your YouTube channel, your podcast, your LinkedIn  content, these are all indexed and can all show  
up as sources AI pulls from. Basically, the more  places your brand appears in a positive topically   relevant way, the more training examples AI has  to learn from. Now, earning mentions is one thing,  
but you also need to keep track of them because  mentions can disappear. pages get updated, lists   get refreshed, and your brand can get removed  without you ever knowing. And on the flip side,  
sometimes AI picks up wrong information about  your brand from outdated or inaccurate sources.   So, it's worth doing a regular mention audit. In  Brand Radar, you can track your overall mention   trends over time. Look for any drops in mention  volume. If you find a page that was previously  
mentioning you, but your brand was removed during  an update, that might be worth investigating. And   pay attention to sentiment, too. Are AI responses  about your brand accurate? Are they positive?   If you spot misinformation, your best move is  to update your own content first and then reach  
out to the publisher to request a correction.  The faster you fix inaccurate mentions, the   less time AI has to learn from them. So, between  this lesson and the last one, you've got a clear   playbook for earning AI visibility through content  and mentions. But there's one platform we haven't  
talked about yet that deserves its own lesson, and  that's YouTube. I'll see you in the next lesson.   Hey, it's Ammo and welcome to lesson three, which  is on YouTube optimization for AI visibility. Now,   in the last two lessons, we covered how to create  content that gets cited and how to earn mentions  
on other people's pages. But there's one platform  that deserves its own conversation, and that's   YouTube. Why? Because YouTube is the most cited  domain in Google's AI overviews. And get this,  
according to our data, YouTube mentions have  a 0.737 correlation with chat GBT visibility.   That's the strongest correlation of any factor  we study. And there's a reason for that. GBT4  
was trained on over a million hours of YouTube  transcripts. So YouTube isn't just a platform   AI sites, it's a platform AI learns from. So,  in this lesson, I'm going to walk you through   a three-step process to get your YouTube  videos in front of AI. Let's get started.  
So, the first step is to find what's already  working on YouTube in your niche. And the key   idea here is that you want to target what I call  search hits rather than viral hits. A viral hit   can get you a spike of views and it can keep  spreading based on a YouTube user's interests.  
But once the YouTube algorithm has exhausted  the interested people, it dies. a search hit.   It gets you consistent traffic from both Google  and YouTube search month after month because   people are actively searching for that topic. And  search hits are exactly the kind of videos AI is  
likely to pull from. And that's because if Google  is already ranking a YouTube video for a keyword,   there's a good chance AI overviews will site it,  too. On top of that, titles for search videos   tend to be very clear about what the video is  about, whereas viral hits, not so much. So,  
here's how to find these topics. Go to site  explorer and enter www.youtube.com/watch. Then, open the organic keywords report. This shows  you every keyword that YouTube videos are ranking   for in Google. Now, filter for keyword rankings  in the top three and add your niche terms to  
the include filter. What you'll get is a list of  topics where YouTube videos are already ranking   at the top of Google and are related to your  space. Now, let's move on to the second step,   which is to actually create videos that rank.  Once you found your topics, you need to make  
sure your video is set up to rank in Google. And  there's a checklist I follow for this. First,   your title needs to contain the keyword people are  searching for. This isn't the place for clever or   clickbaity titles. If the keyword is how to use  Google Docs, that should be right in your title.  
Save the creativity for the thumbnail. The  title handles the keyword. The thumbnail   sells the click. Second, your description needs  to be a real summary of the video. Just write a   summary of what your video covers and add your  target keyword in the first couple of lines.  
Google reads this, AI reads this, and viewers  read it, so make a count. Third, add timestamps.   Timestamps turn into YouTube chapters, and those  chapters can show up in Google for specific  
queries. So, if your video covers five tips and  someone searches for tip number three, Google   can link directly to that chapter. It's basically  free extra visibility for like 2 minutes of work.  
Fourth, say the keyword in your video. And  this one's important. Google understands audio.   Liz Reed, who's VP of search at Google, has  said that Google can understand audio content   and video content. So, if your video is  about the best protein powder for repair,  
actually say those words in the video. Don't put  it in the title and hope for the best. And fifth,   match the format to what's already ranking.  If tutorials are dominating the search results   for your keyword, make a tutorial. If listicles  rank, make a listicle. Don't fight the format.  
Look at what's working and match it because  you're matching the intent of the searcher.   Now, none of this is complicated, but it's the  difference between a video that gets buried and   a video that shows up every time someone searches  for that topic. And that brings us to the third  
step, which is to layer in AI visibility. So, at  this point, you've got a video that's optimized   for Google search. But here's where you take it a  step further for SEO. Go to Brand Radar and enter   a popular brand or YouTube channel and go to the  topics report. Then set a filter where the domain  
mentioned is YouTube.com. Now, you can see exactly  which queries AI is pulling YouTube videos into.   So, go and create content around those topics.  And as for how to actually rank there, it comes  
back to what we already covered, pick the right  keyword, make a thorough, well optimized video,   and be comprehensive without wasting people's  time. And here's one more thing to keep in mind.  
Remember, every video you publish on YouTube  is potentially training data for AI models.   So even if a video doesn't get cited right away,  the content is being absorbed. The more helpful,  
specific, and well ststructured your videos  are, the more likely AI is to learn from them   and eventually recommend them. So, that covers  content mentions, and YouTube, but there's   a technical side to AEO that most people skip  entirely. Things like structured data, robots.txt,  
and making sure AI can actually access your site  in the first place. And that's what we'll cover in   the next lesson. I'll see you there. Hey, it's  Samo and welcome to the fourth lesson in this   module, which is on the technical side of AEO.  Now, I know technical can sound intimidating,  
but this lesson isn't about rewriting your site's  code. It's about making sure that AI can actually   access and understand your content. And while  access and understand might sound rudimentary   for some of you, the reality is a lot of sites are  accidentally blocking AI without even knowing it.  
According to our data, around 5.9% of 140  million websites are blocking GBTO, which is   Open AI's crawler. That's millions of sites that  are invisible to ChatGBT. So, in this lesson, I've  
got six technical checks and tips for you to make  sure AI can find you so that it can promote you.   Let's get started. So, the first thing  you need to check is your robots.txt file.   Robots.txt txt is a file on your site that  tells crawlers what they can and can't access.  
And the thing is, it's not just Google's crawler  you need to think about anymore. There are now   dozens of AI specific bots that crawl the web.  The main ones you should know about are GPTbot   and OAI searchbot from OpenAI, Claudebot from  Anthropic, and Google extended from Google. If  
any of these are blocked in your robots.txt, txt,  you're asking those AI platforms not to crawl your   content. And assuming they obey your rules, they  sure won't be recommending your pages then if they  
don't know what's on them. Now, you might not have  blocked these bots intentionally, but a lot of   sites inherit robots.txt rules from templates or  old configurations, and some platforms add blocks  
by default. For example, Cloudflare has a feature  called instruct AI bot traffic with robots.txt.   That's now enabled by default. When this is on,  Cloudflare automatically updates your robots.txt  
to signal that your content shouldn't be used for  AI training. So, if your site is on Cloudflare,   you could be blocking AI crawlers without even  realizing it. So, the first step is simple. Go to  
your doommain.com/root.txt and look for any lines  that mention GPTbot, Claudebot, Google extended,   or OAI searchbot. If you see a disallow rule next  to any of those, you're blocking that AI crawler.  
You can also use HF site audit to check this. Run  a crawl on your site and it'll flag any robots.txt   rules that might be blocking AI crawlers. Now,  while we're on the topic of files AI reads,   I want to make a quick note on something  you might have heard of called LLM.txt.  
This is a proposed standard kind of like  robots.txt, but specifically designed to   tell AI systems about your site. The idea is  that you create a file atyoudommain.com/lms.txt
that gives AI a summary of who you are, what your  site covers, and where to find your most important   content. It's useful in theory, but as of right  now, no major LLM provider officially supports it.  
OpenAI doesn't use it. Anthropic publishes one  on their own site, but hasn't confirmed their   crawlers actually read it, and Google hasn't  adopted it either. So, should you create one?  
Well, I don't think it'll hurt you, but I wouldn't  prioritize it over the other things we've talked   about in this lesson. Robots.txt is still the  file that actually matters most right now.   All right, the second thing to check is how your  site handles JavaScript. Some AI platforms can  
render JavaScript and some can't. Without getting  too technical, Gemini and Copilot can render JS   while ChatGpt's crawler does not. So if  your content relies on JavaScript to load,   which is common with single page apps and  some React or Angular frameworks, ChatGBT  
literally can't see your content. It visits  the page and gets an empty shell. The fix here   is serverside rendering, which means your server  sends the fully rendered HTML to the crawler  
instead of relying on JavaScript to build the page  in the browser. If you're already doing this for   SEO, you're covered. If not, it's worth looking  into, especially if AI visibility matters to you.  
A quick way to test this is to disable JavaScript  in your browser and visit your own site.   If the content disappears, you have a JavaScript  rendering issue that's affecting AI crawlers, too.   The third thing to consider is page speed. Now,  you might be thinking, page speed is an SEO thing,  
not an AEO thing, but it can actually matter more  for AI retrieval than for traditional search.   When AI systems retrieve information in  real time, they're fetching, parsing,  
and chunking your pages on the fly. And  if your page takes too long to load,   it can get dropped before it's even scored.  So, it won't be making it into an AAI response,   even if the content is great. The good news is  that if you've already optimized your core web  
vitals for SEO, you're most of the way there. Fast  loading pages with clean HTML benefit both Google   and AI systems. And that brings us to the  fourth tip. Create clean HTML structure.  
This one's straightforward. AI systems parse your  content by following your HTML structure. So,   if your headings are logical, your sections are  well organized, and your paragraphs are focused   on one idea each. AI has an easier time extracting  the right information. This ties directly back to  
the content principles we covered in lesson 3.1,  bluff, atomic content, and entity rich writing.   Those principles aren't just about writing style.  They're about making your content technically  
parsible for AI. So, when you're structuring  your pages, use proper heading hierarchy.   H1 for the title, H2s for the main sections, and  H3s for subsections. And make sure each section  
can stand on its own because AI might chunk your  content at any heading boundary. The fifth tip   is about schema markup. Schema markup, which is  also called structured data, is code you add to  
your pages to help search engines understand your  content. Things like article schema, FAQ page,   how-to, and local business. Now, does it help  with AEO? Honestly, the evidence is mixed. There's  
no confirmed data that adding schema directly  improves your chances of being cited by AI, but it   doesn't hurt. And if you're already using it for  SEO, there's no reason to remove it. I wouldn't  
spend a ton of time on schema specifically for  AEO, but if you're setting up a new page, adding   the right schema types is a good habit that makes  your content easier for any system to understand.  
All right, the sixth tip that I have for you  is to optimize for AI hallucinated URLs. AI   assistants sometimes make up URLs that don't exist  on your site. They'll recommend a page to a user,   the user clicks it, and they hit a 404 error. And  this happens a lot more often than you'd expect.  
According to our data, AI assistants send visitors  to 404 pages 2.87 times more often than Google   search does. And Chat GPT is the biggest offender  with about 1% of its clicked URLs leading to 404  
pages. Now, rather than letting that 404  be the end of a visitor's browsing journey,   you should either fix or optimize those pages to  get more out of them. You can do that by checking   your analytics for pages that are getting traffic  from AI referers but returning a 404 status.  
If you spot a hallucinated URL that's getting  consistent traffic, set up a redirect to the most   relevant real page on your site. That way, you're  capturing traffic that would otherwise be lost.   Now, while creating content and getting cited is  a big part of AEO, it's only part of the picture.  
You also need to know if it's actually working.  And that's exactly what we'll be covering in   module four, which is all about measuring and  tracking your AI visibility. I'll see you there.   Hey, it's Samo and welcome to module four in  HF's AI search course. Now, in module three,  
we covered how to create content, earn mentions,  and optimize YouTube videos to get mentioned and   cited in AI search. And we also talked about how  you can optimize for the technical side of AEO.  
So, now you've got a playbook for execution.  But here's the thing. None of that matters   if you can't measure whether it's working. And  that's what this module is all about. Now, I'll   be upfront with you. Measuring AI visibility is a  lot harder than measuring results from traditional  
SEO. With SEO, you've got Google Search Console  giving you impressions, clicks, and rankings.   With AI search, a lot of that data either doesn't  exist or is hidden from you. But that doesn't mean   you're flying blind, because there are three ways  to track your AI visibility. And in this lesson,  
I'm going to walk you through how to  set each one up. Let's get started.   So, the first thing you should track is AI  referral traffic to your site. This is when   someone clicks a link from Chat GPT, Perplexity,  Claude, or another AI platform and lands on your  
site. In analytics tools like Google or HF's web  analytics, it'll show up as a referral visit.   This is helpful in knowing how many people are  finding you because of AI. But here's where it   gets tricky. Not all AI platforms pass referral  data properly. Some of them strip it out entirely,  
which means the visit shows up as direct traffic  in G4. And if it does that, then there's no real   way to know that the traffic came from AI. For  example, ChatGpt source links in search results  
pass referral data properly. But incontent links  on paid accounts use a no referral attribute,   so those visits are actually invisible. Claude  tracks properly. Perplexity tracks on the web  
but not on its desktop app. Copilot tracks on the  web but not on Windows. And Grock doesn't pass   referral data at all. So the key takeaway here is  this. Use AI referral traffic to understand the  
general trend of which AI platforms are sending  you customers and visitors. But understand   that what you see in your analytics is likely an  undercount. Now with that caveat out of the way,  
group that isolates AI traffic. Go to admin,  then data display, then channel groups. Copy   your default channel group, and add a new channel  called AI traffic. Then set the source to match a  
reax that includes chatgpt.com, perplexity,  gemini.google.com, copilot.microsoft.com,   claw.ai, and deepseeek.com. Once that's set up,  go to reports, then acquisition, then traffic  
acquisition, and select your new channel group.  Now you can see the traffic AI is sending you,   the pages it's directing people to, and how  those visitors behave compared to other channels.   Now, if you want something simpler than G4, Hrefs  has a free tool called Web Analytics that does  
this automatically. It has a built-in AI search  channel, so you don't have to set up any custom   groups. It also separates unknown traffic  from direct traffic, which J4 doesn't do.  
That means you can get a clearer picture of  where your traffic is actually coming from.   All right, so once you've got AI traffic set up,  there are two things I'd pay attention to. First,   look at which pages are getting AI traffic. These  are the pages AI is already recommending to users.  
Make sure you keep these up to date, keep them  accurate, and have clear calls to action. If AI is   sending people to a page that hasn't been updated  in a year, it'll probably stop sooner than you'd   like. And second, look at your important pages  that aren't getting AI traffic. If you've got a  
key product page or a highv value blog post that's  getting zero AI referrals that should be getting   it, that's worth investigating. It could be a  content issue, a crawling issue, or it could just   mean AI isn't surfacing that topic yet. The second  thing to track is AI bot activity on your site.  
This one's a bit different. Instead of tracking  the humans who click through from AI, you're   tracking the AI bots themselves, the crawlers that  visit your site to read and index your content.   And here's something most people don't realize. AI  bots visit your pages far more often than humans  
do. So, the pages they're hitting the most are  likely your strongest citation candidates. Now,   there are two types of AI bots to know about. The  first type is training bots like GPTbot and Google   extended. These take your content and use it to  train AI models. The second type is search and  
citation bots like ChatGpt user and OI search  bot. These fetch your pages in real time when   a user asks a question. These are the ones that  can actually drive referral traffic to your site.  
To track bot activity, you can use server logs  if you have access to them. But the easier way is   through HF's bot analytics, which has a Cloudflare  integration that shows you exactly which AI bots   are visiting your site, how often, and which  pages they're focusing on. And this works with  
a free Cloudflare plan, too. Now, what you're  looking for here are patterns. If a citation bot   is hitting a specific page repeatedly, that page  is likely being used as a source in AI responses.  
And if there are important pages that bots aren't  visiting at all, that could mean they're hard to   discover, which ties back to the internal linking  and site structure we talked about in module 3.   All right, the third thing to track is  self-reported attribution. This one's the simplest  
to explain, but it might be the most important  for proving the value of AEO to your team   and to your clients. You see, a lot of the impact  of AI visibility doesn't show up in your analytics   at all. Someone asks ChatGBT for a recommendation.  They get your brand name, then they go to the  
browser bar and they type it directly in there.  That shows up as direct traffic or they Google   your brand name after hearing about you from AI.  That's going to show up as organic search traffic.   So the only way to capture this is to ask people  directly. Add a how did you hear about us question  
to your signup flow, your checkout process or  your post purchase survey. include options like   AI assistant, chat GPT, Perplexity, etc.,  and AI search like Google AI overviews.  
At Hrefs, around 3% of our conversions came from  AI over the last year based on self-reported data.   And our AI visitors convert at a much higher rate  than organic search visitors. But we would never  
have known that without asking. So, if you only  do one thing from this lesson, add that question   to an entry survey. It's the most direct way to  connect AI visibility to actual business results.  
Now, these three pillars work best when you  use them together. Referral traffic tells you   what AI is sending to your site. Bot analytics  tell you what content AI is paying attention to.  
And self-attribution tells you what's actually  driving revenue. No single source gives you   the full picture, but together they give you a  pretty clear view of how AI is interacting with   your brand and where you should focus your  efforts. And on top of these three pillars,  
you've still got brand radar tracking your AI  visibility across platforms, which I showed you   how to do back in module 2. So now that you've  got your analytics set up, the big question is,   is all of this actually worth it? What's the ROI  of AEO? and what should you actually be doing on a  
weekly and monthly basis to keep growing your AI  visibility? That's what we'll cover in the next   and final lesson. I'll see you there. Hey, it's  Ammo and welcome to the final lesson in HF's AI   search course. Now, throughout this course, we've  covered everything from how AI search works to how  
to find the right keywords and prompts to creating  and optimizing content, how you can earn mentions,   and even how to set up your analytics for AEO. But  there are three questions I haven't answered yet  
that are super critical to know. Is AEO actually  worth it? How do you know if you're making   progress? And what should you actually be doing  on a regular basis to keep growing? So that's what  
this lesson is about. Let's get into it. So let's  start with a big question. Is AEO actually worth   your time? Now I'll be honest. If you look at raw  traffic numbers, AI search is still relatively  
small. According to our data, AI referral traffic  accounts for about 0.25% of a site's total traffic   on average. And Google still sends about 210 times  more traffic than the top AI platforms combined.  
So, if you're looking at this purely from a  traffic standpoint, you might think it's not worth   it. But here's where the story gets interesting.  At Hrefs, our AI visitors convert at 23 times the   rate of organic search visitors. And we're not  alone. Verscell is seeing 10% conversion rates  
from AI traffic. Tally says AI is their largest  acquisition channel and helped boost their ARR by   a million dollars. Think about that for a second.  The traffic volume is small, but the quality of  
traffic is significantly higher. And the reason  for that is simple. When AI recommends you,   it's already explained to the user why you're a  good fit. So the traffic arrives prequalified.  
They're not browsing. They're ready to act. And on  top of conversion quality, AI traffic is growing   fast. It's grown about 9.7 times since last  year. Chat GPT alone has grown 85% since January,  
it now sends more traffic than Reddit or LinkedIn.  So, while the numbers are still relatively small   today, the trajectory, it's pretty clear.  AI traffic is only going to grow from here.  
But here's what I think a lot of people miss  about AEO. The real value isn't just the   clickable traffic. It's the brand awareness that  happens inside the AI conversation. Every time AI   recommends your product or mentions your brand,  that's an impression you never had before. And  
most of those impressions never result in a click  to your site. They result in someone googling your   brand name later or engaging with your brand  on social media when they see it in their feed   because they recognize it. So, the way I think  about it is this. AEO isn't an alternative to SEO.  
It's a new layer on top of it. And the brands  that build that layer now, while it's still early,   are going to have a massive advantage as AI  search continues to grow. Now, the main downside,   which we talked about in the last lesson,  is that it's not perfectly measurable.  
But just because you can't measure something  precisely, it doesn't mean it's not working.   It's kind of like brand marketing. You  can't track every billboard to a sale,   but you know it shapes how people think about you.  AEO is the same way. So, how do you know if you're  
making progress? Well, if you followed along in  module 2, you set up your brand radar baseline   and did your first brand gap analysis. Now, it's  time to go back and check how things have changed.   Pull up your brand radar dashboard and look  at four key things. First, AI share of voice.  
How have you moved relative to your competitors?  If your share has grown, your efforts are working.   If it stayed flat while a competitor has grown,  you need to dig into why. At HFS, we're tracking  
queries and prompts at scale. So, brand radar, it  gives you a really good look at the overall trend.   Second, look at cited domains. Are new domains  citing you that weren't before? This tells you  
if your mention earning efforts from module 3 are  paying off. Third, topic coverage. Have you closed   the gaps you identified? Are there new topics  where you're showing up where you weren't before?  
And fourth, mention sentiment. Is AI saying  accurate and positive things about your brand?   This is especially important because AI  doesn't just repeat what you put on your site.   It synthesizes information from everywhere. I'd  recommend doing this as a quick monthly check and  
do a deeper competitive audit quarterly. Now,  while we're on the topic of what AI says about   you, there's something important that you need to  be aware of. AI is vulnerable to misinformation.   And I don't mean it in a theoretical way. We  actually tested it. We created a completely fake  
luxury brand, planted three contradicting sources  about it across a blog, Reddit, and Medium, and   then asked eight different AI platforms about the  brand. The results were pretty alarming. After the  
fake sources were planted, Gemini and Perplexity  repeated the misinformation in 37 to 39% of their   answers. They cited fake founders, fake cities,  fake pricing stories, all presented as verified  
facts. Now, the good news is that Chat GPT was  much more robust. It stayed under 7% and cited   the brand's official FAQ in 84% of its answers.  But the point is that not all AI platforms are  
equally resistant to bad information. So, what  does this mean for you? It means you need to fill   every information gap about your brand with  specific and official content. Create an FAQ  
that directly addresses common questions.  Use specific numbers and dates and facts,   not just vague claims. Because when AI has to  choose between vague truth and specific fiction,  
it tends to choose the specific fiction. Also,  monitor what AI is saying about you regularly.   If you spot inaccurate information, the fastest  fix is to publish content on your own site that   directly contradicts it and then work on getting  the third-party source corrected. All right,  
so let's bring this all together with your AEO  action plan. If you're wondering what to do first,   here's what I'd focus on this week. Check your  robots.txt for AI bot access. This takes about  
5 minutes max, and it's the most common technical  blocker. Make sure AI can crawl your site. Next,   set up your AI analytics. Create the AI traffic  channel in G4 or set up HF's web analytics. Add a  
how did you hear about us question to your signup  or checkout flow. Start measuring from day one so   you can see how you progress over time. Next,  update your most important pages for freshness.   Remember, AI content is 25.7% fresher than what  ranks in traditional search. Pick your top five  
to 10 pages and make meaningful updates, new  stats, updated examples, current information.   Next up, run your brand gap analysis. Set  up brand radar if you haven't already, and  
go through the process we covered in module two.  Know where you stand before you start optimizing.   And finally, identify your top 10 mention earning  targets. Use the cited domains and pages reports  
in content explorer to find the pages where  getting a mention would have the biggest impact   on your AI visibility. And then on an ongoing  basis, do a monthly brand radar check to track   your progress in a quarterly competitive audit  to catch bigger shifts and just rinse and repeat.  
Everything I just mentioned is covered step by  step in this course. So if you need a refresher   on any of it, just go back to the relevant lesson  and then execute. AI SEO or AEO is still early.  
The tools are evolving, the data is getting  better, and the opportunity is only going to grow.   So the biggest advantage you can have right now  is simply starting before everyone else does.  
Thanks for watching our AEO course  and I'll see you in the next