(0:05 - 0:14)
Welcome to the Conquer AI Search podcast. We're your AI monitor team source for figuring out this constantly changing world of digital visibility. I'm Catherine.

(0:14 - 0:34)
And I'm Avi. Great to be back. So we are well into our 11 part series now, looking at the most effective generative engine optimization techniques, you know, Geo, we've already covered implementing LLMs.txt files, the maybe surprising impact of Reddit and using authoritative citations strategically.

(0:37 - 0:53)
And today we're taking all that and sort of zooming out a bit. We're looking at how to evolve your entire traditional SEO playbook, your best practices for this new world of generative engine optimization or answer engine optimization, AEO, as some folks call it. Yeah, exactly.

(0:53 - 1:03)
And it's not about throwing out everything you learned about SEO, not at all. It's more about adapting it, building on it for AI. Okay, but here's something that jumped out immediately, something maybe a bit counterintuitive.

(1:04 - 1:27)
We found that the standard advice on keyword density, you know, the 1% to 2.5% range, every SEO consultant mentions, well, that might actually have zero impact now or get this, it could even hurt your Geo performance. Can you tell us more about what the AI monitor team found there? Yeah, this is really quite crucial. We ran hundreds of tests, literally hundreds of trials and errors here at AI monitor.

(1:27 - 1:40)
And what we consistently found is that right now, as of today, June 20th, 2025, a keyword density of 0.8% seems to hit the sweet spot. 0.8%? Yeah. That specific? That specific.

(1:40 - 1:52)
It yields the best combined results for both traditional SEO and your Geo efforts. It's data backed from all that testing. It really shifts how you should think about optimizing your content.

(1:52 - 1:55)
Wow. Okay. That 0.8% figure is gold dust for our listeners.

(1:56 - 2:04)
So keeping that very specific number in mind, let's talk bigger picture. People often say SEO is dead, but you're saying that's not quite it. Not really, no.

(2:04 - 2:13)
It's more accurate to say SEO isn't being replaced, it's expanding. It's evolving into this bigger thing, Geo. Your traditional SEO skills, they're still absolutely vital.

(2:13 - 2:25)
They're the foundation, the baseline. So what's the key difference then, SEO versus Geo? Well, think about it like this. Traditional SEO is mostly about getting ranked in that list of blue links on Google, right? You want to be high up on that page.

(2:26 - 2:39)
Geo though, is optimizing for the AI to generate a comprehensive direct answer. Like in Google's AI overviews, or when you ask ChatGPT or Gemini a question. The goal is to be part of that synthesized answer, not just a link below it.

(2:39 - 2:52)
And that shift has real consequences, doesn't it? We're seeing things like Google AI overviews, grab clicks that used to go to websites. Some reports show users getting their answers directly from the AI overview and never clicking through. Absolutely.

(2:53 - 2:58)
Click-through rates from traditional results can definitely take a hit. Yeah. But it's not always a simple replacement.

(2:59 - 3:13)
We're seeing this interesting reverse causality, you could call it. Reverse causality, what do you mean? Well, someone might first encounter your brand or product in an AI-generated answer. Maybe it's mentioned in a list or as a recommendation.

(3:13 - 3:25)
Then they might go to Google search to specifically look you up, you know, to check reviews or pricing. Ah, so AI introduces and then Google validates a new kind of customer journey. Exactly.

(3:25 - 3:31)
It's a different funnel, starting with AI discovery. Okay, that makes a lot of sense. So if people are following along, the next big question is probably keywords.

(3:32 - 3:48)
How does our keyword strategy need to change right now for AI? Right. The old way of just chasing high volume head terms, those short, broad keywords, it's just not cutting it anymore. AI tools, they seem to really like long tail keywords, conversational phrases, questions.

(3:48 - 3:56)
Like how people actually talk. Pretty much. Think about Google's own data, they say 15% of all searches are queries they've never seen before.

(3:56 - 4:09)
That tells you there's a massive appetite for niche-specific content. So what are the concrete steps? How do we adapt? Okay, first, go low. Don't automatically filter out keywords just because their monthly search volume looks low.

(4:09 - 4:15)
You need to map out more terms, including those lower volume, more specific ones. More niche terms. Exactly.

(4:15 - 4:21)
Second, be conversational. Tap into the actual questions people are asking. Look at Google's People Also Ask boxes.

(4:22 - 4:27)
Use tools like Answer the Public. Check out forums, like Reddit, which we've talked about before. Right.

(4:27 - 4:33)
Tying back to that previous deep dive. And it's not just the exact phrase they type, is it? Semantic search comes into play. Absolutely.

(4:34 - 4:38)
You need to use variations. Optimize for semantic search. The underlying meaning.

(4:39 - 4:42)
The intent behind the words. Use related phrases. Synonyms.

(4:43 - 4:50)
A really effective tactic here is creating helpful FAQ sections right within your content. Like an FAQ page. Or within blog posts.

(4:50 - 5:00)
Within the content itself often works best. It lets you target multiple long-tail questions naturally. We saw Kiehl's, for example, get something like over 550 AI overview rankings.

(5:01 - 5:07)
Partly because they structure their content well with FAQs. That's impressive. And again, remember that keyword density point we made.

(5:07 - 5:16)
That .8%. That's a direct, actionable change to your keyword strategy, moving away from that old 1-2.5% advice. Got it. Precision matters.

(5:16 - 5:18)
Okay. Moving on. Content clarity and structure.

(5:19 - 5:26)
This sounds fundamental for AI readability. How do we tackle this? Yeah, it's critical. AI models are built to pull out concise, structured information.

(5:26 - 5:33)
So your content needs to be, well, easy for them to digest. That means clear, descriptive headings. H1, H2, H3 is outlining the structure.

(5:33 - 5:40)
Striking things down. Exactly. Break up big walls of text, use short paragraphs, bullet points, numbered lists.

(5:40 - 5:46)
Make it snackable. It's also really good practice to put key takeaways right near the top. Like a summary up front.

(5:46 - 5:52)
Sort of, yeah. Helps both users and AI grasp the main points quickly. And in terms of writing style.

(5:52 - 5:56)
Yes. Lean into natural language. Right? Like you're answering a question someone actually asked.

(5:56 - 6:04)
Use common question formats. Avoid super technical jargon unless your audience absolutely expects it. And featured snippets.

(6:04 - 6:09)
Still relevant. Very much so. AI often pulls from featured snippet content.

(6:09 - 6:15)
So optimizing for those with concise definitions and bullet points helps. Oh, and don't forget tables. Tables.

(6:16 - 6:22)
Really? Yeah. AI can parse structured data in tables really well. Especially if they have clear, descriptive headers.

(6:22 - 6:30)
It makes complex information much more accessible for the AI. Interesting. So clear structure, snackable content, natural language, maybe some tables.

(6:30 - 6:40)
But none of that matters if the AI can't even find or access the page, right? Let's talk technical SEO. Is it still the bedrock? Absolutely. Maybe even more so now.

(6:40 - 6:49)
Technical SEO determines if your content is even visible and interpretable by AI crawlers. Without a solid technical foundation, you're invisible. Okay.

(6:49 - 6:54)
What are the non-negotiables? Well, HTTPS. Secure connection. It's a ranking factor.

(6:54 - 6:59)
Plus, essential for trust and privacy. That's table stakes. Then mobile optimization.

(6:59 - 7:04)
Huge. Google uses mobile-first indexing. Most searches are mobile.

(7:04 - 7:10)
And AI often previews mobile layouts, you said. Right. So a bad mobile experience is a bad signal to AI.

(7:10 - 7:18)
Then you've got to be fixing crawl errors. These block search engines and AI bots check for them regularly, like monthly audits. Keeps the pathways clear.

(7:18 - 7:20)
Exactly. And site speed. Super important.

(7:21 - 7:28)
AI prioritizes fast sites. Remember something like 53% of mobile users bounce if a page takes more than three seconds to load. Speed matters.

(7:28 - 7:42)
What about site structure itself? That falls under crawlability and indexation. A logical site architecture, good internal linking, it helps search engines and large language models, the LLMs, navigate your site efficiently. Okay, now, schema markup.

(7:42 - 7:47)
This keeps coming up as increasingly important for AI. Tell us about that. Schema is massive.

(7:48 - 8:04)
Yeah. It's like adding little labels to your content that explicitly tell search engines and AI what the content is about. Is it for product? An FAQ? A how-to guide? An article? A review? Video? Local business info? So, it removes ambiguity for the AI.

(8:04 - 8:12)
Precisely. It helps AI surface accurate info and really boost your chances for rich snippets and appearing correctly in AI answers. And one more technical point, JavaScript.

(8:13 - 8:16)
Oh, yes. The bane of some crawlers. Still can be.

(8:16 - 8:30)
Many AI crawlers, including GPTBot from OpenAI and CloudBot from Anthropic, don't execute JavaScript effectively or at all. So, critical content needs to be in the raw HTML, server-rendered if possible, not hidden behind client-side JavaScript execution. Good tip.

(8:30 - 8:45)
Okay, technical foundations sorted. How do we then build that trust and authority? We talk about Google's E-E-A-T signals, experience, expertise, authoritativeness, trustworthiness. How does that apply to AI? AI models are explicitly trained to prioritize credible, authoritative information.

(8:46 - 8:52)
So demonstrating E-E-T is key. Practically, that means showcasing author expertise, detailed bios, credentials. Show your work, basically.

(8:52 - 9:02)
Yeah. Site-reputable sources, link out to original studies, keep those references up to date. High-quality backlinks from trusted sites still matter a lot for authority.

(9:03 - 9:26)
And what about the human edge? AI can generate text, but what can humans provide that AI can't replicate easily? That's where the experience part of E-E-T really shines. AI struggles with genuine lived experience, original thinking, deep product expertise based on actual use. So include unique frameworks, specific use case details only an expert would know, bring in distinct voices and points of view.

(9:26 - 9:39)
Make it unique? Yes, and present balance views. AI aims for neutrality, so showing pros and cons, strengths and weaknesses that resonates. Also, proprietary data, your own surveys, case studies that makes your content inherently unique and valuable to AI.

(9:40 - 9:49)
Right. Now, something often seen as separate, but you're saying it impacts GEO, user experience, UX. How does a good UX help with AI visibility? Because a bad UX sends bad signals.

(9:49 - 10:00)
A slow, confusing site frustrates users, increases bounce rates, lowers dwell time. And Google and AI interpret these as signs that the content isn't high quality or helpful. So dwell time matters.

(10:00 - 10:13)
How long people stay on the page? It seems to be a significant engagement signal, yes. Keep users engaged with visuals, clear calls to action, good internal linking, and accessibility. Don't forget website accessibility, AI considers this too.

(10:14 - 10:30)
What about things like images and videos? High quality multimedia is great for UX, but make sure you include descriptive alt text for images and transcripts for videos. This helps users and helps AI understand the content of that media. Okay, and keeping things current, content freshness.

(10:30 - 10:38)
Oh, absolutely essential. AI-driven search relies heavily on timeliness for accuracy, outdated information gets penalized. Keep content fresh.

(10:39 - 10:45)
Use a content calendar, plan regular updates for stats, strategies, especially for fast moving topics. Makes sense. Now, this is fascinating.

(10:46 - 10:58)
Where does AI actually get its information? It's not just our websites, is it? Not at all. AI pulls from a really diverse range of sources. Yes, your site, but also independent blogs, niche affiliate sites, industry publications.

(10:58 - 11:07)
We've seen before that while top tier sources are good, AI often cites mid-tier sources too. So you don't have to be the New York Times to get cited. Definitely not.

(11:07 - 11:20)
And the preferences vary. ChatGPT, interestingly, seems to cite lower ranking Google results positioned 21, plus almost 90% of the time. Often content focused on very specific use cases.

(11:20 - 11:27)
And for Google's AI overviews. Quora comes up top, followed by Reddit. Again, highlighting those community and Q&A platforms we discussed earlier.

(11:28 - 11:41)
But you also mentioned our own research showing business sites get cited too. Yeah, about 50% of ChatGPT links point to business service sites. So there's definitely potential for direct citation if your content is well-optimized, LM-friendly.

(11:41 - 11:48)
And there are even direct submission programs emerging. Right. Things like Perplexity's Merchant Program, OpenAI's Search Product Discovery Initiative.

(11:49 - 11:55)
Brands can submit product catalogs directly for AI recommendations. It's a very direct route. This is all changing so fast.

(11:55 - 12:07)
Which brings us to measurement. How do we track success in this new GEO world? Traditional analytics don't quite cut it, do they? Not directly for AI-driven traffic, no. So we need new approaches.

(12:07 - 12:17)
First, monitor your referral traffic. Look for hits coming from domains like chat.openai.com, perplexity.ai, claude.ai, and so on. Check the referrers.

(12:17 - 12:30)
What else? Watch your branded search volume. If you see more people searching Google specifically for your brand name, maybe adding reviews or pricing, that could be an indirect sign they discovered you via AI first. That inverse customer journey again.

(12:30 - 12:40)
Makes sense. How do we check if our content is actually showing up? You have to actively test. Regularly put relevant prompts into chat.gpt, gemini, claude, perplexity.

(12:41 - 12:49)
See if your brand, your content, appears in AI overviews or GenAI results. Are you in lists? Cited as a source. Take screenshots.

(12:49 - 12:54)
Track it. So manual checking is still part of it? For now, yes. But tools are catching up.

(12:54 - 13:14)
SEO platforms like RFs and SEMrush are starting to integrate tracking for AI overview appearances. We're seeing new metrics emerge. Like what? Things like an AI visibility score tracking how often you appear, how prominently, sentiment, also citation frequency, share of voice specifically within AI results, and sentiment analysis of how your brand is mentioned in AI outputs.

(13:14 - 13:25)
These give you a much clearer picture of your GEO performance. Okay, so wrapping this all up. The big message for everyone listening seems to be SEO isn't dead, but it's definitely transforming.

(13:28 - 13:37)
Well, that about wraps up today's Conquer AI search session. Quite a journey into how SEO is evolving. And remember the big takeaway on keywords.

(13:37 - 13:54)
For top performance across both old and new search, our AI Monitor team's latest findings as of June 20, 2025, suggest aiming for that 0.8% density. Yeah, that's the number for now. And hey, if you found this useful, please do follow or subscribe wherever you listen.

(13:54 - 14:04)
Spotify, Apple Podcasts, Audible, Amazon Music, maybe right here on YouTube. And if you got value out of this, maybe consider leaving us a rating or a review. It genuinely helps us reach more people like you.

(14:04 - 14:10)
We'd really appreciate it. Absolutely. We'll see you next Saturday to talk about the fifth generative engine optimization technique.

(14:10 - 14:18)
And just a thought to leave you with. In this AI-first world, your website isn't just content for people anymore. It's data for machines too.

(14:18 - 14:22)
The better you structure that data, the better you can shape how AI understands your brand.