I’ve been doing digital marketing and SEO for 30 years and I don’t remember a time where there was so much confusion over a topic related to SEO. Not even the Penguin and Panda updates that basically destroyed article marketing about 15 years ago. On the topic of AI Search Optimization, it has sparked thousands of conversations, a whole new group of snake oil salespeople, a whole group claiming to be AI Search Optimization Experts and not to mention the new batch of “SEO is Dead” chicken little posts. So, I wanted to create this AI Search Optimization FAQ to clear a few things up and add a few tips that might help you by framing it in a different way. I hope this helps.
Basic Questions
Is SEO Dead?
Just wanted to get this one out of the way first. No, SEO is not dead in 2026, no more than it was 5 years ago, 10 years ago, 15 years, ago or 20 years ago when people were saying it. SEO is still the foundation of discoverability just as it always was. Platforms change, methods change, strategy changes, usually based on new data, but it's still SEO and AI cannot fundamentally change that.
What Is SEO in 2026?
This question defines the rest of the questions on this list because if you ask a bunch of SEO's what good SEO is, you will get a lot of different answers, so I'll define it my way to help you understand the rest of the answers in this post. TIER 2 SEO: "Google doesn't care about your content. It's all about keywords and backlinks and Programmatic SEO. TIER 1 SEO: High-quality content + link attraction over link building. Building authority over time instead of chasing short-term gains.
What Is AI Search Optimization?
Some call it AI SEO, GEO, AEO, LLM SEO or some other made up acronym. It's about brand visibility, trust, authority and quality. For Tier 1 SEOs it's not a huge difference from what you have been doing. If you fall into the Tier 2 category don't you have another 1000 AI generated pages to build anyway?
AI SEO Claims And Myths
Do I need to add an LLMs.txt file and MD files?
Fact as of 3/16/26: There is no official announcements by any of the major AI companies like OpenAI, Anthropic, or Google that says they have adopted any specific file as a default. It is not equal to robots.txt which has been widely accepted and used for many years. That being said, it won't hurt you to add the files if you are predicting they will become the standard.
Myths: 1. If you don't have these files, you won't get brand mentions in AI responses. 2. If you don't have these files AI will ignore you and not include you in responses or training data. 3. If I have an llms.txt file, the quality of my actual website doesn't matter as much.
The Reality: An llms.txt file is just a map to your content. If an AI model uses it at all, it still evaluates the original source, your content, for accuracy, depth, relevancy and value (Think EEAT). If your Markdown files point to thin, AI-generated, or unhelpful content, the AI will simply learn more quickly that your site isn't a high-quality source.
Tip: Add the files. Don't try to game them thinking you are smarter than the AI or Google. Having them doesn't hurt, but don't expect them to be magic.
Will Schema Markup Will Get Me More Brand Mentions in AI Search?
Fact: Schema.org was founded in 2011 as a joint effort by Google, Microsoft (Bing), Yahoo, and Yandex. They have spent the last 15 years building the large language models (LLMs) and search algorithms we use today. While the vocabulary was originally designed to help search engines understand "entities" (like people, places, and products), it has now become the training manual for modern AI.
Myths: 1. Schema is only good for getting rich snippets in Google. 2. We didn't need schema before AI search. 3. The only schema types you need are the ones Google will consider for rich snippets. 4. Without schema markup your website is basically invisible to AI.
The Reality: Schema has been used for SEO purposes since around 2011. Schema markup was never "just" for Google rich snippets. Google just recognized certain types of schema for rich snippets. There is a difference. There are more than 800 types of schema at schema.org. AI can read all of them. If it is relevant to what is on your page, it may not qualify for a rich snippet but it will help clarify what your page is about, even for Google.
Tip: Don't limit your schema vocabulary to just what Google says qualify for rich snippets. Here's a good use for AI. Show AI your web page and use this prompt or something similar; Scan schema.org for me and give me a list of the types of schema that are legitimately good to use for this page.
Can I Track How Many Brand Mentions I Get?
Fact: Currently you cannot track brand mentions reliably across AI search tools and LLMs.
Myths: 1. Our tool shows you every time any AI mentions your brand anywhere. 2. You’ll get an accurate, complete ‘share of voice in AI’ score from our dashboard. 3. Once you plug this in, you’ll finally know exactly how many times AI recommends your competitors instead of you.
Reality: Right now, any brand‑mention tool is only sampling a tiny slice of prompts, models, and regions, and the answers can change every time someone runs the same query. At best, these tools give you directional signals on a small set of tracked prompts, not a full census of AI mentions, so treat them like spot checks, not a magical truth meter.
Tip: Quit thinking about how many brand mentions you get. Make sure that you monitor what AI is saying about your brand when it does get mentioned. This is where you actually do have some control and it matters more.
Can AI Write My Content If It Knows My Style?
Fact: AI cannot replace content written by humans in any meaningful way. When a human writes, they relate personal experiences, anecdotes, stories related to what they are writing about. AI has no experience to reference. It doesn't know you in the sense some people believe it does. It has no emotion so it can't really trigger emotion and good content does exactly that.
Myths: 1. There's an AI that humanizes your AI content! 2. If you don't think AI can write better content, you just don't know how to prompt. 3. My AI knows my writing style because it has samples, so it can write just as good as if I did it myself. 4. AI is going to replace human writers because it writes better and faster and it's cheaper.
Reality: You just don't know how to prompt has entered the chat. It's not 2022. Anyone using AI and also writes content knows just as much about prompting as anyone calling themselves a "prompt engineer". If you think AI writes as good or better than you do, maybe writing was never your forte to begin with.
Content generators aren't new. In the early 2000s article spinning software was all the rage until Google's Penguin and Panda updates. Relying on AI content means a slightly possible initial gain, but in the long run it will hurt the very things you should be focused on. Trust, Authority, Engagement, and Conversions should be your goal, not some shortcut.
And my favorite myth: Asking an AI to humanize your AI generated content? That's like auto‑tuning an auto‑tuned song. Using 2 AIs doesn't make it more human. If anything, whoever is selling that is secretly laughing at the fact people actually bought it.
Tip: Your workflow should be more like this. Write the content yourself. Get your thoughts down and structure how you would normally do it. Then ask AI to look it over with a prompt something like; Are there any points you think I forgot to cover? Fact check any statements I included as fact by performing a deep search. Check my grammar and spelling as well as any content that isn't explained clearly enough. Do not use the content that it offers as a rewrite. Just look at the suggestions that you feel are good and implement them yourself.
Okay Now The Real AI Search Optimization FAQ
What is an "AI Search Engine"?
They aren't search engines. Most tools branded as "AI Search" are actually Answer Engines or LLMs utilizing Retrieval-Augmented Generation (RAG). Instead of returning a list of links, they generate an answer by first relying on their vast internal training data (parametric memory). If the query requires real-time facts, or the user asks for a search, they will crawl the web, reading websites, forums, and databases to provide a direct response. "Thinking Models" sometimes reveal the process as it deliberates.
If I rank well on Google, will AI automatically recommend me?
No. Studies consistently show that only about 15% to 35% of AI-generated answers correlate with the traditional top 10 organic search results. AI models evaluate information using a different scoring system (if you can call it that) than traditional search. They look for direct answers, trustworthy information, and entity authority, meaning traditional ranking metrics do not guarantee an AI citation.
That being said, the signals AI does look for are rooted in doing genuinely good SEO. When it searches the web, PR, brand visibility, social signals, reviews, and authoritative citations all heavily influence the answers AI provides. When AI relies on its training data rather than a live search, it leans on clear page structure, strong E-E-A-T, and precise schema markup to understand and surface your content.
Where is AI actually getting its information if not just search engines?
LLMs are trained by directly scraping the web, including websites, social platforms, Reddit, forums, and data repositories. They use this data to understand entities and relationships. When an AI generates an answer, it is relying on this training data first, before it ever triggers a live web search.
Do backlinks still matter for AI visibility?
Yes, but the math has changed. AI models don't care about traditional link juice or arbitrary domain authority scores (DA). They view a link as a contextual bridge between two entities. And it doesn't actually have to be a link. Referencing your brand is enough if you have built up enough reputation and authority that helps AI easily identify the reference. A mention from a highly relevant, trusted industry source acts as validation for your entity. Spammy or irrelevant links are simply ignored by the model's training data. One link from a highly authoritative, trusted source relative to your topic means more than 100 DA50+ random blogs or sites with thin content.
Should I be optimizing for specific keywords for AI?
Keyword stuffing is entirely useless for AI. While keywords establish baseline relevance, AI relies on semantic understanding, how words and concepts relate to each other. Instead of chasing search volume, focus on answering the complex, long-tail questions your audience is actually asking with deep, structured, and highly specific content.
Some of the most overused key phrases lately are "chunking", "content clusters", "waterfalls", "semantic SEO", "FAQs" and "Schema markup". However, they are all ways to clarify what your content is really about making it easier for humans, Google and AI to understand and trust it.
Should I block AI crawlers (like GPTBot) from my website?
If you want to be part of the conversation, no. Some publishers block AI crawlers out of fear their content will be "stolen." However, if an AI cannot crawl your site, it is more difficult to include your brand in its training data, it cannot cite your expertise through the primary source of that information, your website or blog, and it cannot recommend your services if it doesn't know what they are. Blocking them guarantees your website content will be invisible to those models.
How do I get an AI to trust my content?
Consistency is everything. AI models look for corroboration. If your website claims you are an expert, but your local directories, social profiles, and industry citations have conflicting information, the AI doesn't know which is correct and will avoid citing at all. A clean, consistent, and widely corroborated digital footprint is the only way to establish authority.
Here is a simplified guide. This is nowhere near a complete strategy. 1. Pick 3 things you want to be known for. 2. Make sure all of your citations and profiles reference those same 3 things. 3. Make sure you write articles, posts, pages and even comments that include one of those things often without being spammy about it.
What is the actual ROI of optimizing for AI?
The ROI is long-term brand authority and capturing high-intent users who want direct answers, not a list of 10 blue links. Building a website that fits into a comprehensive marketing plan and establishes your entity will drive actual conversions.
At the end of the day, ranking is vanity. Sales are sanity.
I will add this since we are talking about ROI. Studies of tens of thousands of sites put the average AI referral traffic, as of August, 2025, ChatGPT and Perplexity combined drive approximately 0.15% of all web traffic. (Source SERanking.com)
To put that in perspective: Google organic search still sends roughly 300 to 350 times more raw traffic to websites than all the AI chatbots combined.
My Advice: Spend the majority of your time on what actually moves the needle right now. If you were doing a Facebook ad campaign that was producing very little traffic to your website and you had a TikTok campaign that was driving 350 x that Facebook traffic, where would you focus your money and time?
Side Note: Since you should already be focused on doing SEO the right way, you can optimize for AI search at the same time. Maybe be a little more conversational, answer a few more questions, be more helpful and be more consistent in your messaging across the web.
What Really Matters In The Age Of AI Search? Entity SEO / Entity Optimization
What is an Entity?
An entity is a "thing" rather than a "string of keywords". It is a distinct, well-defined concept that can be uniquely identified. An entity could be a person (like a company founder), a physical place, a brand, a product, or even an abstract concept like a tagline or slogan, think "You're in good hands with Allstate" or "Where's the beef?". To AI and modern search engines, an entity isn't just a word on a page; it is a specific node in a massive web of data (a Knowledge Graph) that has established relationships, attributes, and a verifiable identity.
What is Entity SEO or Entity Optimization?
Entity Optimization is about building a clear, verifiable digital footprint over time. Instead of trying to trick an algorithm into ranking a specific page for a specific phrase, you are organizing your digital presence so that Google and AI models understand the 5 Ws, the who, what, where, when, why and how of your entity.
Why is Entity SEO important?
Modern search engines and AI models don't read text the way humans do; they connect the dots between entities in a massive knowledge graph. If your entity is weak, confusing, or has conflicting information across the web, AI models won't trust you, which means they won't cite or recommend you. Your entity can also be weakened if there are similar names to deal with, because AI hates ambiguity. Think of Entity SEO as an added trust layer: AI hates to be wrong, so if the trust signals for your entity are higher than a competitor's, it's going to choose the source it can apply the most trust to. Ultimately, it is the foundation for future-proofing your business. When your entity is strong, you stop being reliant on a single platform and start showing up wherever your customers are searching.
What are the most important things to do to strengthen an entity?
- A Consistent Digital Footprint: Make sure your brand's core information, messaging, and data are perfectly aligned across your website, social profiles, local directories, and industry citations. This is more than just NAP. Same products, services, value proposition, slogans, etc.
- Advanced Schema Markup: Using structured data to spoon-feed search engines and AI the exact details about your organization, its founders, and its core concepts without making them guess.
- Corroboration: Earning mentions, reviews, and links from other already-trusted entities (like major industry publications or authoritative local businesses) to validate your brand.
- Top-Tier Content: Publishing deep, structured content that clearly ties your brand to your core topics of expertise.
How does Chris McElroy SEO Agency help companies with Entity Optimization?
We don't just chase vanity metrics or isolated keywords; we build your Digital Identity Architecture. We start with a comprehensive entity analysis to find the gaps in how search engines and AI currently view your brand. This is like a snapshot, a frozen moment in time to use as a baseline.
From there, we deploy a unified strategy utilizing advanced schema, Search Everywhere Optimization, and Integrated Content Marketing (Not the 2010 version, more like becoming a digital publisher) to solidify your authority across several authoritative, platforms that we know are historically sourced for AI answers.
We ensure your website actually fits into your broader marketing plan to build a trusted entity that drives highly qualified traffic. Because at the end of the day: Ranking is Vanity. Sales are Sanity.
Visit ChrisMcElroySEO.com to learn how we incorporate everything we do toward helping you build a strong entity for your name, your business or brand and your products. You can also set up a 15 minute chat to ask me questions about it here.
