Your #1 Google Rank Means Nothing to ChatGPT

Right now, some brands are showing up constantly in chat GPT answers. Others are completely invisible. And the brands that are invisible, a lot of them rank number one on Google. That number used to be 76% and it's now 38% and it's still falling. So, if Google rankings don't protect you anymore, what does? That's exactly what I went and found out. I'm Neil Patel and my agency NP Digital helps some of the largest companies in the world with their marketing. And we even won Ad Age's performance marketing of the year for the results that we're driving to our clients. And over the last several months, we've been tracking which brands get pulled into chat GPT answers and which ones are getting skipped entirely. I cross-referenced that with an independent study from Right Sonic. 119 conversations, 1,161 citations, I went through all of it. Five patterns keep coming up. But before I get into them, there's one concept that you need to understand first because it explains everything that follows. Retrieval. That's what AI search is built to do. It's one job is to retrieve the most trustworthy, relevant, and extractable source for any given question. Not the highest ranked page on Google, not the most popular website, the most retrievable one. And that's a completely different game. The brands that are winning in AI search have figured out how to make themselves easy to retrieve. The brands that are losing, they've been optimizing for Google rank, which is a different algorithm with different rules pulling from different places. Everything in this video comes back to that one word. Now, let's get into it. Google rank barely predicts whether AI sites you. Here's what most marketers assume. They assume that if they rank well on Google, they're protected. That their SEO investment carries over into AI search automatically. The data says otherwise. Google's top 10 used to account for 76% of ChatGPT citations. The number is now 38%. And 75% of all AI citations now come from sources that don't appear in Google's top results at all. We studied 500 commercial keywords at NP Digital, ran 4,308 related prompts through LLMs. We found that ranking number one on Google gave you a 31.4% AI mention rate. By ranking four, that dropped to 2.6%. But here's the part that should really get your attention. 90% of the pages ChatGPT sites rank 21 or lower on Google. Meaning, AI is pulling sources that Google itself barely acknowledges. I've seen this firsthand with clients. One brand ranking number one for their main keyword, solid SEO, strong domain authority, zero AI citations. Because every piece of content they ever produced lived on their own website. They never built [music] presence anywhere else. The brands getting cited consistently, different story entirely. So we started working with this mid-size SaaS company. They never really did any SEO. They wanted us to specifically focus on ChatGPT for them. And they don't really rank on Google for anything. But they're now showing up consistently in ChatGPT answers. Why? Because they're mentioned in G2 reviews, Reddit threads, YouTube explainer videos, and niche industry publications. AI pulls from all of those. So here's what you need to actually do. One, Google your brand name plus your core topic. Look at what non-brand sources appear on page one. Those are places AI's already pulling from. Are you mentioned in them? Two, identify the top five sources ChatGPT sites in your niche. Run a few prompts and look at what is getting linked. Reddit threads, YouTube channels, review sites, niche blogs, find the pattern. Three, get your brand mentioned in those places. PR outreach, podcast appearances, product reviews, go where AI is already looking. Now, while you're thinking about where AI finds you on the web, there's something happening with your own website that almost no one talks about. And the data around it completely flips the conventional wisdom. That's because your website now matters more, but only if it's structured right. Everyone's been saying the same thing for the last year. AI is answering questions directly. Your website matters less now. The data just showed the exact opposite happened in a single model update. With GPT 5.3, the default model most people are using, only 8% of citations went to brand websites. With GPT 5.4, the premium thinking tier, that number jumped to 56%. That's a 7x increase in just one update. >> [music] >> Here's why. GPT 5.4 runs 8.5 subqueries per prompt. [music] GPT 5.3 runs one. And GPT 5.4 goes to brand websites first when it knows who you are. 37% of GPT 5.4 query types are site colon operators, meaning it goes directly to your domain. It already knows your brand. It's just checking whether your site is worth pulling from. >> [music] >> And for a lot of brands, the answer is no. Not because the content is bad, because it's unextractable. Walls of text, no clear headings, no FAQ sections, no structure that lets AI cleanly pull a two-sentence answer. So, it skips the page entirely, even if the brand is well known. High-level is example of the other side of this. Structured content, clear H2s, FAQ blocks, data-backed sections, GPT 5.4 goes to their domain first because it trusts both their brand and the layout. It's worth noting that Ahrefs scanned 140 million websites and found that nearly 6% were accidentally blocking AI crawlers in the robots.txt file. So, if your site is that 6%, none of this matters. AI can't even see you. So, what do you do? One, go to your 10 best content pages. Ask yourself, could AI extract a clean two-sentence [music] answer from each section? If not, restructure it. Add clear headings, put answers at the top of each section. Two, add FAQ sections to your highest value pages. >> [music] >> Write the questions the way a real person would type them into ChatGPT. Three, check your robots.txt file and make sure you're not blocking GPT bot or Perplexity bot. This is the easiest fix on this entire list. Now, you're fixing your Google problem and your website problem. But, here's where most marketers trip up. They assume AI search is one unified ecosystem. It's not. And that assumption might be costing you more than anything else on the list. That's pattern number three. The platforms barely overlaps and most brands are only on one. When most marketers say AI search, they mean ChatGPT. But, AI search isn't one thing. It's five completely different ecosystems. Each pulling from different source pools, serving different audiences, and rewarding different types of content. ChatGPT has 1.2 billion users and drives 78% of all LLM referral traffic. Gemini has 750 million users and has grown 5x in visitors since 2024. ChatGPT only grew 1.64x in that same period. Claude is the fastest-growing referral traffic source right now. And meta AI, 1 billion users. Barely any markers are optimizing for it at all. And here's a mistake I constantly see. A brand tracks their chat GPT visibility, sees good numbers, declares victory on AI search. That's not the real case. Meanwhile, their actual buyers, who are Google workspace users, are living on Gemini, and Gemini doesn't mention their brand at all. The brands getting this right are doing something smarter. But before I go into that, if you just want my company NP Digital to just fix all of this for you and help you dominate the main LLMs, just hit us up. Now, one B2B SaaS company map which platforms their buyer personas use most. Finance directors, Gemini. Developers, Claude. Marketing teams, chat GPT. They built distinct strategies for each platform based on where that buyer actually asks questions. That's how you think about this. A, search your brand name and core topic in chat GPT, Gemini, Claude, and Perplexity separately. Screenshot all four. They will be different. B, ask in your sales calls which AI tools your buyers actually use. You'll be surprised. C, for each platform, identify what source types it prefers and whether you have presence there. Now, here's where most brands that do all of this still get it wrong. They assume that getting mentioned more, publishing more, producing more content is what builds AI authority. The data shows something completely different, and it has to do with how AI decides who to trust, not how often it finds you. That's pattern four. Entity association beats everything else. Traditional SEO optimizes for backlink count. AI search optimizes for something different. It's called association, and it's about how confidently that model connects your brand to a specific topic across the entire web, not just on your site. Everywhere. The volume of content on your own domain is almost irrelevant. 68% of marketers responding to the AI search challenge right now are publishing listicles on their own site and listing themselves as the best option. That builds neither real authority nor entity association. Think about it from the AI's perspective. ChatGPT is now weighing who sites you, not just how often. Self-promotion doesn't move the needle anymore. Third-party validation does. And Google has a knowledge graph with 54 billion real-world entities in it. If your brand isn't clearly defined in that system, >> [music] >> connected to a specific topic with third-party sources confirming it, you're effectively invisible to AI search. I've seen brands with 200 blog posts, strong domain authority, and solid Google rankings, but zero entity footprint outside their own site. AI doesn't know who they are. It can't confidently retrieve them. The brands they get cited? Wikipedia presence, industry report citations, podcast appearances, third-party reviews, niche publication mention. AI associates their name with a specific topic confidently because the whole web confirms it. Here's what you do. One, search your own brand in ChatGPT and look at what third-party sources it references when talking about you. Those are your current entity signals. Are they strong? Are they recent? Are they weak? Next, identifying top sources AI already trusts in your niche. Industry publications, Reddit communities, YouTube channels, review platforms. Those are the places you need to build presence. Then, build brand mentions through digital PR outreach, expert contributions, and product reviews. The goal isn't more content on your site. It's more trusted third parties confirming who you are. Now, there's one final pattern and it's the one that even the brands doing everything else right tend to overlook because it's not about what you publish, it's about what. That's pattern five. Fresh content gets cited disproportionately. A lot of brands have great content, well-structured strong entity footprint, good third-party mentions, and they still are not getting cited consistently because they haven't updated the majority of their content in years. AI models have a strong freshness bias. And look at this breakdown by model. GPT 5.2, 33% of citations came from content published in the last 30 days. GPT 5.3, 6%. Now, look at content age by platform. Google sites content that's about 130 days old on average. >> [music] >> ChatGPT, 80 days. Claude, 62. The freshness window keeps getting tighter. I've seen this played out with a brand that had what they called their definitive guide, published in 2022. Well-structured, great depth, strong backlinks, never updated, never cited by AI because newer fresher sources keep beating it out. Flip side, brands that are methodically republishing their highest value pages >> [music] >> every quarter, updating their stats, adding new data, refreshing examples, they're constantly showing up in AI answers. Not because AI cares that they worked hard, because fresh content signals that information is still accurate. Here's the play. One, identify your 10 highest value content pages, the ones that matter most for your business, and put them in a list right now. Two, build quarterly refresh cycles. Update the data, improve the structure, add new examples. Treat these pages like living documents, not finished products. These are five patterns. They all come back to the same question. Can AI confidently retrieve you as [music] a trusted answer? Not the most popular answer, not the highest-ranked answer on Google, the most retrievable one. From the most trusted sources in the clearest structure, confirmed by the right third-party sources, and updated recently enough to still be relevant. The brands winning right now aren't doing more than everyone else. They're doing the right things in the right places that make it easy for AI to find them, trust them, [music] and pull them into the answer. And here's what I want you to take away. This isn't about abandoning SEO. Google still matters, but you need a separate AI optimization strategy that runs alongside it because [music] the game has two boards now, and most brands are only playing on one. If you want to go deeper on how AI's changing search, I just put out a full breakdown on the entire AI search landscape, what changed, what it means for your traffic, and exactly what to do. That video is right here.