The New SEO Playbook for AI Search

Search has finally evolved, and now everyone wants to rank in AI search. But how? Our brilliant team at Ahrefs studied 75,000 brands and 25 million AI overviews and found five key factors that give you the best chance of ranking in AI search with GEO. And it all starts with brand inventions. Large language models like ChatGPT and Google learn by reading the web. And every time your brand name appears on a credible site, that becomes another training example. And the more often an LLM sees your brand connected to a specific topic, the more confidently it associates your name with that subject when generating answers. It's like when you hear Red Bull, you think extreme sports. Or when you hear Tesla, you think electric cars. Or when you think of Ahrefs, you think of Sam Oh, the charismatic SEO beast. Brand inventions are so important that they have the strongest correlation with visibility in Google's AI overviews, higher than backlinks, referring domains, and domain rating. But how do we know which pages or sites to get mentioned on? According to our research, get mentioned on highly linked to pages if you want to rank in Google AI overviews and high traffic pages if you want more visibility in ChatGPT and Perplexity. To find pages and websites worth getting mentioned on, go to Ahrefs' brand radar and enter your brand, competitor's brands, niche, and run the search. Then hit the Cited Domains report, where you'll see the top websites, the number of AI responses, and the exact number of times each brand is getting mentioned and how they're being mentioned. From here, you can join the Reddit threads AI is pulling from, reach out to YouTubers for future reviews, or contact publishers for potential PR campaigns. Basically, you want your brand to appear literally everywhere online, ideally in a positive and topically relevant way. And that's because of the second key factor for ranking in AI search, long-tail queries. In traditional SEO, targeting long-tail keywords was mostly about coverage. The more queries you rank for, the more ways there are for searchers to find you. But in AI search, long-tail queries play a different role. AI assistants like ChatGPT may actually use them to decide who to recommend. And it's because of how AI assistants work. When someone enters a prompt like, plan me a 5-day trip to Japan in November, the AI assistant fans it out into dozens of smaller long-tail subqueries. They then pull information from multiple sources across the web and combine it into one complete answer. 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 our data backs that up, showing that Google's AI overviews are way more likely to appear on longer niche queries. So from a practical standpoint, you should A, create content that answers more complex, specific questions, and B, build content clusters that deeply cover the full depth of a topic. But here's the twist. Even if you've nailed all of this, it might not matter. Because if your content isn't structured the way AI assistants process information, you may not get as many AI search mentions as you should be getting. One way Google AI reads content is by using a tree-walking algorithm, meaning it follows the exact semantic HTML structure of a web page from top to bottom. And this makes well-formatted and structured content easier for it to process. But it's not just about using heading tags and bulleted lists. It's about how your information flows. Your content should be clear, organized, and easy to follow. And this matters because when Google or ChatGPT read your page, they don't process it all at once. They chunk your content into smaller pieces, almost like reading it paragraph by paragraph. And then they decide which parts are most useful to keep. So if your juiciest points are buried deep down or mixed into long, messy sections, it makes it less efficient for AI to process your content, which may have a role in rankings. Now, that doesn't mean you should start writing in short, choppy blurbs like an FAQ or make every paragraph its own idea. Instead, focus on grouping related thoughts together and keeping each section focused on a single takeaway. Think of it like giving AI a clear outline where each part should make sense on its own, but still naturally connect to the next. Now, even if your content is perfectly written and structured, there's another key factor that can put you in or completely knock you out of AI results. And our data shows that this key factor is all about freshness. Our study of 17 million citations across 7 AI search platforms found that AI assistants strongly prefer newer information. In fact, content cited by AI is 25.7% fresher than the content that shows up in regular Google results. And Chats GPT and Perplexity go even further. They tend to list citations from newest to oldest. That happens because most assistants use something called RAG. RAG stands for Retrieval Augmented Generation, which is the part of an AI system that goes out and looks things up. If a model already knows the answer from its training data, it doesn't need to search again, just like you and I don't need to search for directions to places we visit all the time. But if the topic is new or evolving, RAG steps in and fetches the latest information from the web to fill in the gaps. That's why freshness isn't just a ranking signal anymore, it's a retrieval signal in this AI era. And here's proof that freshness can boost your AI rankings. In April this year, HubSpot updated their Small Business Ideas article and soon saw a massive spike to their organic traffic. No surprises there. But their AI mentions to that page also spiked at the exact same time, generating over 1,100 new mentions according to Ahrefs' brand radar. So something you can do is to take note of pages that could benefit from regular updates and add them to your regular refresh cycle. Update facts, new quotes or stats, remove things that have become irrelevant, and re-date the post when it's meaningful. And that one small update could be the difference in getting mentioned across AI platforms. By the way, I probably should have mentioned this earlier, but before you do anything, make sure your site can actually be crawled by AI bots. Kind of sounds silly, but in our study of 140 million websites, about 5.9% were blocking OpenAI's GPT bot alone. So take two seconds to go to yourdomain.com slash robots.txt and make sure you're not accidentally blocking the very systems you're trying to rank in. And that brings us to our final, easiest, and perhaps most important key factor. And it's all about diversifying your strategies. Normally when I talk about diversifying your traffic or marketing strategy, it's because over-reliance on a single channel can wipe a business out overnight, which we've seen happen time and time again. But that's not the reason you should diversify in the context of AI. In our dataset, we compared the top 50 most mentioned domains in Google AI Overviews, ChatGPT, and Perplexity. And get this, only 7 of those domains appeared on all three lists. That means that 86% of sources are unique to each assistant. Google AI Overviews leans heavily on YouTube, Reddit, and Quora. ChatGPT prefers publishers and news outlets like Reuters and the Associated Press. And Perplexity tends to cite niche or regional sites like health, finance, and localized blogs. So even if you dominate one of these ecosystems, that doesn't mean that you're guaranteed to appear in the others. Hence, diversification is a must. To find the pages you need to optimize for each AI platform, go to Ahrefs Brand Radar and add your brand, a few of your competitors, a niche, and run the search. Hover over your brand under the AI platform you want to investigate and hit others only. Finally, go to the Cited Pages report to see all of the webpages that are being recommended in your AI platform of choice where your competitors are mentioned, but you're not. Then go and reach out to these websites to try and get your brand mentioned alongside your competitors. And when you're done with that AI platform, hit the dropdown, choose the next AI platform you want to be mentioned on, and rinse and repeat. Now, a lot of this advice might sound familiar because doing GEO or AEO or LLMO is still just SEO. The difference is that you gotta think broader. It's not just about content and links anymore, it's about building a brand and showing up everywhere your audience already is. But here's the truth no one's really saying out loud. No one actually knows how to definitively optimize for AI search yet. It's still too early. For now, all we have is the data we've analyzed, and now we need more practitioners to go out there, test, and share the results. And one of those practitioners is already testing, so we can build that step-by-step playbook for you to follow. So make sure to subscribe for updates, and I'll see you in the next video.