The New SEO Playbook for AI Search (Top GEO Ranking Factors)

Search has finally evolved and now everyone wants  to rank in AI search. But how? Our brilliant team   at HRF 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 branded mentions.   Large language models like chatbt 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 O,   the charismatic SEO beast. Branded mentions are so  important that they have the strongest correlation   with visibility in Google's AI overviews. higher  than back links, 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 chat  GPT and perplexity. To find pages and websites   worth getting mentioned on, go to HF's brand  radar and enter your brand, competitors, 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 an AI search. Longtail queries.   In traditional SEO, targeting longtail 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, longtail queries   play a different role. AI assistants like Chat GPT  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   longtail 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 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 Chat GPT 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 seven 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 Chat 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, augment, and 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,00   new mentions according to HR 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 redate 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 yourdommain.com/roots.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 data set,   we compared the top 50 most mentioned domains in  Google AI overviews, chat GBT, and Perplexity. And   get this, only seven 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. Chat GPT prefers publishers and news  outlets like Reuters and the Associated Press. And   Perplexity tends to site 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 HF's 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 web pages 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 drop  down, 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 LMO is still just SEO. The difference  is that you got to 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.