If you want to get recommended by ChatGPT or Google's AI mode, it certainly helps if you rank well in traditional search results. And this is because these large language models use a search engine. When someone goes to ChatGPT or Google's AI to ask for a recommendation, the first thing it'll typically do is a series of searches in a traditional search engine.
And the response from the model is more or less a summary of the pages found in those searches. As long as you rank well in traditional search, you'll show up in the AI responses. And while there's a lot of truth in this, it really misses the Mark.
So let's talk about how traditional search engines work and why your ranking in these search engines isn't the entire story when it comes to AI visibility. And for the sake of time, I'm gonna give you the super simple version. There are many factors that determine your ranking in search engines, but those factors could be broken into three categories.
Most importantly, the content on the page needs to be relevant for what the person is searching for. However, for most searches, there are going to be many relevant pages. So Google needs some way to determine which ones are the most authoritative.
The primary factor used to measure authority is called PageRank. PageRank is determined by the quality of the links pointing to the page you're trying to rank and the quality of a Link is determined based on how many links away it is from the most authoritative websites on the internet. And the third category is user behaviour.
Google will pay attention to which websites people click on after they perform the search and whether or not they go back to the search results and click on another page. Google uses this behavioural data to determine if your page satisfied the searcher's intent. Depending on how well Google feels your content is satisfied the searchers intent, they will move you up or down in ranking.
Now of these three categories, two of them are actually very easy to optimize for relevancy and user behaviour are almost entirely determined to buy the content on your page and you have full control over the content on that page. PageRank on the other hand, it takes a lot of time and resources to build but here's a thing that I think is crucial to understand. PageRank is only important if you're trying to compete with pages that have more PageRank than you.
So yes, if you're trying to rank for a very competitive term you're gonna need a high PageRank but here's why GEO is different. Large language model doesn't search for the same competitive terms that humans do. Large language models do something called query fan out where they'll take your prompt or search and do a series of more specific or what are called long tail searches.
Because the searches are more specific, there's a lot less competition. In a lot of cases, the searches are so specific that there's not a single page on the internet targeting that exact term. So if you can track the exact queries that these large language models are searching for and the types of pages that they site, you can create hyper relevant pages that perfectly match search intent.
And because you have no direct competition, PageRank matters a lot less. Now these pages aren't going to get hundreds of visits a month. They might get a couple visits a month.
So for this to work, you need to create a lot of hyper specific pages, but the visits they get and the recommendations you get from large language models as a result and to convert at a much higher rate. Visitors from large language models convert at about eight times the rate as traditional search engines. So as long as you have a process for creating these pages efficiently, it's well worth it.
So yes, PageRank and organic ranking still matter for AI visibility. It broadens the number of terms that your site can rank for. But AI search also opens up opportunity for lower PageRank sites to get visibility.
