AI Search
AI search is the “new” thing that many are talking about in marketing and SEO. The goal is simple: To be seen, mentioned, and recommended when someone searches or asks a question on any AI service such as: Google (AI Overviews, AI Mode, Gemini), ChatGPT, Claude, Perplexity, or other AI services.
How do you optimize for AI search?
The simplest and shortest answer is: Traditional SEO.
All the rules and guidelines that have come and evolved within SEO for many years are still relevant. The three pillars of SEO have always been:
- Technology: Have a good technical foundation
- Content: Write and publish good and relevant content about what you want to be seen for
- Links: Get, earn, and acquire good links and mentions
All three parts still apply 100%!
Difference between SEO and optimizing for AI search
AI cares less about which sites are presented compared to the regular search results in Google, which are often dominated by the most well-known names in each industry. The reason is simple. They don’t know or have significantly less knowledge. Since its inception, Google has developed various algorithms to determine how credible, strong, and relevant a page is using links. Information that (most) AI services do not have access to. Many solve it by retrieving search results from various search engines and then processing the results to ultimately deliver the best possible results. And that’s where smaller sites can slip in because little or no checking is done to see if the page is brand new, has been around for 10 years, and who is behind it.
Content – Still king
When you do an AI search, a number of “query fan outs” are created. Variants or interpretations of your search. The new variants are then used to perform new searches. All the results are compiled and processed to provide relevant answers, information, and recommendations for your search. So even if you are highly visible in one or more of those searches, it does not guarantee that you will be one of the sites mentioned. But logically, you should become one of them if you are visible and relevant in all the query fan outs that are made. So, depending on the content today, it may mean that you should supplement with content and possibly create and optimize a couple of extra pages to cover as much as possible. Just like with traditional SEO, in other words.
AI search = GEO, AEO, LLMO
You have probably seen abbreviations such as GEO, AEO, LLMO, etc. mentioned as the “new” thing. The truth is that it is just the same as SEO or “SEO for AI”. The reason why all the new abbreviations have emerged is, in our opinion, that several people and companies wanted to create a new industry when it is not really needed. Those people and companies often want to sell services, sell tools, or be experts in GEO, AEO, LLMO. Because if you already work with SEO (as most large companies do), then you also have to work with GEO and track it with a GEO tool. Because for some reason, the information you get in, for example, Google Analytics is not enough according to them.
AI search tracking
Tracking AI search is difficult today, and it will take a long time before there are good and cheap ways to get a good overall picture of what the visibility looks like. There are 1000s of tools already, and new ones are constantly popping up. Often, they only track a couple of arbitrarily selected searches/prompts. Then you get an answer as to whether your site or brand is mentioned in connection with those searches. And the cost for each search that you choose to track usually costs around 20 SEK/each. And then you should not forget that the proportion that uses AI today is still small.
The tools that provide the best overview today are old established SEO tools such as Ahrefs. There you get the following overview for your site under “AI-citations”:
So in many ways, the following picture summarizes quite well what AI search is like today:
Aware that the figures are not correct for Sweden. Google has about 90% share in Sweden. The figure for ChatGPT varies depending on the site and industry. But there are few examples of sites that exceed 1%.
If you want to delve deeper into AI search, I can recommend the following video from Lily Ray, who goes through AI search in 35 minutes. You can find it here on YouTube.