AI citation diagnostics

Ranking an AI search is so easy right now, and I think the main reason is that no one's tracking it. When I talk with businesses that are doing SEO, or even other SEO agencies, they often talk to me about manually going to Google's AI mode or ChatGPT and asking about their industry to see if their brand is mentioned.

And that blows my mind when the software exists to track this at scale and not just to see where your brand and your competitors are mentioned, but why. So I want to talk about the software we use, why tools like hrefs and Semrush aren't sufficient, and how this data informs most of our strategy.

But first, if you're thinking, why should I care so much about ChatGPT and Google's AI mode, Google's already told us that all search will be AI search soon. I think it'll happen by the end of the year.

And when you start tracking AI search results, you see they're quite different from traditional search results. So while traditional search isn't dead yet, I do think our focus should be on AI search entirely...

y good idea of what people are searching for before they hire a company like them. The goal should be to gather a representative sample of prompts people might type into AI.

And in most of those prompts, you should include a demographic or a use case. Something like, I'm an X looking for y to solve Z.

Can you recommend a company? The idea is to simulate the information an AI would have when it does a search and returns a recommendation.

Like 50 or more prompt examples like this, you can use AI to help you and then start tracking them every day. In ChatGPT, Google's AI Mode, and Google AI Overview, we use peak that AI.

I'm a big fan of their team. But there are a lot of tools to do this.

Since these are probabilistic models, you want to run these prompts many times. We run them four times a day at first and then go down to once per day.

And this will, of course, help you see how visible you are in AI recommendations. But the real magic is in the citations...