If you are tempted to buy into the promises of getting featured by AI, be careful! AI Citations are the wild wild west, just like when SEO first came on the scene. This page will update as new information – from reliable sources- is updated. The tutorials do not need my input or edits, so I will just link to them so you can catch up with the best and most authoritative sources on the topic.

AI Citations – is simply the mention of a website by the AI chat bot in response to a question.

AI bots – if you have banned AI bots then you will not be in AI Citations.

Zero-Click results is an SEO term. It describes what searchers do when they do not go to a website after searching. The more you look at zero-click research, AI Overviews and AI Mode, the more you’ll see how beneficial AI Citations really are.

Measuring citations, by the way, is not the same thing as measuring traffic from citations.

I looked at major AIO (AI Overviews)monitors like Otterly.ai and SEMRush. This is how we measure if you are cited by an AI bot:

  1. the tool sends a question prompt to AI,
  2. AI answers back,
  3. The answer is recorded

They record how often a company is shown, what % of the citations it occupies in that field, and the sentiment. Most programs do this daily against a list of question prompts that you supply. In the first tool, Otterly.ai the basic program gives you 15 question prompts. So 15 prompts each day sounds good in theory. In actuality, the results are completely meaningless.

To myself, I was thinking, This doesn’t make sense. If 3 of my friends ask ChatGPT that exact same question, we will all get different answers! How do we accurately track that? And Rand Fishkin (who is killing it on Linkedin) popped up in my feed with this report on new research by SparkToro and Gumshoe.ai. What we learned from that study is essentially:

To get mathematical about it, there’s a <1 in 100 chance that ChatGPT or Google’s AI, if asked 100X, will give you the same list of brands in any two responses. ” ~ Rand Fishkin (Tracking AI Visibility)

What this means, is that you would need thousands of data points to get actionable insights. A second thing coming out of that study is that visibility does remain consistent – it depends on the market. For example, their question about headphones resulted in the same brands 90% of the time. A question about marketing agencies, ran a top result of 30-40%. So is “good visibility” 30% or 90%?

FOR NOW: I don’t believe anyone should be putting any money into monitoring citations. Unless you run enterprise grade analytics.

…any tool that gives a “ranking position in AI” is full of baloney. ~ Rand Fishkin

You know what is a great idea? Tracking the traffic that is coming from AI! Use this DIY approach here using Google Analytics. At least you’ll know that someone is actually clicking through and visiting your site. This will tell you if you are visible.

There are a few proven things to do – but 99% of the stuff out there is complete and utter… um… crap. At best, the advice is good marketing redressed, at worst, it is conjecture where you put your reputation on the line to test it out.

Here is what we know and what we do not know:

  1. Overview of Best Practices for AI worthy content (trustworthy search engine journal)
  2. AI Favors These (Correlation only) by AHrefs.
  3. Google’s Guidelines on Appearing in Google AI
  4. Standard SEO still matters. Correlation says that oft- ranked pages will tend to do better than others in AI.

New! Welcome-Email AI Agent

Optin-Welcome Series GPT

Cathy Mitchell

Single Mom, Volunteer, Lifelong Learner, Jesus Follower, Founder and CEO at WPBarista.