How AI Visibility Is Measured for Dental

As practices start asking “are we showing up when patients use AI,” a certain kind of promise is going to start appearing: we’ll track every time ChatGPT recommends you. It’s worth knowing up front that nobody can do that, and anyone claiming otherwise is either confused or selling you something.

Here’s how this channel actually gets measured, because understanding it changes what you should expect from anyone doing this work.

Why tracking every conversation is impossible

When a patient asks ChatGPT or Google’s AI (or any AI platform for that matter) for a dentist, that conversation is private. It happens between the patient and the AI. It doesn’t pass through your website, it isn’t logged anywhere you can access, and there’s no report that lists every time your practice was mentioned in someone’s private AI chat. The data simply isn’t available to anyone outside the AI companies themselves.

So a tool that promises to show you “every AI citation you received” is promising something that doesn’t exist.

What can actually be measured

You can’t watch every conversation, but you can take a repeatable sample of the landscape, which is the only way to really measure AI citation success.

The approach works like this. You build a fixed panel of prompts, like the real questions you anticipate patients to ask, like “who’s a good implant dentist near me” or “how much do veneers cost”, and you run that same set of questions against the AI engines, repeatedly, over time. Because the questions stay constant, you can measure real things. Such as, how often your practice shows up in the answers compared to competitors, which of your pages get cited, and how the AI describes your brand.

That measure is called Share of Voice. It’s not a count of every mention in the wild. It’s a repeatable sample that tells you whether you’re gaining or losing ground.

Why a sample is the right tool

A consistent, repeatable sample tells you what you actually need to know (such as the direction and the trend) without pretending to a completeness that no one can deliver.

If you run the same 100 patient questions this month and next month, and your practice went from showing up in 10 of the answers to showing up in 25, that’s real, measurable movement. And because the method is repeatable, it’s consistent, so the comparison is honest.

What this means for you

When you evaluate anyone doing AI-visibility work, the tell is in how they describe measurement. “We track every time AI mentions you” is a red flag, because it’s not possible. “We measure your Share of Voice against competitors across a consistent set of patient questions, and track whether it’s improving” is the answer that wins.

We use the second approach because it’s the only one that’s true, and because it’s the only one that can actually tell us whether the work is moving the needle. 

Let’s Talk