Eighteen months of tracking AI visibility as a lead metric moved a medical research non-profit from absent in AI-generated answers to a source those systems cite by default.
13x
Growth in AI-driven traffic. From a baseline of ~100 visits a month to 1,283.
636
Citations, up from 15 at the start. AI-generated answers citing the site grew by a factor of 42.
40%
Improvement in average citation position.
Project Overview
Background
A non-profit research institute came to us with a question most marketing teams can’t yet answer: when people ask ChatGPT, Google AI Overviews, Gemini, or Perplexity questions in their field, does their site show up in the answer?
At the start, it didn’t. Their AI-driven traffic was effectively zero, and nobody on their team had a way to track whether that was changing.
The Approach
We built AI-visibility tracking into their monthly reporting from day one and treated it as a lead metric, not a footnote. Every reporting cycle, the data showed us two things: which existing content was close to earning a citation and needed strengthening, and where a citation gap signaled a topic worth building from scratch. Both fed directly into the content plan — and the cadence held for a year and a half, not just a single campaign push.
Project Outcomes
When a specialist asks an AI tool a question in our clients field, they’re now part of the answer — and has been, month over month, for a year and a half.
From a rounding error to 1,283 visits a month
At the start, AI tools sent roughly 100 visits a month. That’s noise against every other channel, and not enough for anyone to act on. By June 2026 it peaked at 1,283 — a 13x climb on the same site, for the same audience.
The research went from unmentioned to cited 636 times
Citations count how often an AI-generated answer names the site as a source. There were 15 at the start. Every one added since is a researcher’s question where the Institute’s work made it into the response.
From the bottom of the source list to the top
Position measures where the site ranks among the sources an AI answer pulls from. In April 2025 the average sat at 16.3 — far enough down that almost no one reaches it. Through 2026 it held between 5.9 and 7.2.
Start a conversation
The first call is an honest assessment, not a pitch.
30 minutes to figure out whether we’re the right fit. If we’re not, we’ll tell you who is. Month-to-month contracts. If you don’t approve the strategy, we refund your deposit.