Clients can watch their Google rankings move in a dashboard any morning they like. AI visibility has no such dashboard, which is why most agencies either ignore it or hand-wave it. We track it manually, on a fixed routine, and report it monthly next to the ranking data. Here is the routine.
Build the query set once
For each client we maintain 10–20 buyer-intent questions — the things a real customer would ask an assistant. "Best X in city." "X vs Y for small teams." "Who should I call for Z." The set stays fixed month to month; consistency is what makes the trend line honest.
Run, log, screenshot
On the first of the month we run the full set across ChatGPT, Claude, and Google’s AI Overviews. For every answer we log three things: was the brand named, was it cited as a source, and which third-party domains the engine leaned on. Screenshots go in the client folder — when your brand shows up in a ChatGPT answer for a money query, the screenshot is worth more than any chart.
The domains an engine cites for your category are not trivia. They are next month’s target list.
What counts as a win
- Named in the answer: the model knows you exist in this category.
- Cited as a source: the engine treats your footprint as evidence.
- Your placements appearing in its source list: the flywheel is connected.
Movement is slow, lumpy, and impossible to guarantee — we say that in writing. But after a quarter of consistent placements, the trend line usually starts answering for itself.
Talha Mahmood
Founder of Editorial HQ. Writes up what the team learns from placing links and PR for clients.
Related service
See how our Brand Mentions service works