How to track your brand mentions in ChatGPT and AI search

Your brand shows up in AI answers — or doesn't — every time a buyer asks. And unlike a Google ranking you can look up once, AI mentions shift constantly: engine to engine, week to week, and even run to run. Tracking them is how you catch what's working, spot where competitors pull ahead, and notice when an engine starts describing you wrong. Here's how.

What "tracking AI mentions" actually means

It's monitoring, over time, whether AI engines name your brand when people ask questions in your category — across ChatGPT, Perplexity, Gemini and Google's AI Overviews. Not just "did I get mentioned once", but how often, next to which competitors, from which sources, and whether that's trending up or down.

The manual way (and why it breaks)

You can do a first pass by hand: list the buying-intent questions your customers ask, put them to each engine, and note whether you're named. It's a useful reality check. But it falls apart as ongoing tracking, for three reasons:

• Answers are probabilistic — ask the same question twice and you can get different brands, so one check proves nothing.
• There are several engines, each with its own answers — coverage means repeating everything across all of them.
• It's a moving target — models update and sources change, so a one-time snapshot is stale within weeks.

What to track

Focus on the signals that tell you something you can act on: your <strong>presence rate</strong> (how often you're named), your <strong>share of voice</strong> versus each competitor, the <strong>sources</strong> engines cite, and <strong>sentiment</strong> when you are named. (For how these metrics work, see our guide on measuring AI visibility.)

How to track it per engine

• ChatGPT — ask your category prompts; with search on, note which sources it pulls. Repeat across sessions.
• Perplexity — the most transparent: it lists cited sources under every answer, so you can see exactly where a mention came from.
• Gemini / Google AI Overviews — check whether your buying-intent queries trigger an AI answer and who it names and links.
• Across all — run each prompt several times; treat the aggregate, not any single answer, as the signal.

Why automated tracking wins

Doing that by hand, weekly, across engines and dozens of prompts, is a part-time job. A tool automates the sampling — running many prompts across engines on a schedule, aggregating into presence rate and share of voice, capturing sources, and alerting you when something moves. That turns "I think we get mentioned sometimes" into a number you can watch and improve.

Set up ongoing monitoring

1. Baseline: run a full check to see where you stand today across engines.
2. Track competitors alongside you, so presence is always relative.
3. Re-measure on a schedule and watch the trend, not the snapshot.
4. Act on the gaps — the sources that cite competitors but not you.

Start with a baseline

Run a free AI visibility check to see how often ChatGPT, Perplexity and Gemini mention your brand right now, your share of voice against competitors, and the sources they cite — then track it over time.