How to track your brand mentions in ChatGPT, by hand and at scale
By haylow.ai editorial team
In short
Tracking brand mentions in ChatGPT means asking a fixed set of customer questions repeatedly and recording whether your brand is named, in what position, how it is described and which sources are cited.
Why is it hard to track brand mentions in ChatGPT?
ChatGPT writes a fresh answer each time. The same question can name different brands on different runs, depending on wording, settings and the sources it draws on. There is no ranking page to look up, so you have to sample answers and measure how often you appear.
That is why one screenshot proves little. A reliable picture comes from repeating the same questions and counting.
How do you build a question set?
Write fifteen to thirty questions your customers would ask, across category, comparison and problem types. Do not include your brand name in most of them. Lock the wording, give each question an ID and keep the list stable so results are comparable week to week.
Category questions
What are the best tools for X, which companies offer X.
Comparison questions
A vs B, alternatives to B, is A worth it.
Problem questions
How do I do X, how do I fix X.
What should you record for each answer?
Record the date, the question ID, whether your brand is named, its position in any list, the exact wording used about you, the competitors named and any sources cited. Keep the full answer text. Those fields let you compute mention rate, average position and sentiment later.
A simple spreadsheet is enough to start. Add one row per question per run.
How do you turn answers into numbers?
Mention rate is the share of runs in which you are named. Average position is where you appear when named. Sentiment is how the answer describes you. Source share is how often each website is cited. Track all four per question and per engine, not as a single blended score.
Compare against the two or three competitors named most often. Your numbers only make sense next to theirs.
When should you move from a spreadsheet to a tracker?
Move when you need more than one engine, more than one market or a weekly trend, because manual runs stop scaling at around fifty questions. A tracker runs the same set, stores every answer and shows trend, position and cited sources without copy and paste.
haylow.ai runs your questions against the assistants' official APIs when you start a run, and keeps every answer so you can read what was said.
Does ChatGPT track brands differently from other assistants?
Yes. Each assistant draws on different sources and handles browsing differently, so a brand can be named often in one and rarely in another. Track each engine separately and compare, instead of assuming ChatGPT results apply to Gemini, Claude or Perplexity.
If ChatGPT is your main concern, start there, but add a second engine early to see how large the differences are.
Key takeaways
- ChatGPT answers vary between runs, so track a mention rate, not a single result.
- Use questions customers actually ask, written once and never edited.
- Record position, wording and sources, not just whether you appeared.
- Manual tracking works for a first look; a tracker takes over when you need trend and scale.
Frequently asked questions
Can I track ChatGPT mentions for free?
Yes, by hand: run a fixed question set and record the answers in a spreadsheet. It works for a first look but takes time to repeat.
How many times should I run each question?
At least two or three times on different days to see how much answers vary. A tracker can repeat runs for you.
Does asking about my brand by name count?
It measures recall, not discovery. Most of your questions should be ones where the customer does not already know you.
What is a good mention rate?
There is no universal number. Aim to rise over time and to be named at least as often as the competitors that appear alongside you.
See what AI says about your brand
Add one domain and get your first visibility report across five assistants. Free to start, no card required.