Step 5: Score each check
Turn the rows into scores for each check date, using the formulas in the next section. Add one line per check to a trend chart.
Step 6: Compare against the last check
List the prompts you won or lost since the last check. For each lost prompt, read the answer text and the sources ChatGPT used.
Loss across many promptsA loss across many prompts at once can point to a model change.
Loss on a single topicA loss on a single topic can point to a content or source gap.
Which metrics show ChatGPT visibility over time?
Visibility rateRuns that name your brand, divided by all runs. This is your headline number.
Prompt coveragePrompts where you appear in at least one run, divided by all prompts.
Stable coveragePrompts where you appear in every run, divided by all prompts.
Citation rateRuns that cite at least one of your URLs, divided by runs that used search.
Share of voiceYour brand mentions, divided by mentions of you and your tracked competitors.
Prompt coverage shows how wide your presence is. Visibility rate and stable coverage show how dependable it is.
Read how prompt coverage is counted and how an AI visibility score combines several signals into one number.
Researchers reached a similar view in the GEO paper on generative engine optimization. They argue that a simple ranking doesn’t capture visibility inside a generated answer.
A free ChatGPT visibility tracker template
Use one tab per check date and one summary tab. Here is an illustrative check: 40 prompts with 3 runs each, which gives 120 runs in total.
Visibility rateThe brand is named in 30 of 120 runs, a visibility rate of 25%.
Prompt coverageIt appears at least once for 16 of 40 prompts, so prompt coverage is 40%.
Stable coverageIt appears in all 3 runs for 6 of 40 prompts, so stable coverage is 15%.
Citation rateOf the 90 runs that used search, 18 cited the brand’s site. That’s a citation rate of 20%.
Next week, the same summary shows whether a change came from new prompts or from steadier answers on prompts you already had.
Add a notes column for events such as new pages, press coverage, pricing changes or a ChatGPT model update. When the trend moves, that column often explains why.
What our tracking data shows about ChatGPT visibility
Ansvisor tracking dataAnsvisor, an open-source AI search intelligence platform, tracks prompts about AI search for its own brand.
From September 15 to October 14, 2026, those prompts produced 103,427 AI answers and 746,702 citations across all tracked AI platforms.
A brand’s share of voice can differ several-fold between AI engines, even for the same prompts and competitor set.
In the same period, ChatGPT cited about 5 links per answer across 16,582 tracked results.
A brand can look healthy in one assistant and be nearly invisible in another. Track ChatGPT as its own channel.
Manual tracking vs a ChatGPT visibility tracker
The manual method is fine for a first audit. It gets slow once you track more prompts, more markets or several competitors every week.
Manual checksManual checks are free and flexible, and they cope with 20 to 30 prompts in one market.
ChatGPT visibility trackerA tracker runs prompts on a schedule, stores every answer and calculates the scores for you.
You can also combine them. Keep a small manual sample to spot-check what the tracker records.
Before you choose a tool, ask how it runs prompts, how often it repeats them and whether it shows the raw answers.
The score is where analysis starts. Dig into how ChatGPT describes your brand, how you compare with competitors and how much traffic ChatGPT sends you.
How to improve weak ChatGPT visibility
1. Missing promptsStart with the prompts where competitors appear and you don’t.
2. Cited sourcesRead the sources ChatGPT cited for those prompts. They show what it relied on.
3. The buyer’s questionAnswer the buyer’s question directly, on a page that OpenAI’s search crawler can access.
4. Third-party mentionsEarn mentions on third-party pages ChatGPT already cites, such as reviews and comparisons.
5. The re-checkRe-check the same prompts two to four weeks after each change.
How Ansvisor helps you track ChatGPT visibility
Ansvisor runs your ChatGPT prompts on a schedule and stores every answer along with its mentions and cited URLs. You see visibility, prompt coverage and share of voice per prompt and over time.
With answer engine insights, you can compare the same prompt set across Gemini, Google AI Overviews, Google AI Mode and Copilot.
You can use the cloud app or self-host the open-source project on GitHub.
Start a 14-day free trial and run your first ChatGPT prompt set this week.
FAQ
How do I track my brand’s performance in ChatGPT responses over time?
Run the same prompt set on a fixed schedule and log whether each answer names your brand. Calculate visibility rate and prompt coverage for every check. Plot both over time and note launches, content changes or model updates next to the trend.
Can I check brand visibility in ChatGPT for free?
Yes, for a small prompt set. Use a Temporary Chat with personalization turned off, run each prompt a few times and log the results in a spreadsheet. A tracker earns its keep once you need more prompts, more markets or a longer history.
What does a ChatGPT visibility tracker measure?
It measures how often ChatGPT names your brand across tracked prompts, which of your pages it cites and which competitors appear instead. Most trackers also keep answer history, so you can see when visibility changed.
Does ChatGPT have rankings like Google?
Not fixed ones. ChatGPT writes a new answer for each request, and the brands and their order can change between runs. Visibility rate across many prompts and runs is a more reliable measure than a single list position.
How often should I track ChatGPT visibility?
Weekly works for most brands, with a fuller review each month. Keep the prompts and settings the same on every run, in the same market. Check more often after a product launch, a site migration or a major ChatGPT model update.
AI Overviews
AI Mode
Copilot
Perplexity
Claude