ChatGPT prompt volatility tracking
chatgptprompt_volatilityAnonymous
Volatility on ChatGPT is dominated by whether an answer triggers web search at all. Two samples of the same prompt can take completely different paths, and the metric captures that.
How it renders
Derived across the sample set.
Sampling
Volatility is what tells you the right n for every other field. We recommend using it to allocate credits: sample noisy prompts deeply, stable prompts lightly.
Gotcha
Search-triggered and pure-model samples of the same prompt are not comparable. We report the search-trigger rate alongside volatility so you can see which mixture you measured.
What the API returns
A single collection call against ChatGPT for the prompt volatility field, sampled n times and returned as one object per prompt.
{
"engine": "chatgpt",
"field": "prompt_volatility",
"prompt": "best ai visibility tracking tool",
"country": "GB",
"language": "en",
"n": 30,
"volatility": {
"jaccard": 0.42,
"mention_variance": 0.11,
"recommended_n": 75,
"band": "moderate"
}
}Response fields
| Key | Type | Notes |
|---|---|---|
| jaccard | float | Mean pairwise citation-set overlap across samples. |
| mention_variance | float | Variance in brand presence across samples. |
| recommended_n | integer | Samples needed for a ±10 point interval on this prompt. |
| band | string | stable | moderate | volatile. |
ChatGPT prompt volatility FAQ
- Why does the same ChatGPT prompt behave differently between samples?
- Search triggering is itself non-deterministic. Some samples ground on the web, some answer from the model alone.
- Should I filter to search-backed samples only?
- For citation metrics, usually yes. For mention share, no — the pure-model answers are what most users see.
Understand what your ChatGPT numbers are actually made of
We'll run a volatility and search-trigger pass across your prompt set.
Other ChatGPT data
Across every engine
Compare how prompt volatility behaves on every engine we track.
Prompt volatility across all engines →