Prompt coverage
Updated 2026-09-15
Prompt coverage is the share of tracked prompts in which a brand appeared at least once across the sampled answers. It measures breadth: how much of your question space you show up in at all.
How it is calculated
For each prompt, the brand either appeared in at least one valid sampled answer or it did not — a binary outcome per prompt. Coverage is the count of prompts with at least one appearance divided by the number of prompts with valid answers. Because it ignores how often you appeared, it is deliberately insensitive to strength and sensitive to gaps.
Worked example (illustrative numbers)
A 30-prompt set produces mention rate 0.23. Coverage shows 7 of 30 prompts ever surfaced the brand, so coverage is 0.23 as well — but the shape is different: the brand is reliably present in seven questions and completely absent from twenty-three. The content work is in the twenty-three, and only coverage shows it.
Common mistakes
- Reporting coverage without listing the zero-coverage prompts, which throws away the most actionable output of the whole measurement.
- Padding the prompt set with easy branded prompts, which inflates coverage while telling you nothing about discovery.
- Comparing coverage across cycles after editing the prompt set — the denominator changed, so the trend is not a trend.
Frequently asked questions
- How is prompt coverage different from mention rate?
- Mention rate measures how often you appear across all answers; coverage measures how many distinct questions you appear in at all. Strong mention rate with weak coverage means you own a few questions and are invisible in the rest of the category.
- What is a zero-coverage prompt?
- A tracked prompt where the brand never appeared in any sampled answer during the period. Those prompts are the practical content backlog: each one is a question your buyers ask that AI answers without you.
