Zero-coverage prompt
Updated 2026-09-15
A zero-coverage prompt is a tracked question where the brand appeared in none of the sampled answers during a period. It is the practical work list that visibility measurement produces: a question your buyers ask that AI answers without you.
How to use the list
After a cycle, prompts split into three groups: reliably covered, intermittently covered, and zero-coverage. The zero list is then read rather than counted — which intent class do these prompts belong to, which sources do the engines cite instead, and is there a page of yours that should have been the answer. That reading turns a metric into a content backlog with a defined success test: the same prompt, next cycle, same N.
Illustrative list (example numbers)
Of 35 tracked prompts, 23 are zero-coverage and 19 of those are discovery questions answered mostly from two review sites. The backlog is not "write more content" — it is four pages answering those specific questions well enough to be cited alongside the review sites.
Common mistakes
- Reporting the count without the list, which throws away the only directly actionable output.
- Attacking the hardest prompt first. Intermittent coverage usually converts faster than a prompt you have never appeared in.
- Rewriting the prompt instead of the content when a prompt stays at zero — that hides the gap rather than closing it.
Frequently asked questions
- Is a zero-coverage prompt always worth fixing?
- Only if buyers actually ask it. Some prompts stay at zero because the question belongs to a segment you do not serve, and removing them from the set is more honest than chasing them.
- How fast should coverage change after publishing?
- Expect cycles, not days: the page has to be crawled, indexed, then retrieved and preferred. Measuring the same prompt at the same N each cycle is what makes the eventual change believable.
