Skip to the tool

Free AI visibility reporting template

Build a reporting pack that does not invent the missing evidence.

Configure a small, versioned reporting pack for prompt runs, brand naming, source observations, accuracy and downstream evidence.

Browser only. No email. No prompt answers, names, confidential facts or client records belong here.

Configure the pack
Schmitdy dolphin arranging measurement instruments in a harbour workshop
A useful report makes its evidence boundary visible.

What this template keeps separate

Answer-engine observations are not web analytics. Identifiable referrals are not qualified enquiries. Booked demos are not qualified without the rule that defines them. A missing source stays visible as unavailable, not collected or not applicable.

Configure the report

Freeze the context before you calculate

Add one row for each engine, topic, market and language cut. Counts must be whole, observed cells, not estimates.

Use a safe reporting layerUse version labels and aggregate counts. Do not enter names, emails, prompt answers, confidential company facts or client records.
Unavailable means the source cannot be accessed. Not collected means it was available but not gathered for this period. Not applicable means the metric cannot exist for this cut. Observed means a dated, frozen collection exists.

Prompt-run context 1

Observed counts

Evidence summary

Evidence summary

Each percentage keeps its count and denominator. The template never blends these layers into a visibility score.

Answer engine 1

Unavailable

Completed prompt-run evidenceUnavailable
Brand naming rateUnavailable
Eligible-prompt coverageUnavailable
Owned cited-source coverageUnavailable
Owned retrieved-source coverageUnavailable
Factual accuracyUnavailable

Referral, qualification and booked-demo evidence

Referral, qualification and booked-demo evidence

Keep this downstream evidence outside the answer-engine figures. A booked demo needs the declared qualification condition before it can be called qualified.

Shipped interventions

Shipped interventions

Export JSON

Build a reporting pack that does not invent the missing evidence.

The template sends no reporting values to a backend, analytics or browser storage. JSON and CSV exports are created on your device. Standard privacy-controlled page analytics may run; this tool's own events contain only the locale and action name.

AI Search

Need an evidence-led starting point?

The free AI Search audit helps frame buyer questions, source gaps and a measurement plan before a reporting period is agreed.

Request a free AI Search audit

Your reporting pack stays in this tab

The template sends no reporting values to a backend, analytics or browser storage. JSON and CSV exports are created on your device. Standard privacy-controlled page analytics may run; this tool's own events contain only the locale and action name.

Use a safe reporting layer

Use version labels and aggregate counts. Do not enter names, emails, prompt answers, confidential company facts or client records.

FAQ

Before you report

Does this calculate one AI visibility score?
No. Naming, coverage, cited sources, retrieved sources and accuracy answer different questions. The template shows each separately with its count and denominator.
What happens when a source is missing?
Choose unavailable, not collected or not applicable. The template does not convert a missing source into zero or estimate a result from another source.
Can I compare engines or markets?
Yes. Add a separate engine cut and freeze its prompt-panel version, interface where known, market, language, run date and method note. Compare only like-for-like cells.
Does this track traffic, qualification or booked demos?
It records their evidence state and aggregate counts when you have them. It does not connect to analytics, CRM or booking systems, and it does not infer a qualified result.
Are the fictional example figures real?
No. The example is fictional and exists only to show report structure, unavailable states and export format.

All free tools