# Reusable client reporting template

## Field provenance

This is a manually assembled Markdown report. No automatic PB client-report export is implied. “Product-derived” means copied from approved saved evidence and checked by the report author; it does not mean the template automatically fetches it.

| Field group | Provenance | Required author work |
| --- | --- | --- |
| Site, collection dates, exposed provider context, recorded statuses/counts | Product-derived where available | Select the correct records; mark unavailable fields |
| Prompt text, aliases, source relationship/ownership, recorded answer evidence | Product-derived where available | Confirm current definitions and approved evidence access; use product view when API gives metadata only |
| Report period, business question, executive summary, prompt groups and comparable subset | Manual | Define the analysis and disclose exclusions |
| Appearance/recommendation rates and owned-citation response rate | Manual calculations from reviewed records | Recalculate numerator/denominator; reconcile with the metrics and methodology reference |
| Distinct cited URLs | Product-derived count or manual calculation, explicitly identified | Preserve relationship, URL-normalization rule and scope |
| Health findings and check dates | Product-derived where available | Relate to the exact URL and explain missing evidence |
| Implemented work, deployment dates, verification and subsequent observations | Manual plus approved external/product evidence | Keep the three stages separate; do not infer causation |
| Competitor interpretation, next actions, owners and review dates | Manual | Tie each proposed action to evidence |
| Referrals, leads, sales and paid-ad outcomes | Separate approved analytics/provider source, if available | State source, window and attribution; unavailable is not zero |


# AI visibility report: [Client / site]

Report period: [Start–end and timezone]  
Prepared by: [Owner]  
Report question: [The decision this report supports]  
Evidence packet: [Private approved location]

## Executive summary

[Describe the defined scope, most useful observation, material limitation, completed work and next priority. Keep the summary to one short paragraph. Do not turn an observation into a causal or performance claim.]

## Method and coverage

| Field | Record |
| --- | --- |
| Brand / approved domains / aliases | [Current definitions] |
| Reference prompt set and version | [Exact questions or approved evidence link] |
| Branded versus non-branded groups | [Report separately] |
| Provider, API/model/search context | [Known settings; mark unknowns] |
| Market and language | [Recorded context] |
| Baseline and current observation dates | [Collection dates, not only report month] |
| Planned responses | [Count for each period] |
| Successful responses | [Count for each period] |
| Failed / pending / unavailable | [Counts and exclusions] |
| Comparable subset | [Matching question/provider positions and exclusions] |
| Method changes | [Questions, providers, aliases, scope or counting rule changes] |

## Answer observations

| Metric | Baseline count / denominator | Current count / denominator | Formula and limitation |
| --- | --- | --- | --- |
| Brand appearances | [n/N] | [n/N] | Successful answers naming or recommending target; once per answer |
| Recommendations | [n/N] | [n/N] | Successful answers recommending target under stated classification |
| Answers with an owned explicit citation | [n/N] | [n/N] | At least one citation to an approved owned domain; not a distinct-URL count |
| Distinct explicitly cited URLs | [Count] | [Count] | Preserve URL normalization and scope; no response-rate denominator |

If a denominator is zero or evidence is unavailable, write “Unavailable” and explain why. State percentage-point changes separately from relative percentage changes. Do not assume these report calculations are exact PB UI metrics; reconcile with the methodology reference.

## Question groups

| Group | Successful answers | Appearance count | Recommendation count | Citation-bearing answers | Observation |
| --- | --- | --- | --- | --- | --- |
| [Group] | [N] | [n] | [n] | [n] | [Evidence-backed summary] |

## Competitors and sources

| Brand/source | Prompt and provider | Observation date | Relationship | Reviewed evidence | Next question |
| --- | --- | --- | --- | --- | --- |
| [Name / URL] | [Context] | [Date] | [Mention/recommendation; explicit/returned/searched; owned/third-party] | [Link or approved excerpt] | [What it suggests investigating] |

## Page health

| Relevant URL | Check date | Access/status/content evidence | Unavailable evidence | Required action |
| --- | --- | --- | --- | --- |
| [Canonical URL] | [Date] | [Actual finding] | [Unknowns] | [Owner and action] |

A shareable PB technical check may support this section. It is not an automatic export of the entire report. Inspect public URL content and expiry before sharing.

## Work and later observations

| Work item | Implemented | Verified | Observed afterward | Attribution limit |
| --- | --- | --- | --- | --- |
| [Change] | [Actual change/date] | [Live URL/check/date] | [Later answer evidence/date] | [What cannot be attributed] |

## Business outcomes, if available

[Separate referrals, leads, sales and paid-ad performance from organic answer observations. Name the analytics source, time range and attribution model. Omit unavailable metrics or mark them unavailable; never fill them with invented zeroes.]

## Priorities

| Priority | Action | Evidence | Owner | Acceptance | Review date |
| --- | --- | --- | --- | --- | --- |
| [1] | [Concrete next task] | [Source] | [Owner] | [Verifiable outcome] | [Date] |

## Limitations and next review

[State sample size, coverage, ordinary answer variability, context changes, data freshness and causal limits. Identify the next comparable collection and what decision it should support.]