# GEO Client Report

Client: [Client]
Reporting period: [Period]
Prepared by: [Agency / consultant]
Benchmark ID: [ID]

## 1. Executive decision summary

Decision this report supports: [Decision]

What happened:

- [Observed finding]
- [Observed finding]
- [Observed finding]

Why it matters to the client: [Business interpretation]

What we recommend next: [Prioritized action]

Confidence and limitation statement: [Coverage, sample, and interpretation limits]

## 2. What we tested

| Scope item | Current period | Previous period | Comparable? |
| --- | --- | --- | --- |
| Collection dates | [Dates] | [Dates] | [Yes / no and why] |
| Frozen prompt set | [ID and count] | [ID and count] | [Yes / no] |
| Neutral / brand-named mix | [Counts] | [Counts] | [Yes / no] |
| Providers and surfaces | [List] | [List] | [Yes / no] |
| Models or visible labels | [List] | [List] | [Yes / no] |
| Market, locale, session state | [Conditions] | [Conditions] | [Yes / no] |
| Eligible coverage | [n/N and %] | [n/N and %] | [Yes / no] |

If the conditions are not comparable, do not present the change as a trend.

## 3. What changed

| Metric | Current | Previous | Change | Client translation |
| --- | ---: | ---: | ---: | --- |
| Neutral discovery visibility | [%] | [%] | [pp] | How often the brand appeared before it was named |
| Prominence | [%] | [%] | [pp] | How often the brand led or made the main shortlist |
| Selected-competitor share of voice | [%] | [%] | [pp] | Brand mentions relative to the named comparison set |
| Claimed-domain citation rate | [%] | [%] | [pp] | How often eligible answers linked to the client domain |
| Representation accuracy | [%] | [%] | [pp] | How often brand facts were materially correct |
| Coverage | [%] | [%] | [pp] | How much of the planned benchmark produced eligible evidence |

## 4. Evidence that explains the result

### Prompt wins

- [Prompt] — [Provider] — [Why the result matters] — [Evidence reference]

### Prompt gaps

- [Prompt] — [Provider] — [Competitor or source shown instead] — [Evidence reference]

### Citation ecosystem

- Recurring independent sources: [Domains / pages]
- Client pages cited: [URLs]
- Competitor pages cited: [URLs]
- Inaccurate or outdated sources: [URLs and issues]

## 5. Interpretation

Observed:

- [Directly supported statement]

Likely explanation:

- [Hypothesis and supporting evidence]

Still unknown:

- [Uncertainty and what would resolve it]

Do not convert a plausible explanation into a causal claim.

## 6. Next-period action plan

| Priority | Action | Evidence | Owner | Completion proof | Validation date |
| ---: | --- | --- | --- | --- | --- |
| 1 | [Action] | [Prompt / answer / source] | [Owner] | [Proof] | [Date] |
| 2 | [Action] | [Prompt / answer / source] | [Owner] | [Proof] | [Date] |
| 3 | [Action] | [Prompt / answer / source] | [Owner] | [Proof] | [Date] |

## 7. Traffic and business outcomes

Report only when instrumentation exists:

- AI referral sessions: [Value and source definition]
- Conversions from identified AI referrals: [Value and attribution limits]
- Assisted or branded-search changes: [Value and why causation is uncertain]
- Revenue associated with tracked referrals: [Value and attribution method]

Keep answer visibility, direct traffic, assisted behavior, and revenue as separate layers.

## 8. Limitations

- This report is a time-stamped directional benchmark, not a guarantee of future answers, citations, rankings, traffic, or revenue.
- The prompt set is not claimed to be a random sample unless the documented design supports that claim.
- Results apply only to the recorded providers, surfaces, models, settings, locale, and collection period.
- API and consumer-product answers may differ.
- [Client-specific limitation]

## Appendix and evidence access

Evidence location: [Owner-controlled location]

Include the frozen prompts, full answers, visible sources, provider metadata, failures, calculations, and change log. Do not expose private client data in a public artifact.