GEOANALYTICA.COM — PROMPT TESTING KIT
Purpose: preserve an inspectable record of a bounded AI-answer observation.

SCOPE
Audience:
Topic:
Market / language:
Branded or unbranded:
Intended page:
Question or decision this observation supports:

OBSERVATION
Observation ID:
Timestamp and timezone:
Assistant / product:
Mode / search setting:
Exact prompt:
Relevant context (no sensitive information):
Valid response, failed request, or unusable result:
Saved response or permitted reference:

REVIEW
Brand mentioned: yes / no / ambiguous
Raw citation URLs:
Normalized reporting URLs (preserve the originals):
Claim being checked:
Supporting passage:
Accuracy: supported / unsupported / ambiguous / not reviewed
Reviewer:
Review note:

ACTION
Confirmed finding or hypothesis:
Affected page:
Proposed next step:
Owner:
Acceptance check:
Review date:

EXAMPLE PROMPT FAMILIES — replace bracketed terms before using
Definition: What is [concept], and when is it useful?
Evaluation: What criteria matter when choosing [category] for [use case]?
Comparison: How do [approach A] and [approach B] differ for [task]?
Requirements: What prerequisites apply before using [process]?
Implementation: How should a team begin [task] under [conditions]?
Troubleshooting: What should someone check when [symptom] occurs?
Branded verification: Does [brand / product] support [capability]?
Limitations: What limitations should users understand about [process]?

COUNTING CONVENTION FOR A SIMPLE PILOT
Count a valid answer once when it has at least one qualifying citation to the target domain.
Citation rate = qualifying valid answers / all valid answers in the stated sample.
List failures separately. Separate branded from unbranded questions.
Do not present a monitored sample as all platform activity.
Do not infer visits, revenue, or causation from a citation alone.

WORKFLOW
Define scope. Keep the initial prompt set stable. Record the conditions.
Preserve the response where permitted. Inspect cited pages. Review important claims.
Repeat observations under the recorded method. Annotate methodology changes.
Assign a concrete action only when the evidence supports it.
Never place confidential information, passwords, or private customer details in a test prompt.