category discovery
Which providers should a [buyer role] consider for [category] in [market], and what evidence should they compare?
Research Methodology
A research framework before any result is claimed.
Xindar's public measurement specification defines the fields required for repeatable AI visibility research. It contains no approved observations or performance claims.
TEMPLATE STATUS
No approved observation dataset is attached.
This page is a protocol, not a result. It does not claim that Xindar or any other entity improved mentions, recommendations, citations, or commercial outcomes. External platform collection remains unauthorized in this specification.
Required study record.
Every published result must fill these fields before observations are interpreted.
Research questionHow consistently do selected AI answer engines identify, describe, compare, cite, and qualify an entity for a defined category, audience, and market?
SampleOne engine response to one frozen prompt in one declared market-language context.
Prompt panel8 frozen, brand-neutral templates with stable query IDs.
Engines and modelsMust be selected before collection and recorded exactly as shown by the product or API. No engine or model has been approved for this template yet.
Region and languageMust be declared per run. Do not merge markets or languages into one score without a documented aggregation rule.
Collection dateNot set. A real study must publish its exact collection window.
Repeat count3 planned repeats per query; no repeats have been collected under this template.
ScoringDimension-level observations only. No composite score is approved until weighting and missing-value rules are documented.
Source IDsSRC-[collection date YYYYMMDD]-[query ID]-[engine code]-R[repeat number]
Version1.0.0
Brand-neutral query panel.
Placeholders are frozen with the study version; the panel does not contain a target brand or collected answer.
selection criteria
What criteria should a [buyer role] use to evaluate a [category] provider for [use case] in [market]?
provider comparison
Compare credible approaches to [use case] for [buyer role] in [market]. Include limitations and verifiable sources.
evidence verification
What public evidence would verify that a [category] provider can support [use case] in [market]?
risk review
What risks, exclusions, or unsupported claims should a buyer check before selecting a [category] provider?
implementation
What implementation steps and evidence are required for a [buyer role] to adopt [category] for [use case]?
regional fit
How should requirements for [category] differ between [market A] and [market B]?
alternatives
What alternatives exist to [approach] for [use case], and when is each option appropriate?
SCORING BOUNDARY
Observe dimensions before inventing an index.
entity Mention
0 = absent; 1 = present.
description Accuracy
0 = materially wrong or unsupported; 1 = mixed or incomplete; 2 = supported by approved entity facts.
recommendation Position
Record ordinal position only when the answer presents an explicit ordered recommendation. Otherwise record null.
citation Presence
0 = no inspectable source; 1 = at least one inspectable source.
citation Support
0 = source does not support the adjacent claim; 1 = partial support; 2 = direct support.
qualification Coverage
Count declared buyer criteria that the answer addresses with evidence; store the denominator separately.
risk Disclosure
0 = relevant limitation omitted; 1 = limitation mentioned; 2 = limitation linked to decision impact or evidence.
aggregation Boundary
Do not publish a composite score until weighting, missing-value handling, confidence intervals, and minimum sample size are approved.
LIMITATIONS & CITATION
Version the method and preserve uncertainty.
- AI answers can change across time, sessions, products, model versions, account context, and geography.
- A fixed prompt panel measures only the declared questions; it does not represent every buyer query.
- Citation presence does not prove that a source supports a claim; support must be checked separately.
- Entity mention and recommendation are observations, not proof of commercial impact.
- This specification contains no collected responses, scores, uplift claims, or approved research result.
How to cite
Xindar. AI Visibility Query Panel and Measurement Specification, version 1.0.0. Template status: no approved observations. https://www.aixindar.com/research-methodology