STUDY PROTOCOL / V1.0
AEO Benchmark Study Protocol
The CBI Study Protocol defines the required disclosures for category scope, prompt design, model surfaces, collection conditions, coding, exclusions, limitations and corrections.
STUDY PROTOCOL / V1.0
Required Disclosure Fields
30protocol fields
6journey stages
3evidence classes
1public correction path
STUDY PROTOCOL / V1.0
Release Gates
01
Scope frozen
Category, entities, buyers, stages and questions are documented.
02
Conditions recorded
Surface, model label, geography, sessions, dates and repetitions are preserved.
03
Codebook locked
Presence, recommendation, caveat and citation rules precede analysis.
04
Evidence retained
Answers, sources, timestamps and exclusions can be audited.
05
Limits written
The report states what cannot be concluded.
06
Corrections open
Documented factual corrections follow a public process.
STUDY PROTOCOL / V1.0
Why The Protocol Is Proprietary
The moat is not a secret score. It is the disciplined connection between commercial buyer questions, repeatable observations, claim-level source evidence and decision-ready reporting. Publishing the protocol makes future findings more credible and harder to imitate superficially.
STUDY PROTOCOL / V1.0
Complete 30-Field Study Protocol
Every required study field and its definition is published below. These are the disclosure requirements applied before an observed AI visibility benchmark can be released.
30 records published online. Browse The Complete Resource
Required
research question
Definition: The decision the study is designed to inform
Evidence Class: Methodological
Version: v1.0
Required
evidence class
Definition: Observed, Curated or Methodological
Evidence Class: Methodological
Version: v1.0
Required
category definition
Definition: Operational category scope and exclusions
Evidence Class: Methodological
Version: v1.0
Required
entity set
Definition: Frozen list of entities and inclusion reasons
Evidence Class: Methodological
Version: v1.0
Required
buyer audiences
Definition: Roles represented in the prompt corpus
Evidence Class: Methodological
Version: v1.0
Required
journey stages
Definition: Discovery through procurement stages included
Evidence Class: Methodological
Version: v1.0
Required
prompt corpus version
Definition: Immutable corpus version identifier
Evidence Class: Methodological
Version: v1.0
Required
surface
Definition: Named AI product or search surface
Evidence Class: Methodological
Version: v1.0
When visible
model label
Definition: Model or experience label displayed by the surface
Evidence Class: Methodological
Version: v1.0
Required
geography
Definition: Researcher and market geography
Evidence Class: Methodological
Version: v1.0
Required
language
Definition: Language of prompts and answers
Evidence Class: Methodological
Version: v1.0
Required
session condition
Definition: Signed-in state, memory and personalization controls
Evidence Class: Methodological
Version: v1.0
Required
collection start
Definition: First observation timestamp
Evidence Class: Methodological
Version: v1.0
Required
collection end
Definition: Last observation timestamp
Evidence Class: Methodological
Version: v1.0
Required
repetitions
Definition: Number of controlled repetitions per prompt
Evidence Class: Methodological
Version: v1.0
Required
answer capture
Definition: Preserved answer text or approved representation
Evidence Class: Methodological
Version: v1.0
Required
citation capture
Definition: Visible URL, title and access timestamp
Evidence Class: Methodological
Version: v1.0
Required
coding codebook
Definition: Predeclared presence, recommendation and caveat rules
Evidence Class: Methodological
Version: v1.0
Required
reviewer count
Definition: Number of coders or adjudicators
Evidence Class: Methodological
Version: v1.0
If multiple reviewers
agreement method
Definition: Inter-reviewer agreement method
Evidence Class: Methodological
Version: v1.0
Required
exclusions
Definition: Excluded observations with reasons
Evidence Class: Methodological
Version: v1.0
Required
unknown handling
Definition: Rules for ambiguous or missing evidence
Evidence Class: Methodological
Version: v1.0
Required
canonicalization
Definition: URL and entity normalization rules
Evidence Class: Methodological
Version: v1.0
If composite used
weighting
Definition: Predeclared metric weights
Evidence Class: Methodological
Version: v1.0
Required
limitations
Definition: Sampling and interpretation limits
Evidence Class: Methodological
Version: v1.0
Required
correction policy
Definition: How documented corrections are reviewed
Evidence Class: Methodological
Version: v1.0
Required
publication version
Definition: Release identifier
Evidence Class: Methodological
Version: v1.0
Required
data dictionary
Definition: Field definitions and allowed values
Evidence Class: Methodological
Version: v1.0
Required
source license
Definition: Permitted use and attribution terms
Evidence Class: Methodological
Version: v1.0
Required
contact
Definition: Research correction contact channel
Evidence Class: Methodological
Version: v1.0
Next Useful Step
Request An AI Visibility Diagnostic
Start with the commercial decision, evidence and measurement conditions that matter most. The first recommendation will define the most useful scope before any larger commitment.