AI source & retrieval dependency audit

What does AI seem to depend on when it decides whether to find and recommend you?

The question this service answers

What observable sources, search systems, citations, and third-party representations appear to influence whether AI systems find and recommend this business?

A one-time investigation that maps the observable discovery environment surrounding your business, and shows where your visibility appears concentrated.

Map My AI Discovery Sources Talk With Us

When this is useful

Is this the right question for you?

AI Source & Retrieval Dependency Audit is a focused service. It answers one defined question rather than bundling in a broader program.

You are recommended sometimes and invisible other times, and you do not know what the difference is.
You want to understand what your visibility currently rests on before you invest in changing it.
A third-party page or platform appears to carry more weight than your own site.
You are planning a site rebuild, a migration or a rebrand and want to know what is load bearing.
You want an evidence map rather than an opinion about what matters.

Scope

What Goldeneye maps

Only what can be observed. Where evidence is insufficient, the dependency is recorded as unknown rather than guessed.

Owned pages and published business information.
Search and index presence for the questions that matter.
Observable AI citations captured in real answers.
Local and business data sources.
Publishers and directories.
Review, travel and industry platforms.
Partner data where it is observable.
Repeated third-party sources that appear across multiple answers.
Unknown dependencies, where evidence is insufficient to conclude anything.

What is assessed

Crawl and index eligibility for the pages that should be discoverable.
Recurring cited sources across captured answers.
Source concentration, where a large share of visibility rests on very few sources.
Stale or conflicting information across the observed set.
Missing sources a business of this type would normally be represented in.
Third-party dependency, including sources you do not control.

First-party platform evidence

Where appropriate, first-party platform evidence such as Bing AI Performance may be used for Microsoft-supported surfaces. It is treated as observed evidence for those surfaces only and is never generalized to other providers.

How findings are evaluated

Evidence first, then interpretation.

Every dependency is graded on evidence, not assumption

Observed dependencies are those captured directly in results, citations or platform reporting.
Supported inference is used where a pattern is consistent enough to draw a reasonable conclusion, and it is labelled as an inference.
Unknown is used where evidence does not establish the relationship, and it stays unknown in the report.
Concentration and fragility are assessed against how much of the observed visibility rests on how few sources.

The Goldeneye evidence standard

Observed. Something Goldeneye directly saw in a captured result, source or platform report.
Supported inference. A reasonable conclusion the observed evidence supports, stated as an inference rather than a fact.
Unknown. Something the available evidence cannot establish. Goldeneye reports it as unknown rather than filling the gap.

What you receive

What lands in your hands.

A map of the observable sources, systems and third-party representations surrounding your discovery.
Where your visibility appears concentrated, and where it looks fragile.
Stale, conflicting or missing sources, with the evidence behind each.
Dependencies that could not be established, stated as unknown.
A prioritized view of what deserves correction, monitoring, investigation or no action.

Important boundaries

A visible citation is evidence of a shown source, not proof of its exact causal weight.
Goldeneye does not have access to private model internals, hidden ranking weights or provider retrieval systems. This audit works from observable evidence only.
This is an investigation. Implementation and remediation are separate and are not included.
Findings describe the environment at the time of the audit. Retesting is appropriate after material source changes.

Pricing

What it costs.

AI Source & Retrieval Dependency Audit: $950 one time
Cadence: One time
Baseline required: None

Add a 45-minute Strategy Mode findings review for $225 when you want help deciding what to ignore, watch, investigate or address next. Strategy Mode is an optional add-on to a completed service, never a standalone product.

Map what your AI visibility rests on.

Tell us what you are seeing and we will confirm scope before the audit begins.