cff-version: 1.2.0
message: "If you use this white paper, please cite it as below."
type: report
title: "AI Visibility Retrieval Dynamics: Tiered Citation Platforms, Decay, and the Governance Gap in LLM Discoverability"
version: "v1.0"
doi: 10.5281/zenodo.17117353
date-released: 2025-09-14
url: https://doi.org/10.5281/zenodo.17117353
authors:
  - family-names: Sheals
    given-names: Paul
    orcid: https://orcid.org/0009-0006-2407-4612
    affiliation: The AIVO Standard™
  - family-names: de Rosen
    given-names: Tim
    affiliation: The AIVO Standard™
repository-artifact: https://github.com/pjsheals/aivo-retrieval-dynamics
keywords:
  - AI Visibility
  - AIVO Standard
  - PSOS
  - Prompt-Space Occupancy Score
  - Tier 1 citations
  - Wikidata
  - schema.org
  - DOIs
  - LLM
  - AI assistants
  - governance
  - audit
  - volatility
  - decay
  - GEO
  - AEO
license: CC-BY-4.0
abstract: >
  This white paper presents findings from The AIVO Standard’s R&D program based on 100,000+ reverse-engineered prompts across industries and LLMs (ChatGPT, Gemini, Claude, Perplexity, Grok). It formalizes a tiered citation framework—Tier 1 (foundational/canonical), Tier 2 (industry validation), Tier 3 (topical/recency)—and shows that brands relying primarily on Tier 2/3 experience 40–60% month-on-month visibility volatility in AI assistants. The paper identifies a governance gap in current AI visibility tools (which report where brands appear now but do not measure durability, decay, risk, or causality) and introduces PSOS™ (Prompt-Space Occupancy Score) as a governance-grade KPI that weights visibility across tiers, adjusts for decay, and enables board-level reporting, benchmarking, and certification.