This paper documents a systematic infrastructure-driven evidence gap that affects brand performance at the AI purchase recommendation stage. Building on the four-cause diagnostic framework introduced in WP-2026-04 (AIVO Orbit), we apply Cause 2 analysis - evidence invisible rather than evidence missing - to four major enterprise brands across distinct industry verticals: The North Face (consumer retail, VF Corporation), DocuSign (B2B SaaS), Akamai Technologies (cloud infrastructure), and Expedia (online travel). Source code inspection of key product pages across all four brands reveals a consistent pattern of AI content invisibility driven by enterprise web infrastructure rather than content quality or production volume. The brands have produced extensive, accurate, commercially relevant content. That content is systematically invisible to AI model crawlers due to four distinct infrastructure mechanisms: dead link generation from CMS-served assets, absence of JSON-LD structured data on key product pages, JavaScript-rendered page architecture that delivers no content to non-rendering crawlers, and knowledge graph entity anchors that have not been updated to reflect brand repositioning. These findings have material implications for AI purchase recommendation outcomes. A brand whose content is invisible to AI crawlers at the decision stage cannot win the T4 recommendation regardless of content quality, brand equity, or advertising spend. The evidence invisibility problem is a pre-content problem - it must be resolved before content optimisation programmes can produce measurable inference position improvement. We describe the four infrastructure mechanisms, their brand-specific manifestations, their mapping to the four-cause diagnostic, and the systematic monitoring and remediation approach required to address them at enterprise scale.
Tim de Rosen (Mon,) studied this question.