How many distinguishable brand positions can a market sustain? We formalize brand perception as points in \ (R⁸_+\) under the Aitchison metric (Aitchison, 1986; Egozcue et al. , 2003), which is isometric to Euclidean distance on a 7-dimensional hyperplane after centered log-ratio transformation. Two brands are distinguishable if their distance exceeds perceptual threshold \ (\). The problem maps onto sphere packing. Because the \ (E₈\) lattice achieves the unique optimal packing in eight dimensions (Viazovska, 2017), its density (\ (⁴/384. 2537\) ) and kissing number (240) supply structural bounds. We derive five results: (1) volume-ratio capacity is at least \ ( (1/) ⁸\), equaling \ (10⁸\) positions at \ (=. 10\) ; (2) each position has at most 240 nearest neighbors decomposing into 112 specialist (two-dimensional) and 128 generalist (eight-dimensional) vectors; (3) 10, 000 brands occupy less than. 01% of the unit 8-ball; (4) category saturation occurs near \ ( (1/) ^dₑff\) ; (5) average inter-dimension correlation \ (=. 3\) collapses effective dimensionality to approximately 2. 6 and capacity by five orders of magnitude. An LLM stability experiment (250 calls, five models) finds null competitive-interference effects, consistent with fixed geometry. The \ (E₈\) connection is structural, not literal: it establishes the mathematical ceiling on positioning capacity. We discuss implications for white-space strategy, correlation management, and observer-dependent capacity. Includes zharnikov-2026g-r4. yaml (Paper Spec v0. 1. 0) — a machine-readable specification of the paper's claims, assumptions, and dependencies. The paper's full machine-first bundle (the SPINE claim/dependency graph and the ONTOLOGY term module) lives in the public repository; see https: //github. com/spectralbranding/paper-spec for the standard. This PDF is generated programmatically from that machine-first source under a research-as-repository model.
Dmitry Zharnikov (Sat,) studied this question.