Definition: The degree to which a machine system can form a correct, usable understanding of business information from the representation available to it.
Example: A product page may clearly show a sunscreen to a human, while a machine may still fail to resolve its product type, SPF, intended use, active ingredients, or restrictions.
Boundary: In IEO, interpretability means machines interpreting the merchant. This is the reverse direction of model interpretability in XAI, where humans interpret the model.
Definition: The practice of making a business's offers, entities, attributes, policies, and source facts machine-interpretable before downstream retrieval, recommendation, comparison, or action occurs.
Operational formulation: IEO structures, resolves, enriches, validates, and represents source business data so machine systems can correctly understand the offer without inventing missing facts.
Category status: Open discipline. IEO does not prescribe a specific vendor, model, data format, protocol, or software implementation.
Canonical framing: IEO optimizes understanding.
Canonical closing line: Semantic discovery is downstream of interpretation.
Scope of the acronym: "IEO" is used here for Interpretability Engine Optimization, the discipline proposed by NSOLVIA Research in 2026 and published under DOI 10.5281/zenodo.22104280. The acronym has other expansions in circulation; this work refers exclusively to the one defined here.
Research note: What the term refers to, where it sits relative to neighbouring practices, and what falls outside its scope are set out in Interpretability Engine Optimization: Terminology and Scope.
Definition: The application of IEO to products, catalogs, brands, offers, and commerce operations.
Scope: Commerce IEO is the domain in which the current NSOLVIA research, instruments, and empirical evidence are most mature.
Definition: A collective term for any machine system that consumes business information and must form a usable interpretation of it, including search engines, answer engines, generative systems, recommendation systems, commerce platforms, and autonomous agents.
Clarification: IEO optimizes source data for interpretation engines. It does not optimize the engines themselves.
Definition: The umbrella condition describing whether machines can correctly understand the three objects required to evaluate a commerce offer: the product, the brand, and the transaction.
Canonical model: Product tells the machine what is being sold. Brand tells it who stands behind it. Transaction tells it under what rules it can be bought.
Definition: The state of preparedness of a merchant's commerce data and surfaces for machine-mediated discovery, interpretation, comparison, recommendation, and action.
Clarification: Readiness is a state. IEO is a practice used to improve one of the foundational properties that readiness depends on.
Boundary: Distinct from any vendor-specific readiness score: commercial instruments (e.g., NSOLVIA's Agentic Catalog Readiness Score™) measure readiness including interpretability-related signals, but no single commercial score should be read as a universal IEO metric.
Definition: The accumulated gap between what a business knows about its offers and what its machine-facing representations make explicit enough for machines to understand correctly.
Example: Years of adding products for human shoppers while leaving taxonomy, attributes, use cases, policies, and structured meaning unresolved can create interpretability debt.
Commercial shorthand: A catalog can accumulate interpretability debt just as a software system accumulates technical debt.