Methodology

Classification model

Organizations are classified independently across executive roles, provider types, community formats, event formats, intelligence types, resources, executive needs, topics, audiences, and geographies. Each record exposes one primary provider category. Additional provider categories describe other material lines of activity; research and publishing capabilities remain in the separate intelligence-type dimension.

Controlled vocabulary

Every semantic term has a stable ID and neutral definition. Closely related phrases resolve through the alias registry to one canonical concept unless they represent a material functional distinction. Community formats describe ongoing structures; event formats describe time-bounded gatherings; first-party research remains distinct from editorial commentary.

Review the controlled taxonomy, aliases and role mappings.

Evidence requirement

Entity associations require reasonable evidence from official organization, event, research, regulatory, government, or reputable third-party sources. A passing keyword mention is insufficient. Semantic relationship exports retain the evidence URLs used for each classified field.

Profile completeness

Each organization receives a transparent 0–100 record-completeness score across thirteen weighted criteria. Values such as founding year, headquarters, operating geography, offerings, industries, classification and executive relevance only receive credit when the record includes applicable evidence. Source breadth, fact depth and freshness are measured separately. The score describes research depth inside the Index; it is not a rating, ranking, endorsement or assessment of organizational quality.

Review the public standard, field coverage and research queue.

Multi-category organizations

An organization may serve multiple roles or appear in multiple dimensions without creating duplicate entity records. Its primary category represents its principal organizational activity. Capabilities such as events, research, peer groups, and resources do not displace that identity unless they are independently documented as a principal line of business. Missing associations remain unknown and are not inferred from a company name or isolated article mention.

Indexability and intelligence publication gates

Arbitrary filter combinations remain application state. A canonical page is created only when it has stable intent, enough substantive entities, distinct analysis, and durable value. Decision tools and benchmarks require at least 60% evidence coverage for headline placement, answer-engine structured data, and sitemap inclusion. Lower-coverage analyses remain downloadable and visible as research coverage, use noindex,follow, and cannot present themselves as decision-ready answers. Dataset v0.9.7 does not create role-format-topic combination pages.

Commercial neutrality

Inclusion does not constitute endorsement. Commercial relationships do not influence inclusion, ranking, scoring, descriptions, or category placement.

Versioning and updates

Schema and dataset versions are published in exports. Stable IDs do not change without a documented migration. Taxonomy additions and aliases are recorded in the taxonomy changelog; breaking entity-schema changes require a schema-version update.