Glossary/Semantic Layer & Metrics

Business Logic Layer

A business logic layer is the component of a semantic layer or data system that encodes business rules, calculations, and transformations, making them reusable and enforced across analytics.

The business logic layer captures the "how" of business metrics: subscription revenue excludes trial periods, customer segmentation includes 60-day activity windows, currency conversion happens at transaction date, product categories follow a specific hierarchy. Rather than embedding these rules in individual SQL queries or BI tool logic, the business logic layer centralizes them as governed, versioned code or configuration.

The business logic layer emerged because business rules were scattered across SQL files, dbt models, Tableau calculated fields, and spreadsheet formulas, each subtly different. This fragmentation caused reconciliation nightmares when metrics didn't match. A centralized business logic layer ensures that revenue is calculated the same way whether the query runs in a dashboard, API endpoint, or batch job.

The business logic layer typically lives in dbt (via dbt models and tests), in the data warehouse (computed columns, views), in metric platforms (metric definitions), or in dedicated semantic platforms. It includes not just calculations but also validation: flagging when data quality issues exist, enforcing constraints, and documenting assumptions. The business logic layer is where organizational knowledge becomes code.

Key Characteristics

  • Centralizes calculations and transformation rules
  • Version-controlled and auditable
  • Reusable across multiple analytics interfaces
  • Includes validation and quality checks
  • Decouples business rules from tool-specific implementations
  • Linked to data lineage and ownership

Why It Matters

  • Reliability: Consistent business rule application prevents reconciliation errors
  • Maintainability: Rule changes happen once, affecting all downstream consumers
  • Compliance: Audit trail shows what rules apply to which metrics
  • Agility: Update business logic without redeploying dashboard code
  • Documentation: Rules are explicit, not hidden in query logic

Example

A business logic layer defines customer lifetime value as: sum of orders minus refunds, where orders must be non-test and customer must be non-deleted. This rule exists in one place. When an analyst needs CLV for reporting, they reference this logic; when the rule changes (add minimum account age), it updates once for all consumers.

Coginiti Perspective

Coginiti separates business logic into two layers: SMDL for analytical definitions and CoginitiScript for transformation logic. SMDL encodes how metrics are calculated (aggregation types, custom expressions) and how entities relate, while CoginitiScript blocks encode data preparation rules as reusable, parameterized SQL. Both are stored as code in the Analytics Catalog with version control and promotion workflows, so business logic changes go through review before reaching production.

See Semantic Intelligence in Action

Coginiti operationalizes business meaning across your entire data estate.