Analytics Engineering Best Practices

Implement analytics engineering practices with dbt and modern data stack.

📊 Data & AnalyticsadvancedAnalytics Engineer✓ Free

The Prompt

You are an analytics engineering lead. Create a best practices guide.

Data warehouse: [SNOWFLAKE/BIGQUERY/REDSHIFT]
Transformation tool: [DBT/OTHER]
Team: [SIZE]
Current state: [DESCRIBE]

1. dbt Project Structure:
   - Folder organization: staging, intermediate, marts, metrics
   - Naming conventions: sources, models, columns
   - Materialization strategy: view, table, incremental, ephemeral
   - Configuration: project.yml, source freshness, tags

2. Modeling Patterns:
   - Staging: 1:1 with source, renaming, casting, basic cleaning
   - Intermediate: business logic, joins, aggregations
   - Marts: dimensional models (facts and dimensions), one big table, metric definitions
   - Reverse ETL: models for operational tools

3. Testing:
   - Schema tests: unique, not null, accepted values, relationships
   - Data tests: custom SQL tests for business rules
   - Unit tests: testing complex logic
   - Freshness: source freshness SLAs

4. Documentation:
   - Model descriptions and column descriptions
   - dbt docs: generation, hosting, adoption
   - Lineage: understanding and communicating data flow

5. Code Quality: SQL style guide, PR review checklist, CI/CD pipeline
6. Performance: query optimization, incremental models, clustering/partitioning
7. Governance: access control, PII tagging, data classification
8. Metrics Layer: metric definitions, semantic layer, consistency

💡 Tip: Replace all [bracketed text] with your specific details before pasting into your AI model.