1. Overloaded Expressions
- Large, nested expressions used in place of reusable logic.
- Business-critical rules hidden in obscure calculated fields.
2. Misuse of Lookups & Joins
- Using lookups instead of pushdown joins.
- Repeating identical lookups across multiple steps.
- Handling join conditions inconsistently across pipelines.
3. Implementing Business Processes in ETL
- Building approvals, email flows, or business workflows inside ETL tools.
- Mixing data processing responsibilities with orchestration or business logic.
4. Excessive Copy-Paste Coding
- Many pipelines copied with slight modifications instead of being parameterized.
- Inconsistency increases maintenance effort drastically.
5. Using ETL as an Application Server
- Embedding procedural logic better executed in an API or backend service.
- Performing file parsing or enrichment workflows better suited to cloud-native tools.
Tool abuse inflates cost and complexity. Automated platforms like Undraleu can detect these patterns at scale, helping teams refactor consistently.