- Filtering during extraction means retrieving only required data from the source itself.
- For example, SQL query pulls only last 12 months of sales instead of full table.
- This reduces data volume, network load, and refresh time.
- Filtering during transformation means applying conditions in Power Query after data is loaded.
- For example, keeping only “Completed” orders from the full dataset.
- It does not reduce source extraction but shapes data for analysis.
- Extraction filtering improves performance; transformation filtering improves report relevance.
- Both are used together for efficiency and clean datasets.
What is filtering during transformation vs extraction?
Updated on February 9, 2026
< 1 min read
