Skip to content
BENIBIX Learn

Data Transformation

Data⚙️ ProcessingData Transformation

Also known as Transformation, Data Transform

Raw data is changed into the shape a report needs.

Think of it like

It is like making soup from a bag of shopping. You wash, cut and mix until it is ready to eat.

Example

Raw orders hold prices as text and dates in three formats. A transformation fixes both, then adds up sales by day.

Why it matters

Raw data is rarely ready to use. Most of a data team's time goes into this step.

Definition

Data transformation changes raw data into the shape a report or model needs. It can fix types, rename columns, join tables, filter rows and add up totals. It is the T in ETL and ELT. Today it is often written in SQL and run inside the warehouse. Each step should be tested, so a bad change is caught early.

Common mistake

Doing all the work in one huge query. Small steps are easier to test, fix and reuse.

Key takeaways

  1. 1
    Raw in, shaped out. Fix, join, filter and sum.
  2. 2
    The T in ETL. Often SQL in the warehouse.
  3. 3
    Small tested steps. Catch a bad change early.