- First, I would analyze how the historical data structure changed, such as column names, table schema, or calculation logic.
- I would compare the old and new data models to understand which fields can still be aligned for year-over-year comparison.
- If possible, I would create a mapping layer that standardizes the fields into a common structure.
- For example, in one project product categories changed between years, so we created a mapping table to align old categories with the new ones.
- Next, I would transform the historical data to match the current structure in the data preparation layer.
- If exact alignment is not possible, I would clearly define assumptions or adjusted calculations.
- I would validate the results with sample checks and discuss the approach with stakeholders to ensure the comparison is meaningful.
- Finally, I would document the transformation logic so future reporting keeps the same comparison method.
Business users want to compare data across years, but historical data structure has changed. What do you do?
Updated on March 9, 2026
< 1 min read
