How to clean sales data before you analyse it
Bad data gives confident but wrong answers. A ten-minute check before you analyse prevents most mistakes.
Duplicates and blanks
Duplicate rows count the same sale twice. Blank cells in the amount column hide sales from totals. Count both before you trust any total.
Dates and numbers
Dates should be in one style. Amounts should be numbers, not text such as “1,200 rs”. Spelling variations like “Imphal” and “imphal ” (with a space) look like two different cities to a computer, so make names consistent.
Keep a copy
Always keep the original file. Clean a copy, so you can start again if you remove the wrong thing.
In Super B.A.: The Data Setup page shows rows, duplicate rows and blank columns, and the Mini-Projects lesson in Learn & Lab walks through it. Open the app
Related guides
- How to save an Excel sheet as a CSV file
- Average, median and why one number can mislead
- SQL for beginners: SELECT, SUM and GROUP BY
General information for learning. It is not financial, tax, legal or investment advice.