Futureproof Column Selection
Learn why deleting unwanted columns is fragile, and why selecting the columns you need and removing everything else is the safer, more professional approach.
A common mistake when cleaning wide tables is to delete the columns you don't need. It works today — but next month, if the source file arrives with two extra columns, your "Remove Columns" step silently discards them too. And if a column you deleted reappears with a new name, Power Query may throw an error on the step that followed it.
The professional habit is the opposite: select only the columns you want to keep, then choose Remove Other Columns. This step is anchored to the columns you specified — not to the columns you didn't. Future changes to the source file's width have no effect on your query whatsoever.
| # | Steps | Directions |
|---|---|---|
| 1 | Create a new query and source data from Excel File 2B. | |
| 2 | Select the GL transactions sheet. | |
| 3 | Promote headers. | |
| 4 | Sure enough — we've got 35 columns of data. Only 6 are useful. | |
| 5 | It is tempting at this point to select and delete column 7 onwards. However, thinking about automation and futureproofing — what if next time there are 37 columns? | |
| 6 | A better method is to select the columns you want to keep, and remove the others. | |