The two most common CSV combination operations — appending and joining — are frequently confused because people use the word "merge" to mean both. The confusion produces wrong results: you end up with one wide file when you wanted one long file, or vice versa. Here's a clear breakdown of what each operation does and when to use it.
Appending: Making One Long File
Appending (also called stacking or concatenating) takes multiple files with the same columns and combines their rows into one file. The output is taller than any individual input but no wider.
When to append:
- Monthly report exports that should be one annual file
- Regional data files that should be combined for a national analysis
- Survey response batches from different collection periods
- Transaction records split across multiple export files
The requirement for appending is structural consistency: all files should have the same column headers in the same order. When that's true, appending is trivial. When columns are different across files (even slightly — an extra column in one file, a renamed column in another), you need to decide how to handle the mismatch.
Output: If you have three files with 1,000 rows each, you get one file with 3,000 rows and the same number of columns.
Upload your CSV and see exactly what's wrong — Sohovi profiles quality in seconds — try Sohovi free.
Joining: Making One Wide File
Joining (also called merging on a key) takes two or more files that share a common identifier column and combines their other columns side by side. The output is wider than any individual input but not necessarily taller.
When to join:
- A customer list with IDs, names, and contact info joined to an orders file with IDs and purchase history
- A product catalog joined to a pricing file, where both share a product SKU column
- Employee records joined to performance data, both keyed on employee ID
The requirement for joining is a shared key column. Every row in one file must be matchable to a row in the other file via this shared key. Rows that don't match may be dropped (inner join) or kept with nulls (outer join).
Output: If file A has 1,000 rows and 5 columns, and file B has 1,000 rows and 4 columns (one of which is the shared key), you get one file with 1,000 rows and 8 columns (5 + 4 - 1 shared key column counted once).
Common Mistake: Trying to Join When You Should Append
The most frequent error is uploading two monthly sales files and selecting "join" instead of "append." The result is either an error (if join mode can't find a common key) or a meaningless wide file where January and February data appears side by side per row.
If your files have the same structure and represent the same type of data over different time periods, you want to append.
Sohovi's free CSV Merger supports both stack (append) and join operations, with a schema union option for appending files with slightly different columns.
Keep Reading
Merge CSV Files Free →