How to use the Join CSV Files by Column
- 1 Drop the main file first, then the file to take columns from.
- 2 In the join step choose the other file, then the key column in each file.
- 3 Choose which rows to keep: all rows of the source (left), only matches (inner), or more.
- 4 Check the match counts and download.
Adding columns, not rows
Merging stacks files on top of each other. A join adds columns: for every row in the first file it looks up the row with the same key in the second and appends that row’s other columns – orders with customer names, products with prices. It is what a VLOOKUP does, but it works on whole files, on several key columns at once, and it does not stop at the first match.
Choose the join
- Left: every row of the source; empty cells where nothing matched. The usual choice for a lookup.
- Inner: only rows with a match in both files.
- Right and full: also keep the unmatched rows of the other file (or both files).
Keys that don’t quite match
Matching ignores case and surrounding spaces by default. Identifier columns often lose their zeros in one system and not the other (01234 against 1234); a switch treats those as equal. Rows with an empty key never match anything – the studio reports how many.
Repeated keys
If a key occurs several times in the second file, the source row is repeated once for each match, as in a database join. The studio tells you how many rows did this, because it is the usual cause of a result with more rows than expected. Columns that exist in both files are kept, with _2 added to the second.
Limits
Both files need a header row. To find rows that differ rather than combine them, use compare CSV files.
Frequently asked questions
What is the difference between merge and join?
Merge puts rows from several files into one list; join adds the columns of one file to another by matching a key.
Can I join on more than one column?
Yes. Tick the same number of key columns in both files; they are matched pairwise.
Why does the result have more rows than my file?
A key occurs more than once in the other file. The notes under the step count these rows.
Last updated: