Merge CSV files

Matching columns by name is the part that matters

Most merge tools stack files by position: the third column of every file becomes the third column of the result. That works right up until one export puts Email before Phone and another puts it after, at which point you have silently interleaved two fields and nobody notices until a mail merge goes out wrong.

With match by header name on, the first row of each file is read as its header, and a column called Email lands in the Email column wherever it appeared. Files that contain a column the others do not simply widen the result, and rows from files without that column are left blank there. Header matching ignores case and surrounding spaces.

Turn it off if your files genuinely have no headers, and columns will be stacked by position instead.

Keeping track of where each row came from

Turn on Add a "Source file" column and every row carries the filename it arrived in. When twelve monthly exports become one table, that column is what lets you find the month a bad row came from, and it is what makes a pivot by month possible without parsing dates.

What happens to the headers, blank rows and encodings

• The header row of the first file becomes the header of the result; the header rows of the others are consumed for matching, not repeated in the data.
• Delimiters are detected per file, so a comma-separated export and a semicolon-separated one merge without you converting either.
• Excel files contribute one entry per sheet, so a workbook with three sheets adds three sets of rows, each listed separately so you can remove the ones you do not want.
• Files are read as UTF-8.

When merging is a monthly chore rather than a one-off

Dropping twelve files into a browser tab once is fine. Doing it every month is a process, and a process wants a tool that remembers.

TableDI is a desktop spreadsheet workspace that runs entirely on your machine. It imports a folder's worth of exports into one table, keeps the formula columns you built on top of them, and can pull from the source directly — HTTP, a database over JDBC, or a mailbox — so next month's rows arrive without a download step. How importing works.