Can I build a load file in Excel?

Yes — the production itself is deterministic text. Convert the sheet to CSV, then to DAT/OPT with the correct delimiters. The traps are Excel's own re-encoding, stripped leading zeros on Bates numbers, reformatted dates, and cell line breaks that must become ® markers.

What are the 7 columns of an OPT file?

Image key (Bates/page number), volume, file path, document break (Y), folder break, box break, and page count — the page count appears on the first row of each document. Empty trailing columns are preserved.

Why not just rename my CSV to .dat?

Renaming changes nothing: the delimiters are still commas. A DAT needs the ASCII 20 field delimiter (or visible ¶ variant), the þ qualifier around every field, ® for in-value line breaks, and a header row in the same grammar.

← LoadfileKit / Guides / Excel → DAT/OPT

Excel → DAT / OPT: building a load file from a spreadsheet

For small productions assembled by hand · what the spreadsheet must become before any platform will import it

Not every production comes out of a processing vendor. Small matters get assembled by hand: a spreadsheet of documents, a folder of PDFs, and a receiving party that expects a proper load file. The gap between "CSV I can edit" and "DAT the platform accepts" is small but full of traps — generic tools don't speak Concordance delimiters, which is why the conversion has to be done deliberately.

What a DAT needs that a CSV doesn't have

Requirement	Detail	Spreadsheet trap
Field delimiter	ASCII 20 / DC4 (the invisible control char); some shops use the visible ¶ = 0xB6 variant	Excel exports commas; renaming the file to .dat changes nothing
Text qualifier	þ (0xFE in Windows-1252) wraps every field; a literal þ inside a value is written twice	Cells containing þ look identical to delimiters — they're not
In-value line break	® (0xAE) stands in for CR/LF so one record stays on one physical line	Alt-Enter newlines in cells become literal line breaks and shred the record
Header row	First line: field names, same grammar, same order for every record	Excel headers with commas or special chars must be wrapped too
Encoding	One family for the whole file — Windows-1252 for legacy targets, UTF-8 (ideally with BOM) for modern ones	Excel re-encodes CSV on save and silently changes the family
Line endings	CRLF	Usually fine on Windows; Mac exports can arrive LF-only

The OPT side: seven columns, one row per page

The image cross-reference is positional — seven comma-separated columns, no qualifier:

#	Column	Content
1	Image key	The Bates number / page key of this image
2	Volume	Volume identifier (e.g. VOL001) matching the volume's folder
3	Path	Relative path to the single-page image file
4	Document break	Y on the first page of a document, blank otherwise
5	Folder break	Y when a folder boundary is crossed
6	Box break	Y when a box boundary is crossed
7	Page count	Number of pages in the document, written on the break row

Empty trailing columns are preserved, not trimmed. Every Y row's page count must equal the inclusive span of keys in that document — the checker in the Bates checker verifies exactly that, along with gaps and duplicates across the whole run.

Excel-specific traps, in order of how often they bite

Leading zeros stripped. ABC0000001 survives as text, but numeric-looking Bates numbers get converted and lose padding. Format the column as text before entering anything.
Dates reformatted. Excel rewrites 6/30/2017 per its locale, and review platforms are picky about date formats. Keep date columns as text.
Encoding drift. "CSV UTF-8" vs plain "CSV" in the save dialog differ by a BOM — and BOM presence is exactly what some importers use to detect encoding. Pick deliberately, then verify bytes.
Commas in custodian names. Harmless inside a qualified DAT, fatal in a naive CSV-to-DAT script that splits without honoring the qualifier.
Cell line breaks. Convert Alt-Enter breaks to ® during export; a literal newline in TEXT is the classic "second half of my email imported as its own document" bug.

The workflow

Sheet → CSV (deliberate encoding) → DAT/OPT via the converter → verify (field counts, Bates spans, page counts) → test import on the target platform. If you need to hand-edit afterwards, convert DAT→CSV, edit, convert back — never edit the DAT in a text editor that might re-encode it (see the encoding trap). The DAT format and OPT format guides have the full grammars.