Merged/blank-cell CSV (flattened formatted workbook)
A CSV produced by exporting a human-formatted workbook that relied on merged cells and multi-row headers. People assume a CSV is a clean rectangular grid; this one is the visual layout with all the merges silently dropped.
The shape, how the file is really laid out
CSV has no concept of a merged cell. When a merge is flattened, the value
survives only in the top-left cell of the merged range and every other cell
it visually covered becomes empty. A banner like "Q1 2026, Northeast Region"
that spanned six columns becomes one populated cell followed by five commas.
A category label merged vertically down a block of rows appears once with
blanks below it. A two-row header (group on row 1, sub-field on row 2)
exports as two physical header rows, so a single column's identity is split
across "Revenue" above and "Net" below. Formatting-only spacer columns and
title rows above the real header produce leading blank columns and ragged
rows with fewer commas than the data rows.
The traps
Blank does not mean zero and does not mean null, under a merged label it means "same as the cell above," which only a forward-fill recovers, and the forward-fill is ambiguous wherever a genuinely empty cell sits inside a merged block.
- Every trap in this export, worked through
- The reconciliation anchor, the total the file asserts about itself
- The refusal cases, when to stop and ask instead of guess
Reference: Microsoft Support, "Merge and unmerge cells" · RFC 4180 (Common Format for CSV Files) · Google Docs Editors Help, "Download & export files"