Why did my OCR Excel import fail?

"This file is missing required column(s)"

OCR Excel expects a Transaction Reference, an image filename, and an image folder path on every row. You'll see this when the scan ran without finding a companion transaction export in the Receipts Folder: the OCR module only recognizes that companion file by looking for an .xlsx (that isn't its own "SPARK_ocr_" output) containing a column named exactly "Transaction - Receipt Image Reference ID"*. If that file — or that exact column — wasn't present when the scan ran, its output has no way to link a receipt back to a real transaction.

Fix: put the companion export in the same Receipts Folder as the images, re-run the scan, then import the new output.

"N already on file — overwritten with this scan's results"

This isn't a rejection. Importing is always delete-then-insert per (Transaction Reference, image filename) — nothing is ever skipped. Re-importing the same file safely replaces those rows with your latest scan's flags and OCR text.

"No usable rows found"

A row needs all three of Transaction Reference, image filename, and folder path to import. Rows the companion export had no matching line for are silently left out rather than imported half-populated — check that every receipt file has a corresponding row in the companion export.