My Mask Re-Run Made No Changes: Why Updates Did Not Land

A mask re-run that reports everything reused and nothing changed is usually correct behavior, not a failure: the import reconciles each row against its existing record and only writes fields that genuinely differ, so an identical file changes nothing. EDGEBIC by User Solutions imports through configurable masks that match, compare, and update rather than blindly overwrite, which is exactly what makes a re-run safe and also what makes an expected update quietly not land.

This post sits in the EDGEBIC troubleshooting guide and is the reconcile-behavior companion to how to build an import mask and import masks explained.

What You Are Seeing

You edited a file, re-imported it, and the result summary shows a stack of reused rows with zero (or far fewer than expected) updates. The data in the app still reads the old values. Nothing failed, nothing aborted, and the log does not name a problem row. The run did work; it just decided there was nothing to write. That decision is the thing to understand.

Why It Happens

An import is a reconcile. For each incoming row it finds the matching record, compares field by field, and writes only what differs. Three situations produce "no change" honestly:

• The file matches the database. If the values you re-imported equal what is already stored, every row reconciles as reused. This is the safe, expected outcome of re-running an unchanged file.
• A blank cell on an update is preserved. An empty cell is treated as "no value supplied," so it keeps the current database value rather than overwriting it with nothing. A file with your changes in some columns and blanks in others updates only the filled columns.
• The column is not mapped to the field you think. If the mask maps your edited column to a different field, the run compares the field it actually maps (unchanged) and reports no update, even though the file looks right.

A fourth, subtler case: units cancel the difference. If the mask applies a conversion factor and both the old and new cells convert to the same stored value, the comparison sees no change.

How to Fix It

Work from the outside in: confirm the file, then the mapping, then the values.

1. Confirm you re-imported the edited file. Re-open the saved file and verify it holds the new values on the sheet the mask reads. Re-importing an earlier copy is the most common cause of "nothing changed."
2. Check the column-to-field mapping. Open the mask and confirm the edited column maps to the field you intend to update. A column mapped one position off, or to a similarly named field, compares the wrong thing.
3. Verify the values genuinely differ. Pick one record, read its current value in the app, and compare against the file cell after any conversion factor. If they are equal, there is nothing to update.
4. To clear a field, supply an explicit value. A blank cell is preserved. Put a zero (for a number) or the intended empty value in the cell so the mask writes it.
5. Re-run and read the summary. After correcting the mapping or the file, the summary should now show updates. A one-record spot check confirms the change landed.

How to Prevent It

• Edit and re-import the same file object. Save the edited workbook and re-open it before importing, so you never load a stale copy.
• Treat blank as "keep," not "clear." Design your update files knowing that empty cells preserve existing values. If a field must be cleared, carry an explicit value for it.
• Name the mask's mapping intent. A mask whose field mappings are documented (which column feeds which field) makes an off-by-one mapping visible before it costs a silent run.
• Spot-check one record after every update load. Reading a single before-and-after value turns "the run did nothing" into an immediate, obvious diagnosis.
• Watch for units that cancel. If a column carries a conversion factor, remember the stored value is scaled; compare against the scaled result, not the raw cell.

If instead the run reported failed rows, that is a different symptom covered in the import said failed rows. If updates landed but created extra records, see an import created duplicate records.