Toast modifier sales report cleanup in Excel
Modifier cleanup is not just deleting blank rows. The goal is a table where each modifier remains traceable to its source and its parent item—or is clearly marked for review.
The problem
Toast modifier data can contain repeated names, zero-priced choices, paid add-ons, special requests, voids, malformed values, and parent IDs that do not exist in the chosen item file. Removing “messy” rows can make totals look neat while destroying the audit trail.
Example Toast CSV columns
business_date,parent_item_selection_id,modifier_group,modifier_name,quantity,net_price,voided2026-06-22,item-204,Cheese,Extra cheese,1,2.00,false2026-06-22,item-999,Add-ons,Avocado,1,2.50,falseKeep these cleanup columns beside the normalized values:
- source filename;
- source row number;
- parent match status;
- parse warning;
- voided status;
- review note.
Cleanup sequence
- Preserve the untouched CSV.
- Normalize whitespace without changing the source field.
- Parse quantity and currency explicitly.
- Keep zero values distinct from blanks and invalid text.
- Check whether each parent ID matches exactly one Item Selection.
- Mark unmatched and ambiguous relationships.
- Keep voided rows visible.
- Build modifier frequency and value summaries from reviewed rows.
Do not deduplicate repeated modifier names automatically. Two “Extra cheese” rows may represent two deliberate selections.
Screenshot and demo

The exception sheet is part of the deliverable, not a failure screen.
Free sample workbook
- Download the modifier cleanup template
- Download sample Item Selection data
- Download sample Modifier Selection data
Use the synthetic item-999 row to test how your process handles an unmatched
parent.
Pricing
Generate up to three successful workbooks per UTC day for free within the free limits. Pro is $19/month or $190/year and adds larger limits, saved mappings, history, email delivery, and 90-day retention.
Frequently asked questions
Should zero-priced modifiers be removed?
No. They may represent included choices or operational preferences. Keep them for frequency analysis.
Can I deduplicate by modifier name and check number?
Not safely. One check can include multiple identical items and modifiers.
Should voided modifiers count?
Keep them visible and decide explicitly which analysis excludes them.
Why preserve invalid raw values?
They show what the source contained and make correction possible without guessing.
Sources
Related: how to combine ItemSelectionDetails and ModifierSelectionDetails.

