CSV / duplicate decision guide

Decide which CSV duplicate to keep before deletion

A shared ID is not enough to justify deletion or merging. Separate exact duplicates from conflicting candidates, then document keep, merge, remove or owner-check before resubmission.

Import a table into Excel

Separate four decisions

Decision	Condition	Next action
Keep	Unique row or confirmed source of truth	Retain it
Merge	All fields are an exact duplicate	Record the basis, then keep one
Remove	Owner has confirmed removal	Keep the confirmation with the handoff
Owner check	Values conflict or a key is missing	Stop instead of guessing

If the same ID has a different email or amount, it is not an exact duplicate.

Reproduce the decision with fixtures

A001 is an exact match, A002 has a conflicting email, and A003 is unique. missing-key.csv is the negative case.

id,decision,reason
A001,keep-one,all fields match; retain one copy after confirmation
A002,ask-owner,email differs; do not merge or remove automatically
A003,keep,only one row; no duplicate action

This page never performs deletion automatically. Create the decision record first, then hand it to the resubmission workflow.

Verify and hand off

Keep the source CSV unchanged and inspect key, name, email and amount.
Use Table to Excel to make field differences visible.
Record exact matches as merge candidates and conflicts or missing keys as owner checks.
Give the recipient expected.csv, checklist.md and the basis for the action.

This guide and Table to Excel work in the browser. CSV files and input are not sent to an external server.

Cases that must stop

If a key is blank, values conflict for the same ID, or the source of truth is unclear, do not guess from a name or row position. Confirm with the data owner.