Collapse duplicate signups down to one row each

Duplicate signups, thanks to a broken import.

IntermediateRead 6 minDoing: 30 to 45 minutes once you've settled on the tiebreak rule, longer the first time the copies genuinely disagree7 steps

▶ Start

What you'll have at the end

One surviving row per duplicate signup, with the before-and-after row counts written down

Before you start

You need

Write access to a signups table where a broken import wrote some signups more than once, with a column that names the real person (an email address is the common case) and a timestamp on every row.

Not covered

Fixing the import job itself, or working out why it ran the batch twice in the first place.

Leans on

Find the key, then measure the damage

Pick the column that actually identifies a person
Count the damage before touching anything

Decide the survivor, then look before deleting

Rank each duplicate group to pick a survivor
Preview exactly what the delete would remove

Remove the extras and prove the count holds

Delete the extras inside an undoable transaction
Check the count, then commit or roll back

Close the door the import walked through

Stop the import from writing it again

Where to go next

Standardize a status column so every capitalization spelling counts as one value

for a status column where the same value shows up under three different spellings.

Checked 26 Aug 2026

Part of the SQL cookbook