How to remove duplicate records from a meeting CSV

A weekly management CSV can contain the same action, customer, or project more than once. That can inflate totals, give one issue extra attention, or leave people unsure which row to discuss. But two rows that look similar are not always duplicates: one may be a legitimate update from a different week.

Use the CSV Editor to find, sort, edit, and remove duplicate rows in your browser. Start with a copy of the exported file, then check what each repeated row represents before deleting anything.

Check the meeting CSV

1. Open the CSV Editor and open a copy of the CSV you plan to review. Keep the original export unchanged so you can compare or recover a row if needed.
2. Check the column headings and identify what makes a record unique. For an action list, that could be an action ID. For another report, it might be a combination such as project, owner, and reporting week.
3. Use the editor's find and sort features to bring likely repeats together. Search for a known ID or name, then compare the relevant fields, including dates, owners, status, and notes.
4. Remove rows only after confirming they are duplicates, then check the edited list and totals before sharing it with the meeting group.

A repeated name by itself is not enough to prove that two records are the same. If your report has no unique ID, decide which combination of fields identifies one record before you start removing rows.

What the repeated rows mean

Every field in both rows matches

These are likely exact duplicate rows, often caused by exporting or combining the same data twice. If the information is truly identical and the report is meant to show each record once, removing the extra copy is usually safe. Keep the untouched export until the meeting is over.

The ID matches but the status or date differs

This may be a change over time, not a duplicate. For example, an action might appear once as open and again as complete because the CSV contains weekly snapshots. Check whether the meeting needs only the latest status or a history of changes. If the history matters, preserve both rows or prepare a separate current-status view rather than deleting one without a record.

The name matches but other details differ

Two people can share a name, and one person can own several actions. Compare IDs, project names, owners, due dates, and descriptions. If those details point to different work, keep both records even if the name is identical.

Some fields are blank in one copy

These rows may be incomplete versions of the same record. Do not assume the row with more filled-in fields is automatically correct. Confirm which source is current, preserve any useful information from both rows, and make one complete record only if your team agrees the entries refer to the same item.

How to clean the file safely

1. Keep an untouched copy. Duplicate the export before editing. Use a clear filename, such as weekly-review-cleaned.csv, so the meeting version is easy to distinguish from the original.
2. Choose a record key. Use a unique ID when the file has one. Otherwise, choose the smallest set of fields that reliably identifies one item. A name alone is rarely enough, and a date may be part of the record rather than a way to identify it.
3. Compare the full record. Sort by the key fields and inspect each cluster of repeated entries. Confirm whether the rows match across the fields that matter to the meeting. A shared project or owner does not make two separate actions duplicates.
4. Remove only confirmed copies. Use the editor's remove-duplicates feature for exact repeated rows. If two rows overlap but contain different information, review them manually and retain the current or complete information. Do not treat a cleanup as permission to erase history.
5. Check the results before sharing. Compare the cleaned row count with the original and make sure the reduction matches the duplicates you reviewed. Recheck important totals, open actions, owners, and due dates. Open the saved file again if possible to confirm it still has the expected columns and values.

Make next week's review easier

Repeated records often come from combining exports from several people or reporting periods. Agree on one source of truth for each list, and ask contributors to update an existing action rather than add a second row for the same item. A stable ID makes this much safer than matching by names or descriptions.

Keep the reporting period in a separate field when the meeting needs to compare weekly history. If the agenda needs only the current state, create that view from the latest confirmed record and retain the full history elsewhere. This prevents old updates from being mistaken for accidental duplicates.

The CSV Editor processes the file in your browser, so the CSV stays on your device while you review it.