The short answer
To keep the first occurrence, preserve the row order, define which columns identify a duplicate, preview the groups, and export only after checking the result. “First” means the first matching row from top to bottom in the chosen sheet or CSV file. If the order is not meaningful, sort or document it before removing anything.
A safe workflow
1. Make a backup
Keep the original workbook or CSV unchanged. Duplicate removal is a data transformation, not a reversible view. Work on a copy and compare the number of rows before and after export.
2. Open the file locally
The browser tool accepts XLSX and CSV files and processes the selected data in the browser. It is useful for a quick cleanup when you do not want to upload a workbook to a server. Do not confuse browser-local processing with a guarantee that your device, browser extensions, or other software never access the file.
3. Select the sheet and header behavior
For an XLSX workbook, choose the sheet you intend to clean. Confirm whether the first row contains headers. If a header row is treated as data, it can be compared or exported incorrectly.
4. Choose the duplicate key
Select the columns that define “same row” for this task:
- All relevant columns: use when two rows are duplicates only if every selected value matches.
- One stable column: use when an order ID, email address, or SKU is the actual identity key and you have permission to process it.
- Several columns: use when a combination such as date plus SKU defines a record.
Do not select a convenient column merely because it is easy. A repeated city or status value is usually not a duplicate identity.
5. Preview duplicate groups
Review the reported groups before exporting. For each group, verify which row will remain and whether the later rows are truly redundant. Pay attention to blanks, leading zeros, capitalization, date formats, and cells that look identical but contain different underlying values.
6. Export and validate
Download the cleaned XLSX or CSV. Compare the output row count, inspect several retained first matches, and open the file in the application that will use it. For important workbooks, also verify formulas, formatting, merged cells, filters, and external links because a browser cleanup should not be assumed to preserve every workbook feature.
Worked example
Suppose a sheet contains these rows:
| Row | SKU | Region | Quantity |
|---|---|---|---|
| 2 | A-17 | North | 4 |
| 3 | A-17 | North | 4 |
| 4 | A-17 | South | 2 |
If SKU and Region are the key, row 3 duplicates row 2 and the first match is retained. Row 4 stays because its Region differs. If SKU alone is the key, rows 3 and 4 belong to the same group, so decide whether that business rule is actually correct before exporting.
Common mistakes
- Choosing all columns when the real key is only one ID.
- Choosing one descriptive column when multiple records can legitimately share it.
- Deleting before previewing the groups.
- Treating blank cells, formatted numbers, and text numbers as automatically equivalent.
- Skipping the post-export row-count and workbook check.
Sources and privacy note
The concept of selecting columns and retaining one instance follows Microsoft's Remove duplicate values support guidance. The tool's current local-processing behavior and supported formats are product-specific. Use authorized data and avoid placing workbook contents in URLs, analytics events, or public examples.
Tool link: https://www.removeduplicatesexcel.org/
Suggested visual: a synthetic five-row table with the selected key columns highlighted, followed by the previewed retained rows. Do not use a real customer workbook.