# Best Practices for Data Import

Dataloader from MuleSoft accepts many CSV layouts, but small formatting mistakes can make processing a file difficult or impossible. Follow best practices to avoid import errors.

## Prepare Your CSV File

- Use a spreadsheet app such as Microsoft Excel to build the CSV file, then verify with a text editor if needed.
- Validate the structure with a tool such as [csvlint.io](https://csvlint.io/).
- Save the file as **UTF-8**.
- Use a single header row with no duplicates or empty header cells.
- Remove empty rows, including hidden empty rows at the end of the sheet (check in a text editor).
- Put column headers in the first row (for example, First Name, Last Name). Use Salesforce field labels as headers when possible. Dataloader from MuleSoft maps them automatically.
- Format date fields by using a [supported date format](/docs/platform/dataloader/guide/supported-date-formats.md).
- Wrap fields in double quotes (`"`) when needed for delimiters.
- Keep files under 50 MB.

## Performance Tips

- For files larger than about 200 rows, select **Bulk API** on the last step of the import wizard for better throughput on large loads.
- If imports fail, review the result files on the task list. They explain row-level errors.
- Schedule large imports during off-hours when possible to avoid resource contention. When many users access Salesforce concurrently with a large data load, processing slows and can hit the two-hour task timeout. Tasks that exceed two hours are automatically terminated.

## See Also

- [Importing Data into Salesforce](/docs/platform/dataloader/guide/import-data.md)
- [Use Supported Date Formats](/docs/platform/dataloader/guide/supported-date-formats.md)
- [*Salesforce Developers*: Introduction to Bulk API 2.0 and Bulk API](https://developer.salesforce.com/docs/atlas.en-us.api_asynch.meta/api_asynch/asynch_api_intro.htm)