Save the import file before starting the CSV run

The CSV import report starts with the file that was previewed and confirmed for execution. Keep the original file unchanged, name it uniquely, and record the import date, author, import profile used, and the Driving History request ID in the same place. The original import file cannot be loaded from the Driving History afterwards, so a separate backup copy is necessary for later comparison of rows and values.

Spreadsheet programs can change product codes, long numbers, dates, and decimal separators when you open the file. Import the CSV into the spreadsheet program as text and set the columns containing the tags to text format before editing. After saving, also check the file in a text editor: the header row, number of columns, quotation marks, character set, and separator must match the base of the import profile used.

Limit CSV to only the intended change

In CSV import of products, each included column is part of the contract of the selected import profile. Use the profile's own export or template, keep the columns required for identification, and only remove the change columns that are allowed to be omitted in the profile. For example, in a price correction, the product's unique identifier and price column are essential, but an incorrect or extra column can either prevent the preview or target the change to the wrong field.

A missing column and an empty cell are not the same thing. Depending on the profile and field, an empty cell can either clear the value, produce a validation error, or be a valid empty value. Don't use an empty cell as a general reset, but check the preview for each row's classification and field-specific changes. Also, don't change the separator to a semicolon just because Excel defaults: keep the template separator and make sure the preview recognizes the headers as separate columns.

Preview the import before executing.

The preview for the CSV import of products shows whether the headers and rows are recognized and whether the row is classified as insert, update, unchanged, or invalid. Pause before confirming if the number of rows is different than expected, the tag matches the wrong product, the change columns show unexpected values, or the preview reports an unknown header, missing column, different column count, or invalid value.

In particular, compare product identifiers and values, which are easily changed by spreadsheet software. If the weight, price, or other number in the preview looks like a date, a rounded number, or a value stored with the wrong decimal separator, do not import. Go back to the saved file, correct the column data type, and recreate the preview.

Open Data Transfer History

The Data Exchange Run History is opened in the Control Panel under Data Exchange → Run History . On the /data-exchange/history page, you can filter runs by profile, status, and date. For each visible run, the import or export, profile, status, creation, start, and completion times, number of rows processed, administrator ID, and request ID are displayed. For failed runs, an error summary is also displayed.

The run history is not a complete line-by-line change log. The centralized history includes imports of products, customers, and product tab and tag associations, but not all CSV import types are reflected in it. The Load action applies to the file produced by a successful export; it does not restore the original CSV loaded for import.

Compare to the product's History tab

The Product History tab helps you verify whether a CSV import was performed on that product. The history shows the date of the event, the administrator, the action, the page, and any old and new values. A CSV import of products is recorded as an add or update entry with the action csv_import and the page csv_history.php .

However, the general product tag in the CSV import does not include field-specific old and new values or a link to the Driving History request ID. Therefore, the product History proves that the import affected the product, but does not alone prove which CSV row or column caused the change. A separate price history for the price can supplement the analysis, but the original CSV and Driving History data are still needed side by side.

Resolve unexpected CSV change

To investigate an unexpected product value change from a CSV import, start with the time and author on the product's History tab. Find the import from the same time period on the Data Transfer → Run History page, verify the profile, status, row count, and request ID, and then open your own backup of the exact CSV file that was run. Search for the product by its ID or SKU and review all the columns included in that row in a text editor.

For example, if the weight of a product changed unexpectedly, check the raw value, title, separator, and column position of the weight column. Also, make sure that Excel didn't convert the value to a date or decimal, and that an extra column didn't move values under the wrong titles. The same troubleshooting model works for price, inventory, category information, and other values that are updated with CSV; don't just assume the cause from the end result.

Fix error with restricted new import

Incorrect CSV import of products will not be undone from the Run History. To correct the error, create a new CSV with the correct template for the profile used, limit it to only the verified products and fields, and review each change in the preview before confirming. Also, keep the patch file and its request ID so that the original event and the fix can be distinguished.

If the original file does not exist, the centralized Run History is missing the import type, or the product History does not identify the changed field, do not perform a bulk restore import based on guesswork. For the support request, collect the product ID or SKU, changed field, before and after value, estimated time, Run History request ID, profile, administrator ID, and the retained CSV and preview report. This information can be used to narrow down the change without new side effects.