CSV files imported into BigQuery complete successfully but the imported data does not match the source byte-for-byte. What is the most likely cause?

BigQuery's default encoding for CSV files is UTF-8. If the source CSV file uses a different encoding (e.g., ISO-8859-1 or UTF-16), characters that are not representable in UTF-8, or are represented differently, will be converted or replaced during the import process, leading to a byte-for-byte mismatch even if the data appears visually similar. To ensure an exact match, the encoding must be specified correctly during the BigQuery load job. Incorrect options: "The imported data was not specified as CSV during load" would likely cause the load to fail or result in completely garbled data, not just a byte-for-byte mismatch with successful completion. "Invalid rows in the CSV were skipped during import" implies that some data would be missing, but the question states the import completed successfully and the data doesn't match byte-for-byte, not that it's incomplete. "The CSV files were not ETL-processed before loading" is too general; while ETL can transform data, the core issue described is a low-level data representation problem, not a lack of higher-level transformation.