Remove exact duplicate records and review likely matches.
Choose what makes a record identical, preview which rows remain, and export a separate CSV. Exact rules can remove duplicates. Similarity scores remain suggestions for a person to assess.
Deduplicate complete rows
Use exact-row deduplication when every field must match. The first record in source order is preserved, so the result follows a predictable rule instead of choosing a winner silently.
Deduplicate by a business key
Select stable columns such as a customer ID or an approved combination of identifiers. Selected-key deduplication compares only those values, which keeps the decision explicit and repeatable.
Treat fuzzy matches as evidence
Spelling differences and similar names can indicate duplicates, but they can also identify different people or companies. Fuzzy results show pairs and scores without modifying the file.
Check the result before export
Review the row count and preview after each deduplication step. Undo a step if the key was too broad, then download a new CSV without overwriting the source.