Choose the identity columns that define a duplicate, preview the retained rows, and export a deduplicated copy locally.
Audit CSV deduplication by selected key columns with row counts, duplicate groups, retained rows and key completeness evidence.
Use the same deduplication core with keep-first or keep-last behavior, optional comparison normalization, and evidence showing which input row each removed duplicate matched.
Choose key columns in the primary tool above, then review the exact retention policy before export.
CSV has no universal schema or type system. Delimiter detection and type summaries are practical heuristics, while repairs create a new copy instead of silently altering the source. Excel conversion focuses on cell values and simple worksheets rather than promising perfect preservation of formulas, macros, charts, external links, formatting or every workbook feature.
Use the tool first, then use these checks to interpret, verify and hand off the result without displacing the primary workflow.
An entire-row match, email match, order-ID match, or multi-column business key can produce very different results. Select the key columns deliberately and review duplicate groups before export so distinct records are not collapsed merely because they share one convenient field.
Keeping the first row favors earlier input order; keeping the last favors later input order. Neither rule knows which record is newest or most accurate unless the source is already ordered accordingly. Sort by a trustworthy timestamp or priority field first when recency determines the winner.
Whitespace and letter case may be irrelevant for some identifiers but meaningful for others. Normalize only when the business rule permits it, and preserve the original cell values in the retained row. Review the removed-row preview to verify that normalization is merging the records you intended.
Compare input, kept, removed, and duplicate-group counts, then inspect rows with blank key fields. Multiple blank keys can collapse into one group even though the underlying records are unrelated. Use the existing blank-key audit and a representative sample before replacing the source file.
WebToolArc