Clean a Contact List Without Merging Different People

Two rows say “Alex Hamilton.” One belongs to a customer; the other belongs to a supplier. A cleanup tool sees a match. Your business sees two different relationships. Separate tidying fields from deciding that records represent the same person. The second decision deserves more scrutiny.

Start from a recoverable copy Use an authorized export or backup, retain original row IDs, and keep the live list unchanged during the first trial. Our file-handling guide helps keep source and working copies identifiable. For a Google-based document, consult version-history guidance, but verify what your actual contact system can restore. A spreadsheet copy is not a complete backup of every connected customer record.

Standardize before suggesting merges Trim stray spaces and use a consistent display format for phone numbers while preserving the original values. Do not remove punctuation from every email address or assume one shared inbox means one person. For Google Contacts, follow the documented merge-and-fix process. Treat its suggestions as candidates to inspect.

Show the evidence side by side For each proposed match, show source IDs, names, organizations, contact methods, conflicting fields, and linked history. Fictional example: two rows share an email, but one says Accounts team and the other says Daniel. That may be a shared mailbox—not a duplicate individual. Another example: the same person has changed companies. You may need updated contact details while preserving the old relationship history, not a blind overwrite. Use the tracker approach to record candidate, decision, reviewer, reason, and next action.

Preserve permissions and relationships Do not turn one record’s marketing permission into permission for every address in a merged record. Keep source evidence and follow the business’s approved resolution process. Consult the FTC’s data-protection guide and your access checklist. AI can highlight similarities. Keep the identity decision within the approval process and assign uncertain cases using the handoff guide.

Test the cleanup before scaling it Review ten candidate groups containing exact matches, similar names, shared inboxes, changed details, and conflicting permissions. Record both correct merges and false suggestions with the test checklist. Measure correction work using the time-saving framework. Your goal is trustworthy records, not the largest number of deleted rows. Leave a candidate unresolved when the evidence is insufficient.