The Scenario
RevOps is consolidating three contact databases into one Excel workbook before the CRM migration. You've been handed the merged file: 600 rows, pulled from a legacy CRM, a LinkedIn campaign download, and an event registration system. Some contacts appear in two or all three sources under slightly different names, different email formats, or just a LinkedIn URL in one record and only an email in another. Your job is to find and collapse the duplicates before the file goes anywhere near the new CRM. The migration window is next week.
The bad version:
- Run a duplicate check on the email column — finds 200 obvious matches, leaves the other 400 rows you still need to evaluate
- Try to match the remaining rows by name and company using a formula — "Jon Smith" and "Jonathan Smith" at the same company don't match because the name strings are different
- Spend two days manually reviewing the fuzzy cases, flag 30 rows as "suspected duplicates — check this one," and hand off a workbook with a disclaimer that a few rows might still be wrong
The CRM migration is on a hard deadline. Your disclaimer is not part of anyone's plan.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It reads the contact data across the rows and uses PDL's person identity resolution to assign a unique person ID to each contact, then flags rows that share the same ID as duplicates.
Use PDL's person identify endpoint on every row in my Excel Combined Contacts table — look up by email in column A and LinkedIn in column B, write the PDL person ID to column C, and mark duplicates in column D.
What You Get
- Column C populated with a unique PDL person ID for each row that resolves
- Column D flagged with "Duplicate of row X" for any row whose PDL person ID matches another row in the table
- Rows where PDL cannot resolve using email or LinkedIn left blank in column C with a note in column D so you know exactly which contacts need manual review
- A ready-to-review workbook where duplicates are explicitly labeled, not just suspected
What If the Data Is Not Quite Ready
Some rows have no email and no LinkedIn — only first name, last name, and company
For rows where column A and column B are both empty, attempt PDL identity resolution using first name from the Name column and company from the Company column. Write the resolved PDL person ID to column C if found. Flag unresolvable rows in column D as "Insufficient signals — manual check."
You want to know which source database each contact originally came from to help choose the primary record
For each contact, resolve via PDL and write the person ID to column C. Flag duplicate rows in column D and note the source database — CRM, LinkedIn, or Event — in column E. When two records resolve to the same PDL ID, recommend the one with the most complete fields as the primary in column F.
You want to automatically merge duplicate records into one row
For each contact row, resolve via PDL and write the person ID to column C. Group rows by person ID. For each group with more than one row, merge into a single row by keeping the most complete value for each field — prefer rows with a work email, then rows with a LinkedIn URL, then rows with a full name. Write the merged contacts to a new Deduplicated sheet.
Resolve identities, merge duplicates, score completeness, and flag records missing key fields for outreach
Resolve all contacts via PDL identity resolution. Merge duplicates into single rows on a Deduplicated sheet. Score each surviving contact from 0 to 5 based on how many of these fields are populated: email, LinkedIn, company, title, phone. Flag any contact with a score below 3 in a Data Quality column as "Incomplete — enrich before outreach."
One prompt handles identity resolution, deduplication, merging, and quality scoring — the CRM admin gets a clean, import-ready file with no ambiguous rows.
Try It
Get the 7-day free trial of SheetXAI and open any Excel workbook with a merged contact list from multiple databases. Ask it to resolve each contact via PDL identity resolution and flag the duplicates. Then see how to enrich surviving contacts with job title and company data or go back to the People Data Labs overview.