Quick Fix
Find Duplicate Customers in Excel With AI
Upload a copy of your customer spreadsheet and use AI to identify exact and possible duplicate customers without deleting anything automatically.
The problem
Customer lists often end up with the same person entered more than once. The duplicates may not be identical: John Smith and John M Smith may be the same person, or two rows may use the same phone number with slightly different names. AI can help identify records that look like duplicates so you can review them. It should identify them, not automatically delete or merge them.
What you need
- Excel customer spreadsheet
- ChatGPT or Claude
- A copy of the spreadsheet
Safety first
Make a copy of the spreadsheet first. Do not let AI automatically delete, merge or overwrite customer rows: the goal is a review list only. Customer spreadsheets can contain private information, so remove information you do not need to share with AI and follow your business and privacy requirements before uploading customer data.
Copy-paste prompt
I have attached a copy of my customer spreadsheet. Please find possible duplicate customer records using fields such as: - name; - email address; - phone number. Separate the results into: 1. Exact duplicates 2. Likely duplicates that need human review For each likely duplicate, explain briefly why you think the records may belong to the same person. IMPORTANT: - Do not delete any rows. - Do not merge any rows. - Do not overwrite the spreadsheet. - Do not guess when the evidence is weak. - I want a review list only. Show me the possible duplicate groups in a simple table so I can check them myself.
Steps
- Make a copy of your customer spreadsheet. Use the copy for this tutorial.
- Attach the copied spreadsheet to ChatGPT or Claude, then paste the prompt above.
- Review the two groups AI returns: Exact duplicates and Likely duplicates. Exact duplicates look essentially identical. Likely duplicates need you to decide whether they are really the same customer. For example, row 14 may show John Smith, john@example.com and 082 123 4567, while row 51 shows John M Smith, john@example.com and 0821234567. AI may flag this as a likely duplicate because the email and phone number match even though the names are slightly different.
- Check each suggested duplicate before making any changes in Excel. Do not automatically delete everything AI flags.
What success looks like
Before: you have a customer spreadsheet and do not know which records may be duplicates. After: you have a short review list, such as possible duplicate group 1: rows 14 and 51; reason: same email and phone number, with a slightly different customer name. The spreadsheet itself should still be unchanged.
If it goes wrong
The duplicate list is too broad. Please check again using stronger evidence. Prioritise: 1. matching email addresses; 2. matching phone numbers; 3. name similarity only when another field also supports the match. Do not delete or merge anything. Return a smaller review list.
Next useful tutorial
Clean a Messy Customer Spreadsheet With AIFix Any Excel Formula With AI in 2 Minutes