Sample data

Two demo files let you try FuzzyMatch without your own data. Click Try with sample data under each upload box, or download them:

• crm-contacts.csv: 52 contacts exported from a CRM, with Full Name, Company, Email, Phone, City and State.
• billing-accounts.csv: 51 accounts from a billing system, with Contact Name, Organization, Email, Phone Number, City and State.

All names, companies and addresses are made up, and phone numbers use the fictional 555-01xx range.

What's in them

36 customers appear in both files, written differently in each the way real systems drift:

• Nicknames plus a typo: Richard Garcia in the CRM, Rick Gacria in billing
• Reversed names: Zoe Thompson and Thompson, Zoe
• Typos: Patricia and Patrcia
• Company suffixes added, dropped or spelled out: Kestrel Freight Inc and Kestrel Freight, Incorporated
• Different case: emails and cities in capitals
• Phone formats: 227-555-0171 and +1 227 555 0171, plus (…), dotted and digits-only styles

Each file also contains near-duplicates of its own, the same customer entered twice, and about 15 records that appear only in that file.

Try this

1. Load both samples and click Next: pick columns.
2. Check the suggested pairs. Pair Full Name with Contact Name and Company with Organization if they aren't already, and leave Customer ID and Account No unpaired.
3. Run on Balanced, then switch to Loose and compare the extra matches you get.