Fuzzy Matching

Fuzzy matching identifies records that are similar but not exactly the same. While exact matching only catches identical values ("John Smith" = "John Smith"), fuzzy matching catches variations: "John Smith" matches "Jon Smith," "J. Smith," "Jonathan Smith," and "John P. Smith." It uses algorithms that measure how similar two strings are and returns a confidence score for each potential match.

Why It Matters

Exact matching catches less than half of real duplicates. People abbreviate names, misspell companies, use different email formats, and enter phone numbers in different formats. If you only match on exact values, "Acme Corporation" and "Acme Corp" look like two different companies. Fuzzy matching closes this gap and catches the near-duplicates that exact matching misses entirely.

Common Fuzzy Matching Algorithms

Example

A database has "Robert Johnson, Acme Corp" and "Bob Johnson, Acme Corporation" with different email addresses. Exact matching finds nothing. Fuzzy matching detects: name similarity 78% (Robert/Bob are known nicknames), company similarity 92% (Corp/Corporation), same area code on phone numbers. Composite score: 87% match probability. Flagged for review.

Related Terms

Related Resources

Duplicates slipping through exact matching?

We use fuzzy matching across name, company, email, and phone to catch the near-duplicates your CRM misses.

See What We'll Find