Phonetic matching finds values that sound similar.

It is useful for names because the same name can be spelled in several ways.

Examples:

  • Steven and Stephen
  • Smith and Smyth
  • Jon and John

How phonetic matching works

Phonetic algorithms convert words into a code based on pronunciation. If two words produce the same or similar code, they can be treated as a match.

Common phonetic algorithms include Soundex and Double Metaphone.

🔍 Good to know

Phonetic matching is useful for names. It is not a good default for emails, domains, IDs, or product SKUs.

Phonetic matching vs fuzzy matching

Phonetic matching focuses on sound.

Fuzzy matching focuses on string similarity.

Use phonetic matching when names may be spelled differently but pronounced similarly. Use distance matching when you want a similarity threshold across text values.

Datablist workflow

Datablist supports Metaphone matching in the Duplicates Remover.

Use it for name columns when exact or smart matching misses likely duplicates. Combine it with multi-column deduplication, such as first name + last name + company domain, to reduce false positives.