Visa uses a matching logic system to compare the name you provide (source name), with the name on file held by the issuer (target name).
This system uses a combination of n-grams and Levenshtein distance to handle slight differences in how names are entered.
Visa splits the names into groups of three letters, and uses Levenshtein distances to get an overall similarity score. This score determines if the names are close enough to be considered a match. The system applies different weights to different parts of the name. For example, it often gives extra weight to the last name because that is usually the most important part for a good match.
The system can return one of three results:
- Match: The names are very similar.
- Partial Match: The names have some things in common but aren't an exact match.
- No Match: The names are too different.
This approach lets Visa handle common errors like:
- Spelling mistakes: "JAMS" instead of "JAMES"
- Shortened names: "Jas" instead of "James"
- Mixed-up letters: "SJAME" instead of "JAMES"