Name Match Clinic
One name, fourteen spellings: transliteration's share of failed transfers, diagnosed and cured.
The job
Measures how many failed person-to-person transfers and remittances are really down to names spelled differently when converted from another alphabet. Fixes the matching, the wording and the bot answer, and holds each receiving bank to account with before and after verdicts.
The moment
The product manager types a common first name into a test box, and a family tree spreads out into 14 spellings.
38% of last holiday season's name-mismatch complaints involved the same first name spelled differently by the sender and the receiving bank.
Two branches of the tree are thick red at one receiving bank. That evidence goes into a letter to the bank.
Within the week, the bank changes how closely names must match. The scorecard before the holidays shows its row turn from red to green.
What it does
- Send the evidence letter
- Draft copy fix and bot answer
- Register a fix
- Drill any spelling
What you see
Clinic layout (diagnosis, exhibits, prescription) around a name family tree, with a live try-a-name box and a per-bank scorecard
What it moves
Name-mismatch complaints per week, by receiving bank and group of spellings, before and after each fix, and the share of all failed-transfer complaints caused by name mismatches.
Built for
- Product & digital
- Partners & ecosystem
- Frontline staff
