Your customer's name does not match their ID. Do not reject the right person.
Charles Archibong, Co-founder
A name mismatch needs an explanation, not an automatic rejection. Preserve the original spelling, identify the kind of difference, and corroborate the person with independent evidence before deciding.
Key takeaways
- Keep the original name and any normalised comparison value separately.
- A spelling explanation supports review; it does not establish identity.
- Use independent attributes and a recorded review outcome.
A name mismatch should send your team to the evidence. It should not automatically approve an applicant or force a genuine customer to start again. First identify why the names differ, then check whether the remaining evidence supports the same person.
An applicant's family name appears first on one record and last on another. Your form strips an accent. A document contains a second given name that the customer did not enter. These are different review questions from a completely different identity.
Find the kind of mismatch
Separate spacing and punctuation, name order, omitted components, transliteration, and a claimed legal change. Keep unexplained differences in their own category.
The W3C's guide to personal names (opens in a new tab) shows why one global first-name/last-name model fails. Use that insight to design your review process, rather than treating a familiar naming pattern as a condition of trust.
Preserve the source spelling. Store any comparison form separately, with the transformation recorded. Removing characters from the only copy destroys evidence a reviewer may need later.
Corroborate without lowering the bar
Compare the relevant independent attributes: date of birth, document details, authoritative records where available, and the evidence connecting the applicant to the identity. A plausible spelling variation cannot compensate for contradictory dates or a failed identity check.
NIST's identity-proofing guidance (opens in a new tab) separates resolving an identity from validating evidence and verifying its owner. Use those separate questions in your review notes.
Ask the customer for the smallest clarification that closes the gap. Do not request a new selfie when the unresolved issue is how a family name was recorded.
Give the reviewer a repeatable decision
Record the two values, their sources, the mismatch category, the corroborating evidence, and the remaining uncertainty. Your outcome can be accepted variation, further evidence needed, or unresolved contradiction.
For an illustrative reordered-name case, matching other attributes may support acceptance. For a claimed legal change, spelling normalisation alone is insufficient; follow a documented evidence route. Neither example is a universal acceptance rule.
Test the cases your form mishandles
Create authorised test cases covering accents, multiple given names, different scripts and name order. Measure review time, subsequent corrections and genuine-customer completion alongside fraud outcomes.
Use Toden's identity-verification overview to choose the checks that answer your evidence question, then test your workflow. Keep the acceptance policy explicit; a product result does not explain a mismatch on its own.
Sources
Charles Archibong
Co-founder
Charles Archibong co-founded Toden. He writes about identity verification, financial technology, and the practical work of building trusted digital services.
