Loyalty glossary · 7. Data and analytics (31)
Identity Resolution
Identity Resolution is the process of linking different identifiers, transactions, and interactions to the same individual or household, creating a single unified view for the programme.
Loyalty programmes run on the assumption that every transaction maps to a known person. Identity resolution is the process that makes that assumption true, by linking identifiers such as email addresses, card numbers, and partner accounts to one member. Without it, accrual is a guess, because the programme cannot tell whether a point earning event belongs to a new member or an existing one.
Most programmes have fragmented data: a member books a stay with one email, earns via a partner with another, and calls support with a third. Activity-based qualification rules then fail, because a rule that requires three stays in a year cannot see that the same person made all three. The programme rewards three different impostors instead of one loyal member.
Consider two accounts that each hold 500 points. Identity resolution recognises them as the same member, combining the balances into 1000 points. If the programme's top tier requires 1000 points, resolution moves that member from 0 tiers to 1 tier. The difference is not cosmetic; it changes the member's status, earn rate, and future accrual.
The actuarial model that prices breakage and liability depends on the number of unique members. Unresolved duplicates inflate that number, making every member-level assumption too optimistic. Resolve the identities first, and the model stops treating one person as two liabilities.
The danger is not only under-merging. False positives, where two different people are merged, are worse than no resolution. They create a single profile that contains points and stays from unrelated individuals, which then accrues benefits neither one earned. Identity resolution must be conservative, not aggressive.