The Ghosts in Your Data: Duplicate Profiles Haunting Personalization

Publication date: October 26, 2026

Laura has a personal email, a work email, and an app on her phone. To her, that is ordinary life. To a customer database, it may look like three different Lauras, each patiently awaiting a welcome email. Nobody invited those ghosts, but they all show up in the campaign.

A brand knows Laura by email, phone number, and mobile-app ID. Its systems treat her as three customers. Laura remains one person, even as she receives three welcome messages. Before blaming creative, we should ask how identity is resolved.

Laura, Laura, and the other Laura

Adobe Experience Platform works with identities and profiles from different sources. Teams must decide which identifiers are reliable, how they relate, and under which conditions to join them. Incorrect merges can be worse than duplicates, introducing problems in audiences, measurement, and customer experiences.

The temptation is to merge everything and celebrate the arrival of a single customer view. But a bad match can be worse than a duplicate: it can combine histories that belong to different people.

Duplicate profiles inflate audience counts, lead to repeated impressions, and undermine frequency controls. Fragmented purchase and engagement data can also make a successful journey look ineffective. Marketers end up optimizing an incomplete picture.

The mystery lives between systems

Map sources and namespaces, examine identity rules and event quality, and compare distinct people with the identifiers recorded. Test actual fragmentation scenarios and measure failed suppressions before changing production settings.

Next time a dashboard proudly announces three new customers, it may be worth asking whether they are three people or Laura switching devices. It sounds like a small question—until marketing decides how much to spend reaching them.