The challenge
An AEP implementation that was technically functional but practically unusable. Anonymous web users could not be connected to authenticated app users. Campaign activation was limited to known customers. Personalization was impossible for the majority of the user base.
The engagement began with a full inventory of identity signals available across every touchpoint — web, app, email, call center, and CRM. The goal was to understand what identifiers existed, where they were collected, and how they could be connected.
The identity graph design
We designed a probabilistic-to-deterministic identity resolution approach that used email hash as the primary deterministic key, supplemented by device fingerprint and cookie-based probabilistic matching for anonymous users. This approach was reviewed against the organization's privacy and compliance requirements before implementation.
The identity namespace strategy in AEP was restructured to support this model — creating a clear hierarchy of identity confidence levels that downstream activation systems could use to make appropriate personalization decisions.
Activation enablement
With a functioning identity strategy in place, we worked with the marketing team to define the first set of activation audiences — prioritizing high-value segments that had previously been inaccessible due to identity gaps. The first campaign using the new audiences launched within six weeks of the identity strategy going live.
Outcome
Resolvable anonymous users increased by 68%, more than doubling the organization's addressable audience for personalization and activation. The AEP investment, which had previously been difficult to justify, became a clear competitive advantage.