Retail media has historically depended on identity to connect advertising to sales. For many retailers, that meant relying on loyalty programs or logged-in users to establish a deterministic link between exposure and purchase.

This episode revisits how that problem was originally solved at Walmart. Without a loyalty program or widespread logins, Walmart developed an approach based on consistent, anonymous identifiers built from available signals such as browser, device, and transaction patterns. This allowed the business to connect behavior and outcomes without relying on third-party identity or persistent user accounts.  

As retail media evolved, much of the ecosystem shifted toward third-party identity and audience models. Data was collected, matched externally, and returned as audiences for activation. While this enabled targeting at scale, measurement remained largely within the retailer’s control, limiting the effectiveness of these approaches.

The discussion in this episode focuses on how that model is changing.

New approaches emphasize capturing data at the point of interaction, resolving identity within the retailer’s own infrastructure, and enabling real-time activation. Rather than relying on batch processing and external identity graphs, retailers can now operate with a combination of known and anonymous identifiers across a much larger portion of their traffic.

This shift reduces dependence on third-party identity and expands the scope of retail media beyond logged-in users. Loyalty programs and deterministic identifiers still provide value, but they are no longer the sole mechanism for enabling targeting and measurement.

The result is a model where retailers maintain control over their data and measurement while activating audiences across channels in real time.