Losing Identity Hurts Advertising. It May Break Measurement.
Losing Identity Hurts Advertising. It May Break Measurement.
By Brian PozeskyFor more than 20 years, the advertising industry has been having the same conversation. Everyone knows the drill.
Cookies are disappearing. Mobile identifiers are restricted. Match rates are declining.
At this point, it’s background noise.
Chrome threatens restrictions, pulls back, and then moves forward again. Safari and Firefox limit tracking without opt-ins. Apple’s App Tracking Transparency framework continues to pressure mobile identifiers and restrict cross-app tracking.
Meanwhile privacy policies have increasingly become competitive tools in the fight for marketing dollars. Identity graphs, clean rooms, retail media networks, and an ever-expanding ecosystem of probabilistic identity solutions have emerged to help marketers preserve consumer recognition and the personalization that drives performance in the face of all this flux.
But there is a second implication of this environment that receives far less attention:
If consumer identifiers are becoming increasingly unstable, measurement that depends on them is in serious trouble.
This challenge is compounded by a structural reality. According to Nielsen and Circana, roughly 85–90% of consumer-packaged goods sales still occur in physical retail stores. That leaves brands struggling to determine how marketing truly influences market-level outcomes.
Consider how products are purchased. Households shop across multiple retailers every month. Each retailer operates within its own data environment, often with limited interoperability. Yet measurement providers continue to rely heavily on identifiers to link exposure to outcomes.
The industry is attempting to measure market-level commercial impact using datasets that capture only a fraction of consumer behavior and that fraction can shift every time a privacy policy changes.
According to the IAB, more than 60% of digital traffic now occurs in environments where third-party identifiers are unavailable. And the composition of that 60% changes with every new or retroactive privacy update.
Even where reliable identity exists, visibility remains incomplete. Exposure-to-purchase match rates in retail media environments often hover around 20%, depending on the retailer and identity framework.
In practical terms, most purchases that occur after a campaign are never connected to the advertising exposure that may have influenced them. They simply never appear inside today’s measurement model.
Unmatched sales must be estimated from unreliable fragments. Cross-retailer halo effects disappear. Offline lift is understated. Market-level impact becomes fragmented and harder to interpret. The result is measurement that can look extremely precise while being fundamentally incomplete.
None of this means identity-based marketing is flawed. Identity remains powerful for targeting, frequency management, and channel optimization across both digital and offline environments.
85–90% of CPG sales still happen in physical retail stores
60%+ of digital traffic now occurs where third-party identifiers don’t exist
In many retail media environments, exposure-to-purchase match rates hover around ~20%
But measurement must evolve.
Fortunately, today’s toolbox is far more advanced than it was even a few years ago.
Advances in machine learning and causal inference modeling combined with the ability to process billions of data points quickly now allow us to analyze commercial systems in ways that were previously impractical. Rather than relying on user-level matches that project upward to national purchase aggregations, modern measurement approaches can evaluate patterns of sales across a brand’s full retail portfolio of 1000s of stores, isolate the effects of media pressure, account for seasonality and day-of-week variation, and uncover the true causal drivers of growth.
McKinsey & Company has written extensively about the importance of robust causal measurement frameworks as ecosystems grow more complex and privacy constraints tighten. As retail environments remain fragmented and identifiers remain unstable, measurement must increasingly rely on models that evaluate total economic impact across every store and every channel in which a brand operates.
The opportunity ahead is not simply finding new ways to track consumers.
It is building measurement frameworks capable of understanding what truly drives growth even when individual identity is no longer reliable.
The solution: OmniChannel Performance Platform
If measurement frameworks built around identity are becoming structurally incomplete, the industry needs a different approach.
At Pēq, we’ve built the Omnichannel Performance Platform to measure the true incremental impact of marketing across retailers, channels, and markets, even when consumer identifiers are missing.
By Rikki Marler
