At POSSIBLE 2026, industry leaders filled panel after panel with the same refrain from 2024 and 2025: retail media measurement is fragmented, walled gardens won’t talk to each other, and standardization is “just around the corner.” U.S. retail media spend continues to climb, and most of that budget is still evaluated network by network, with no unified view of what’s actually driving incremental revenue. If you’re a performance marketer waiting for Amazon, Walmart Connect, Kroger Precision Marketing, and Instacart Ads to agree on common measurement standards before you build a coherent attribution framework, you’ll be waiting forever. Here’s why — what to do instead.
Why Walled Gardens Have Zero Incentive to Standardize
This isn’t cynicism. It’s economics. Every retail media network monetizes proprietary purchase data by making it accessible only within its own ecosystem. The moment Kroger and Amazon agree on a shared attribution methodology, they lose the asymmetric information advantage that justifies their CPMs. Enabling true apples-to-apples comparison across networks would commoditize their inventory overnight.
The standardization bodies — IAB, MRC, ANA — have been workshopping frameworks for three years. Progress has been incremental at best. Top-down standardization requires voluntary cooperation from parties with zero economic incentive to cooperate. The measurement problem has to be solved from the demand side: by you, the advertiser, using your own data and your own methodology.
Establish a Channel-Agnostic Conversion Source of Truth
Before you can compare retail media networks to each other (or to channels like programmatic direct mail, paid social, or CTV) you need a conversion dataset that doesn’t belong to any single network. That’s your first-party transaction data: CRM records, POS logs, ecommerce order histories, and subscription events.
Stop relying on network-reported conversions as your primary KPI. Each network uses different attribution windows, different match methodologies, and different definitions of “incrementality.” When Instacart reports a 4x ROAS and Amazon reports a 6x ROAS, those numbers aren’t comparable — they were calculated using different rules applied to overlapping customer populations.
Instead, pipe your transaction-level data into a clean room environment — AWS Clean Rooms, Snowflake’s data sharing layer, or a purpose-built solution like InfoSum or Habu — and match conversions back to exposure logs from each network on your terms. You set the attribution window. You define the match logic. You control the deduplication. This is table stakes for any cross-network measurement framework that produces trustworthy numbers.
Layer in Holdout-Based Incrementality Testing
A unified conversion dataset tells you who converted after exposure. It doesn’t tell you whether the exposure caused the conversion. That’s where incrementality testing comes in — the single most important discipline missing from most retail media strategies.
Run holdout-based tests on every network and every channel in your mix: suppress a randomized control group from receiving impressions on a given network, then compare conversion rates between the exposed and holdout populations using your first-party transaction data — not the network’s self-reported metrics.
This is how performance direct mail already works at Postie. Every campaign can include holdout groups to measure true incremental lift — the additional revenue mail caused versus what would have happened organically. The same discipline applies to retail media. If you’re spending $500K per quarter on Walmart Connect and you’ve never run a holdout test to validate that spend, you’re optimizing to a fiction.
Incrementality tests won’t just tell you which networks perform — they’ll reveal which networks are claiming credit for conversions that would have happened anyway. Expect uncomfortable findings. That’s the point.
Stitch It Together With Media Mix Modeling
Clean room matching gives you unified attribution at the individual level. Incrementality testing validates whether that attribution reflects real causation. Media mix modeling (MMM) ties it together at the strategic level, quantifying how each channel — retail media networks, direct mail, paid search, CTV, email — contributes to total revenue.
Modern MMM platforms like Meridian (Google’s open-source framework), Robyn (Meta’s), or commercial solutions like Measured and Recast can ingest your cross-network incrementality results as Bayesian priors, which dramatically improves model accuracy. Feed in your Postie matchback attribution data alongside retail media spend and paid social investment, and you get a single model that tells you where your next marginal dollar should go.
The critical insight: MMM doesn’t require network cooperation. It runs on your spend data, your conversion data, and your incrementality results. It’s entirely demand-side infrastructure. No one needs to agree on standards for you to answer the question every CMO is asking: “What’s actually working?”
Use Direct Mail as Your Incrementality Benchmark
Here’s a practitioner tip most retail media teams overlook: programmatic direct mail is one of the cleanest channels for incrementality measurement because exposed and holdout populations are defined at the household level before the campaign runs. There’s no probabilistic matching, no cross-device guessing, and no viewability ambiguity. Either the mailpiece arrived at the household, or it didn’t.
That makes direct mail an ideal calibration channel for your broader measurement stack. If your MMM shows direct mail driving strong incremental ROAS on reactivation audiences, and you have the holdout data to confirm it, you now have a high-confidence benchmark to pressure-test other channels against. When a retail media network claims superior ROAS but can’t produce holdout-validated incrementality data, you know exactly how much skepticism to apply.
This is the same principle behind using direct mail matchback data in platform negotiations: walk into Walmart Connect or Amazon conversations with household-level, independently verified conversion data, and challenge self-reported figures with deterministic results.
Stop Treating Standardization as a Prerequisite
The retail media measurement problem isn’t a technology problem. Clean rooms exist, incrementality testing is well understood, and the tools to unify your media mix are ready right now. The real problem is waiting for walled gardens to voluntarily give up their asymmetric information advantage.
You don’t need to wait for an industry consensus that may never come. You can take control of your measurement strategy today using your first-party data, built-in incrementality testing, and a platform that provides true, deterministic truth.
Postie’s Programmatic Direct Mail platform is built exactly to solve this measurement gap. By operating on a completely independent identity layer at the household level, Postie gives you the deterministic matchback attribution and native holdout testing you need to prove true incrementality. We turn your offline and online touchpoints into a unified, measurable growth engine.
Ready to stop guessing and start scaling with true incremental measurement? See how Postie’s measurement works →
