Retail media has grown from a search-only Amazon budget to a full media channel that now includes display, video, CTV, and offsite programmatic — all built on the retailer's first-party purchase data. The appeal is real: you're reaching in-market shoppers using actual purchase behavior, not probabilistic audience segments. Amazon knows someone bought your competitor's product last month. That signal is genuinely valuable, and no amount of DMP audience modeling can replicate it. The problem is that every retailer uses their own attribution model, defined by them, to report the ROI of their own inventory. Amazon's ROAS is measured by Amazon. Walmart's is measured by Walmart. There is no independent auditor.
The standard retail media attribution uses last-touch within the retailer's ecosystem, often with a 14-day attribution window and a definition of "conversion" that includes any purchase in the category — not just your specific SKU. This systematically overstates ROI for everything. The only credible independent measurement approach is incrementality testing: running holdout experiments to measure the lift your retail media spend actually drove versus what would have happened without it. Very few brands are doing this systematically. The gap between reported retail media ROAS and incremental retail media ROAS is often the largest measurement gap in the entire marketing budget.
Every retailer's self-reported ROAS looks excellent. Incrementality testing consistently shows the incremental lift is a fraction of the reported number. The gap varies by category and retailer, but the gap is always there.
CPG brands now manage retail media budgets across Amazon, Walmart, Kroger, Instacart, Target, CVS, and dozens more — each with different UIs, measurement methodologies, and attribution windows. The operational overhead is enormous.
On-site retail media (sponsored search) is high intent and relatively defensible. Offsite retail media (display and CTV using retailer data) competes directly with standard programmatic — but costs significantly more, justified by the data premium.
The dominant retail media network. Includes Sponsored Products, Sponsored Brands, Sponsored Display, Amazon DSP, and streaming TV ads — all powered by Amazon's first-party purchase data.
Consolidated retail media buying platform that connects advertisers to 225+ retailer networks in a single UI. The aggregation layer that addresses the fragmentation problem.
Walmart's retail media network. Access to 140M+ weekly shoppers across Walmart's physical and digital properties, with in-store, on-site, and offsite inventory.
Grocery retail media across the Instacart marketplace and 1,400+ retail partners. High-intent grocery shopping signals with category-level targeting.
Target's retail media network. On-site search and display combined with offsite programmatic using Target Circle loyalty data and first-party purchase history.
White-label retail media technology that powers the retail media networks of 80+ retailers including Woolworths, Albertsons, and Petco. The infrastructure behind many retailer networks.
Not sure how to evaluate your retail media spend?
Answer 5 questions and get a framework for measuring retail media lift independently of the retailer's own attribution.
Get a stack recommendation →Clearpath Analytics builds independent measurement frameworks for retail media — incrementality testing, holdout design, and cross-retailer reporting normalization.