How Agentic Commerce and Retail Media ROI Are Reshaping Digital Strategy
The Drum's piece on OpenAI and agentic commerce, alongside Supermarket Perimeter's Groceryshop '26 agenda, converges on one conclusion: AI-mediated discovery is now a primary planning variable for retail media buyers.

MediaPost breaks the week's headline metric: Kroger and Instacart outperform Amazon on retail media ROI. The companion signals — OpenAI's agentic commerce trajectory, the Groceryshop '26 agenda, and Vusion's acquisition of In-Store Media — compress into a single operating picture for digital commerce operators.
The ROI anomaly
- Metric: Study cited by MediaPost shows Kroger and Instacart delivering higher ROI for advertisers than Amazon.
- Volume vs. efficiency divergence: Amazon retains the largest share of retail media spend despite trailing on advertiser return. Spend concentration and return concentration are not the same variable.
- Operating read: Smaller networks extract more measurable value per dollar. CPM-weighted allocation and ROI-weighted allocation produce different output. Share leader is not the efficiency leader.
- Planning implication: Re-run incrementality tests on non-Amazon networks before the next quarterly budget cycle. The cited study is directional, not a substitute for first-party measurement.
Agentic and physical shifts
GLP-1-driven basket composition sits alongside AI as a leading agenda item, confirming that demand-side and supply-side signals are both in flux.
- Discovery path extending from keyword search into agentic recommendation. New touchpoints require new instrumentation.
- Last-click attribution will undercount upper-funnel agentic influence. Expect deterministic attribution rebuilds.
- Sixteen:Nine confirms Vusion's acquisition of In-Store Media. Infrastructure consolidation in digital signage and in-store inventory raises shelf-edge addressability.
- Expect CPM parity pressure between digital and physical retail media surfaces as the in-store stack matures.
Broader coverage of social and consumer behavior shifts affecting retail demand patterns appears at Lefty Magazine.
Binary summary
Pros: Higher ROI optionality across non-Amazon networks. New agentic inventory channels opening. Consolidated in-store media infrastructure. GLP-1 basket data offers a measurable demand-side signal.
Cons: Attribution drift as AI mediates discovery. CPM convergence between digital and physical surfaces. Measurement transparency depends on retailer-side disclosures of limited studies. In-store consolidation may reduce vendor competition in specific regions.
Track through Q3: OpenAI SDK and checkout integration changes, Kroger Precision Marketing and Instacart Ads methodology disclosure, Vusion inventory availability and minimum spend thresholds, and category-level GLP-1 basket data.