Amazon DSP: What Brands are Actually Doing

TL;DR: Despite industry narratives around full-funnel activation, AI-driven optimization, and advanced measurement, most brands still run relatively immature Amazon DSP programs. Our research suggests optimization cadence is infrequent, AMC adoption remains shallow, and AI experimentation has yet to consistently prove value. We dive deep in our report to examine why and explore where competitive edge opportunities exist.

1 minute read

Walk any retail media conference and you'll hear that brands are running Amazon DSP as a full-funnel platform, optimizing it continuously, and putting AMC and AI to work.

We surveyed 95 consumer brands, ran a peer share group, hosted an AI-on-DSP webinar session and revisited our 2026 State of Retail Media data to see if those things were true.

Surprisingly, we found the gap between what the industry says and what brands actually do to be quite wide. Wide enough, in fact, to offer a competitive edge opportunity.

A few of the findings:

• DSP is largely set-and-forget: Only 27% of brands optimize weekly, and 73% optimize monthly or less. The platform is too manual to manage well by hand, so most programs just roll forward unchanged.
• It's still a lower-funnel tool: 42% cite retargeting as their strongest use case; just 15% lead with upper-funnel awareness. Amazon is being treated as an extension of search, not a media platform.
• AMC adoption is shallow four years in: 37% don't use it or aren't sure how, and even brands that have dug in struggle to turn its outputs into decisions they'd actually make differently.
• AI is widespread but unproven: 63% have tried GenAI in retail media; only 25% see clear results.

Our report also covers where AI agents are starting to close the manual gap, the structural reason DSP gets neglected, and a self-assessment worksheet to benchmark your own program against the data.