The hidden cost of convenience: What Amazon’s DSP means for advertiser control

As programmatic advertising has grown more complex, marketers have learned to love simplicity. Since launching in 2012, Amazon Advertising’s DSP has aimed to answer that call by becoming more expansive, integrated, automated and self-contained. Simple can be appealing for marketers.

As one ad tech executive told me: “When running a campaign, if you take into account all the platforms, channels and ad formats, you end up with trillions of permutations.”

And yet, as Amazon’s DSP evolves, it increasingly resembles the kind of fully developed walled garden long associated with Google’s DV360. That evolution raises important questions about control, transparency and long-term leverage.

Easy doesn’t always mean better.

As convenience increases, advertisers may find that they are ceding more control than they initially realized. Eventually, long-term success will favor those advertisers that maintain data transparency and can operate across ecosystems — pairing the strengths of walled gardens with the scale of the open internet.

Brick by brick

Amazon’s DSP was built on commerce data. Deterministic purchase signals, proximity to transactions and closed-loop measurement created a compelling value proposition for marketers eager to see what drives sales.

From there, Amazon expanded its DSP beyond its retail core. Connected TV (CTV) inventory in Prime Video, along with off-Amazon display and video, extended its reach. Amazon was no longer just a retail performance engine; it was positioning itself as a full-funnel media platform. Adding AI into the mix, the system increasingly manages itself.

This growth journey mirrors that of Google, which started with similar intent data coupled with integrated media activation. Google then expanded from search into YouTube, display and CTV. What began as performance advertising matured into a comprehensive cross-channel environment. Now, with the rise of AI, automation is the default operating mode. Manual levers give way to machine-led decisioning, promising efficiency at scale.

The walled garden facade

For marketers, both Amazon and Google both offer efficiency as well as effectiveness: built-in audience creation, streamlined path from impression to outcome and faster validation of return on ad spend.

But the underlying trade-offs are there as well. Just like Google, Amazon’s data largely stays inside the ecosystem. Visibility into auction mechanics, cross-platform comparability and data portability are limited. Marketers increasingly relied on Amazon’s measurement logic rather than stitching together independent views of performance. The dependency is real too.

As optimization decisions span multiple channels within one ecosystem, transparency is diminished. Budget allocation, inventory prioritization and cross-channel trade-offs increasingly rely on automated platform logic. It can become harder to isolate what truly drove performance — or what did not.

A structural dynamic is at play. Any vertically integrated platform must balance advertiser outcomes with ecosystem growth. That tension is not unique to Amazon, but it is inherent in comprehensive walled gardens.

Automation introduces another challenge. When AI systems make decisions across bidding, placement and creative combinations, the “why” behind outcomes can become harder to establish. If performance disappoints, it may be more difficult to challenge the logic or extract transferable learning. The learning loop becomes platform-centric rather than advertiser-centric, which is why understanding the logic behind the system is important.

Efficiency rises. But interpretability can decline.

Ecosystem diversification wins

While integrated ecosystems can drive performance and reduce complexity, the concentration of execution inside a single environment inevitably centralizes insight, data and decision logic as well. It’s all self-contained.

The advertisers best positioned for long-term success are those that operate across ecosystems — combining the strengths of walled gardens with the reach and transparency of the open internet.

Diversification mitigates risk. It prevents overreliance on a single measurement framework or optimization logic. It enables comparative intelligence: Performance can be evaluated across environments rather than within a single one. It preserves leverage in negotiations and planning.

The open internet offers scale beyond logged-in environments, greater transparency into auction dynamics and the flexibility to integrate independent data and measurement solutions. Walled gardens offer closed-loop effectiveness and operational simplicity. These strengths are complementary.

Amazon’s DSP has become a formidable one-stop shop. For many campaigns, it will be a central growth engine. But no single ecosystem should become the sole operating system for a brand’s media strategy.

This op-ed represents the views and opinions of the author and not of The Current, a division of The Trade Desk, or The Trade Desk. The appearance of the op-ed on The Current does not constitute an endorsement by The Current or The Trade Desk.