Category: In the News

Trust remains the fault line running through Amazon’s ad ambitions.

It has been there from the start. Marketers learned early that partnering with Amazon meant buying media from a company that not only set the marketplace’s rules but also stocked the shelves and sold competing products. The tradeoff was efficiency. The tension never disappeared. 

In fact, it has intensified as Amazon pushes its ad business beyond its own walls and deeper into the broader web. The pitch is performance at scale. The concern is whether those ad dollars are being steered to the best outcomes or toward the parts of the ecosystem Amazon controls. It’s the familiar strain of any platform that sells ads inside their own gardens while extending their reach across the open internet. Google faced these suspicions. Amazon is confronting it now, at a moment when automated buying and real-time optimization send more dollars through tech as influential as the Amazon DSP.

Continue reading this article on digiday.com. Sign up for Digiday newsletters to get the latest on media, marketing and the future of TV.

,Read More

Susan Wu, associate vice president of marketing research, PubMatic

After decades of fragmentation punctuated by episodic roll-ups, digital advertising has entered a different kind of consolidation phase: one driven by AI and supply path discipline, not just M&A. As marketers push for fewer hops, lower latency and transparent economics, and publishers seek durable yield with less operational drag, the path to a unified, two-sided tech fabric is finally practical.

The cost of fragmentation

Buyers today often operate between eight and 12 disconnected point solutions: DSPs, verification vendors, fraud detection, attribution and creative optimization, to name a few. Publishers mirror this complexity with separate yield tools, header bidding wrappers, identity solutions and reporting dashboards that rarely agree with one another. Each handoff introduces latency, data loss and fee opacity. When something breaks, troubleshooting means stitching together logs from half a dozen vendors with incompatible taxonomies.

The real breaking point is strategic. As AI becomes essential, fragmentation caps performance. Machine learning requires unified data. When impression or viewability means different things across systems, intelligence can’t compound. This is why 88% of marketing teams are using generative AI in production, experimentation or exploration environments, and why buyers allocate 53.7% of media budgets to supply path optimization efforts (according to a 2025 “Brand Perception Study” commissioned by PubMatic and conducted by Forrester). But SPO alone only fixes routing, not the underlying infrastructure problem.

Continue reading this article on digiday.com. Sign up for Digiday newsletters to get the latest on media, marketing and the future of TV.

,Read More