Dataline column flags programmatic opacity as Trade Desk changes partner payments

Dataline published a new MartechView column by CEO Paul Sobel arguing that programmatic advertising remains opaque and that The Trade Desk’s new incremental payment model could deepen that problem. The piece says marketers should demand raw data and more independent verification as AI-driven buying tools could reshape the ad-tech stack within about two years.

Why it matters: - Programmatic advertising still leaves marketers unable to see where a meaningful share of their media dollars goes. - The column argues that the latest platform changes could further concentrate control over pricing, measurement and verification inside the DSP. - The article says marketers who accept opaque reporting may keep losing visibility into campaign performance and costs.

What happened: - Dataline published a new MartechView column by Paul Sobel, CEO of Dataline, titled “How Long Will Marketers Settle for Opacity in Programmatic?” - The column examines marketers’ tolerance for campaign transparency gaps across data and buy-side platforms. - Sobel cites a LinkedIn post from Michael Learmonth and says early digital media research found fewer than half of ad dollars spent on premium sites reached those sites. - The column also points to recent UK ISBA research showing just over 51% of spend reaching publishers. - The piece says up to 15% of media buys can disappear during campaign flights. - The column says The Trade Desk plans to change how it pays audience partners through a new “incremental” model.

The details: - Under the new model, payments will be tied to the incremental value a partner’s identity data adds when that data is not already captured elsewhere in the DSP system. - The shift is intended to reduce payment for duplicate data and reward data that appears differentiated. - The column says the model could let The Trade Desk favor one audience provider over another based on its own algorithmic view of incrementality. - In a B2B example, two providers selling similar “Cloud Security Decision Makers” audiences could be paid differently if The Trade Desk’s system says one audience overlaps with existing data. - The article says that outcome could penalize a provider even if its data converts better for an advertiser. - The Trade Desk has promised new tools, including APIs and scoring mechanisms, to explain how data is evaluated under the revised model. - Those tools are not yet available. - The article says advertisers have long complained that digital platforms charge fees while requiring buyers to rely on internal verification tools. - The column says the ANA has long claimed ad fraud can reach $80 billion a year in digital video.

Between the lines: - The column frames the new payment model as a control issue, not just a performance issue. - It argues that platforms may be defining value in ways outsiders cannot audit. - The piece suggests the industry’s transparency debate is also about who captures the unexplained share of spend. - Sobel says marketers should demand log-level data, test data partners directly through PMPs or data clean rooms, and diversify spend beyond major DSPs. - The article predicts enterprise brands will use proprietary AI buying tools that connect directly to data sellers and publishers, reducing the role of traditional DSPs.

What’s next: - The Trade Desk is expected to roll out the APIs and scoring tools it has promised, but the release timing remains unclear. - The column says buyers will keep questioning opaque economics if the new model is applied without outside validation. - The piece forecasts that AI-based buying workflows could start disintermediating traditional ad-tech layers in as little as two years. - The article says that shift could move brands away from hidden fee structures and toward direct machine-to-machine audience buying.

The bottom line: - The column argues that marketers no longer need to accept platforms grading their own homework, especially as new payment models and AI buying tools reshape programmatic economics.