For every dollar a brand allocates to media spend, approximately 30 cents actually buys media. The rest disappears into what the industry politely calls the tech tax and practitioners come along for the ride. That figure is a structural indictment of how media buying still operates in a “black box”, and it is the problem Elle Shnayder Donner has spent her career working around.
An advertising executive with deep roots in premium television sales and programmatic execution, Donner recently launched Skye Consortium, a collective of industry experts drawn from supply-side platforms (SSPs), demand-side platforms (DSPs), agencies, and publishers. The initiative was built to help chief marketing officers (CMOs) navigate an increasingly fragmented advertising technology landscape, where the cost of making the wrong technology decision is significantly higher than most brands realize. “20% of my media spend drives 80% of my sales, I just don’t know which 20%” Donner reflects, invoking one of the industry’s most enduring paradoxes. “And we’re still answering that question today.” The right infrastructure makes that problem easier to begin to solve.
Programmatic Does Not Lack Control, It Lacks Infrastructure
The persistent assumption that linear TV offers superior placement control does not hold up to scrutiny. Buying a specific time slot in linear, the first ad break after 8:02 PM, for example, requires expensive manual negotiation over the phone and still does not guarantee the outcome. Programmatic, by contrast, enables more precise audience targeting across a far wider range of supply options and media types. The only gap is show-specific buying, and Donner is candid about why that gap matters less than it appears: buying a specific show title when the target audience is not watching it is counterproductive to the entire logic of audience-based buying.
The real problem is not a targeting gap. It is an infrastructure gap. When brands layer ad tech tools without a coherent strategy, adding whatever the hot platform of the moment is, they create expensive compatibility issues. A data vendor that does not work with the chosen DSP. An SSP buying structure that was never evaluated against direct alternatives. Technologies that do not communicate with each other waste both resources and signals.
Skye Consortium was built to address exactly this. The firm helps clients evaluate their ad stack components, assess market-rate pricing to prevent overpaying, identify compatible data vendors and data management platforms (DMPs), and navigate barriers to entry that even organizations with CMO-level authority face. Having a CMO title, Donner notes pointedly, does not guarantee access to major platforms like The Trade Desk. Understanding how those relationships work, and how to build them, is what separates brands that scale their media investment from brands that burn it.
Automation Needs Human Judgment More Than It Admits
The TV upfronts still exist. They still fill stages with discussions of content, programming, and new ad-break formats. Programmatic gets minimal stage time, not because it lacks relevance, but because the industry understands that advertising remains a fundamentally relationship-driven business. Pipes talking to each other is not the same as people talking to each other, and no level of bidding automation closes that gap.
Donner applies the same logic to AI-driven planning and execution. The errors are real and frequent: tools producing irrelevant or incorrect outputs, automated reports requiring human review before they reach clients, and innovation requiring human-secured investment that no AI system can fund independently. The discipline that matters is not whether to use AI. It is maintaining the ability to override it when the output is wrong, the judgment to identify when that is happening, and the checks that catch errors before they reach clients or markets.
“There’s got to be checks and balances when using AI,” Donner states. The organizations building a durable competitive advantage in this landscape are those that treat automation as a capability multiplier rather than as a replacement for the expertise that makes the multiplier work. Without that expertise, the 20 cents actually buying media gets smaller, not larger.
Follow Elle Shnayder Donner on LinkedIn for more insights on programmatic strategy and ad tech infrastructure.
