How programmatic fails in the absense of clear KPIs and a measurement-first strategy

Programmatic buying holds the allure of effortless efficiency. With the promise of automated, data-driven decisions at scale, it’s tempting to adopt a "set-it-and-forget-it" mindset. But this hands-off approach often masks a landscape riddled with waste, from MFA sites to convoluted supply paths that erode budgets before a single impression is truly seen.

We sat down with Vlad Chubakov, Associate Director of Programmatic at Delve Deeper and author of the educational series Programmatic101. For Chubakov, the biggest mistake in programmatic advertising happens long before a campaign goes live. He argues that launching without a clear measurement framework isn't just inefficient, it's a fundamental failure of strategy.

• Baby steps, backwards: "If you don't have a plan for how you're going to measure a campaign, and you try to solve for it after the campaign ends, that's a failure in my opinion." Without a predefined plan and concrete KPIs, Chubakov explains, it becomes impossible to prove success, determine clear next steps, or meaningfully evaluate a campaign's outcome. The antidote lies in a mindset shift that prioritizes a continuous, iterative process over a single, automated solution.

"The whole process is about taking small incremental steps and working backwards," Chubakov said. "You start by understanding how you're going to measure success, then you change things on the execution side, find more of those small wins, and scale them to achieve bigger results."

• Curating the curated: This philosophy extends even to the industry's most trusted tools. Pre-vetted inventory packages, like The Trade Desk's SP500+ list, are a valuable starting point for screening out low-quality MFA sites. But according to Chubakov, they are not a complete strategy in themselves.

"You still need to curate this curated list." That involves moving beyond simple performance metrics and into the nuanced territory of brand values. "It's not only about performance; it should be aligned with brand values," he explains. "If you are selling products for kids, you don't want to appear on news articles about distressing events."

• The reseller problem: A core part of working backwards from measurement is ensuring efficiency, which means tackling the web of intermediaries that plague the supply chain. "This is one of the biggest problems in ad tech," Chubakov warned. "We have a lot of intermediaries or resellers who are just reselling the same inventory and applying their fees on top, making the whole thing significantly more expensive for buyers." A measurement-first approach demands auditing reports to cut inefficient suppliers and using direct-to-publisher tools to ensure you’re not paying multiple middlemen for the same impression.
• Streaming's transparency gap: "In CTV, transparency is one of the biggest issues that prevent the category from growing even further," Chubakov says. "You sometimes miss important things like show-level targeting because sellers simply don't pass it via the bid request." This lack of programmatic transparency has a direct and significant consequence on buying behavior. "As a result, a lot of buyers are still buying CTV and streaming via guaranteed deals," he notes. "But it's not ideal because it involves a lot of communication. It's more like an old-school, non-programmatic way to get your inventory."

Ultimately, tools like curated lists are an excellent first step on the path, not the destination. The real work—and the real value—lies in applying a rigorous, measurement-led, and incremental process. In programmatic, the most valuable strategies are architected, not automated.