Can media companies ever claw back advertising revenue from big tech?
Adstra CEO Rick Erwin explained why open web ad products face a structural disadvantage.
For years, publishers have told themselves some version of the same story: if they could improve targeting, automate more of their ad sales, and make inventory easier to buy at scale, they might begin reclaiming some of the advertising dollars that have flowed to big tech. That belief animated much of the optimism around programmatic advertising and sustained the idea that publishers could eventually narrow the gap with platforms like Google and Meta. Yet the gap has only widened, with Amazon now emerging as a third major force. That raises a harder question: whether publishers have been trying to solve a structural problem with tactical tools.
That question framed a recent conversation I had with Rick Erwin, CEO of Adstra, whose career has spanned the evolution of identity resolution, data-driven marketing, and modern ad tech. Erwin does not dismiss the progress the industry has made, but he argues the dominant platforms built advantages that go much deeper than publishers originally understood. Better targeting alone was never going to erase those advantages. The bigger issue, in his telling, is that the platforms combined superior data, better-performing ad products, and simpler buying systems into a reinforcing flywheel that has been extraordinarily difficult for the open web to match.
In our interview, Erwin explained why tech platforms have built such a durable advantage over publishers, where publishers could still add value, and why he believes AI could fundamentally change how ad campaigns are planned and executed.
Let’s jump into it…
Why Better Data Still Wins
If there is a core thesis running through Erwin’s argument, it is that the platforms’ dominance begins with the depth and quality of their data. “When they work better, they work better because the data is better,” he said. His point was not merely that companies like Google and Meta possess more user data than publishers. It was that they possess a unified and constantly compounding view of consumer behavior that is extremely difficult to replicate in the fragmented environment of the open web.
Erwin illustrated that advantage with a story from his early interactions with Facebook. He recalled an engineer trying to explain the sheer volume of signals the platform could use to predict behavior. For instance, Facebook is able to track a mobile user’s “four-finger scrolling speed,” a tiny behavioral detail that nonetheless reveals the granularity at which the platform could model users. The anecdote is striking because it captures the scale of the asymmetry. Publishers may have valuable first-party signals, but they rarely have access to that kind of behavioral depth, let alone the infrastructure to operationalize it at comparable scale.
That, in Erwin’s view, is why the original promise that programmatic would level the playing field was always incomplete. It addressed access to targeting, but it did not solve the underlying imbalance in consumer intelligence. Publishers gained tools to make their inventory more addressable, but they did not suddenly acquire the same quality of data as the platforms they were trying to compete against. That distinction matters because in performance advertising, even small improvements in prediction can produce enormous advantages when multiplied across billions of impressions.
Why Programmatic Didn’t Restore Publisher Economics
Erwin resists the idea that programmatic was a failure. In his telling, it succeeded in making sophisticated advertising tools more accessible to brands. “Many brands that would have felt the need to go through an agency no longer need one,” he said, arguing that demand-side platforms and related infrastructure democratized access to data-driven buying. That achievement is real. But it did not produce the broader economic reversal publishers once hoped for.
Part of the problem is that programmatic solved one problem—making inventory easier to transact—without solving the structural disadvantages that continued to favor the platforms. Publishers still operate in a fragmented ecosystem where identity, audience signals, and measurement are distributed across countless sites and intermediaries. By contrast, the platforms control both advertiser demand and the environments where ads are delivered. That closed-loop structure allows them to optimize in ways the open web often cannot.
Erwin repeatedly returned to what he called the “black box” problem to explain why that imbalance persists. Platforms increasingly ask marketers to trust systems they do not fully understand. He pointed to products like Google’s Performance Max, where advertisers can hand over budgets, creative assets, and goals while algorithms determine much of the targeting and optimization. “The black box just did what the black box does,” he said. “And it’s fairly inscrutable.” That opacity creates discomfort, but it has not prevented marketers from continuing to spend because the systems often deliver results. For publishers, that has been a difficult reality to compete against.
The Advantages Publishers Still Haven’t Matched
In our interview, I argued that platforms’ edge extends beyond data. Another advantage lies in ad format itself. Even excellent targeting can only do so much if the underlying ad unit is weak. Publishers have long struggled with the limitations of display advertising, while many of the platforms built performance engines around ad products that feel native to the user experience. Search ads align directly with intent. Sponsored social posts blend into feeds. Commerce ads sit close to transactional behavior. Those formats often outperform traditional display not simply because they are better targeted, but because they are structurally better ad products.
A second advantage is simplicity. During the interview, I argued that one of the platforms’ greatest strengths is “easy self-service advertising,” where even a small business can begin spending modestly on a platform like Google and understand how the system works. Meanwhile, many smaller advertisers will never touch programmatic because the barrier to entry still feels too high. Erwin mostly agreed with that point.
This matters because simplicity is not merely a convenience; it is a moat. The easier it is for millions of small advertisers to enter a system, the more demand that system aggregates. That dynamic has helped the platforms build enormous advertiser bases that publishers, even collectively, have struggled to attract. In that sense, publishers have often been competing not only against superior technology but against a fundamentally more accessible business model.