Every industry develops narratives that begin as useful simplifications and eventually become accepted truths. Advertising is no different. But the problem is that many of these narratives survive long after the market has stopped validating them.
The industry says streaming television is becoming a performance channel, then acts surprised when television behaves like television and does not provide immediate direct response gratification. Companies position themselves as alternatives to walled gardens while consolidating more of the advertising stack under their own control. Billions of dollars continue flowing into identity infrastructure even as evidence mounts that much of the underlying data is way less accurate than the industry would like to admit.
None of these observations is a secret. Most people reading this have probably witnessed some version of them firsthand. But somewhere between what the industry says and what the market rewards is often where the truth lives.
Here are the three industry storylines most in need of debunking.
Television optimized into a search box
The advertising industry increasingly talks about connected television in the language of paid search and social. Outcomes-based buying, performance optimization and return on ad spend have become the dominant vocabulary of the channel.
For example, MNTN bills itself as “a platform that turns connected television into a performance marketing channel and enables campaigns similar to paid search and social.”
The appeal is obvious. If television can become directly accountable, marketers get the influence of TV with the measurability of digital.
The problem is that television and search solve fundamentally different problems. A consumer watching a streaming drama on a Tuesday evening is in a different state than someone actively entering a query into a search engine. One is consuming content while the other is expressing intent.
Television’s influence is cumulative and delayed, unfolding over weeks and months in ways that cannot be conveyed in a neat trend line. Yet television is being judged against compressed attribution windows built for channels designed to capture existing demand rather than create future demand. The result is predictable. Advertisers fail to see immediate signals, conclude television underperformed and reduce investment in a channel that may be doing exactly what it was designed to do.
The answer is not less measurement but a more realistic understanding of what different measurement approaches can and cannot account for. Marketing mix modeling, incrementality studies and attribution all provide useful perspectives. A channel that builds memory deserves measurement capable of observing memory.
Independence packaged into a new garden
The open internet sells itself as the antidote to the walled gardens. The logic is straightforward: Owning inventory creates conflicts of interest, so independent platforms provide a more neutral alternative.
Today, major social and search platforms own supply, control measurement and set the rules. Their metrics have been questioned, restated and, in some cases, materially overstated. Meta famously acknowledged inflating video metrics for two years. Advertisers noticed and continued spending because the buying remained easy, the testing remained cheap and the results remained sufficiently defensible.
Meanwhile, the industry’s definition of platform independence has become suspiciously convenient. A company can own no content whatsoever while controlling identity, measurement, optimization, supply access and content classification. Viant added IRIS.TV for content classification, Lockr for identity and TVision for measurement. The Trade Desk built OpenPath for supply access, Unified ID 2.0 for identity, Kokai for optimization and Ventura as a CTV operating system.
None of these initiatives are inherently problematic on their own. But they reveal where the market is heading. Many of the industry’s largest independents now influence the same decisioning layers they once criticized others for controlling. Firms that built their brands as alternatives to walled gardens are, deal by deal, assembling gardens of their own.
Whether a platform owns media has become a far less interesting question than whether buyers can independently understand how decisions are being made. If advertisers cannot audit the logic behind optimization decisions or evaluate the quality of the underlying data, neutrality becomes difficult to distinguish from opacity.
A precise measurement of almost nothing
Corporate strategies assume that whoever controls the strongest identity layer controls a decisive competitive advantage. This conviction often exceeds the evidence.
In 2025, Truthset and CIMM benchmarked nearly one billion IP records from six major data providers against verified ISP and MVPD truth sets. IP-to-household linkages were accurate roughly 13% of the time. IP-to-email linkages were accurate roughly 16% of the time.
Advertisers want to know who they are reaching. The industry built increasingly sophisticated systems designed to answer that question, then began competing on the answer itself. Publicis has spent roughly $4 billion in 14 months on identity and data assets. WPP paid around $150 million for InfoSum.
A 90% match rate against a 13% accurate graph is a precise measurement of almost nothing. The industry knows this but continues investing.
Clean rooms are often presented as a solution to the matching problem, but clean rooms assume the signals feeding them are accurate. At sufficiently high error rates, they simply share flawed data more efficiently.
And scale does not repair a broken ID linkage. It replicates it. You reach the wrong household with greater confidence.
The fix? Demand error-rate disclosure from identity vendors the same way you demand viewability from publishers. If a vendor cannot tell you how often it is wrong, that’s all the answer you need.
None of the above means these systems are useless. Television should be measured. Identity matters. Independent platforms are different from walled gardens in meaningful ways. The problem begins when useful signals are taken as universal truths, and assumptions stop being challenged.
The healthiest industries are willing to revisit their assumptions, especially when commercial concerns encourage us not to question the prevailing storylines.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Irina Dzyubinsky, David Nyurenberg and AdExchanger on LinkedIn.
