Identity Is CTV's Most Expensive Guess
Better identity in CTV means less inferring, not just more data. A stronger, consented signal before the app opens can improve measurement, monetization, and the ad experience for everyone.
Turn on a smart TV and the operating system does something the digital advertising business spends heavily trying to manage - it creates a direct relationship with the household. Before a single app opens, someone creates an account, selects a profile, and makes choices. That relationship isn’t built from a third-party graph; it starts when the household engages directly with the TV.
Yet, the way many of today's operating systems and apps interact forces the ad market to reconnect with that same household through inference. It takes an IP address, runs it through a model, and hopes the match is good enough for targeting, measurement, and frequency control. Having spent two decades building platforms across streaming, search, and TV, I see this disconnect as a missed opportunity for advertisers and the industry. CTV has strong identity signals; the challenge is carrying the strongest of those from setup to the apps, ads, and measurement systems that rely on it. That challenge only grows as more media investment flows into CTV. Get identity right and CTV doesn't just measure better, it becomes premium.
As CTV investment grows, identity has more work to do
The issue becomes more evident as the market gets bigger. Truthset's 2026 State of Data Accuracy report estimates that roughly 40% of open programmatic CTV spend is wasted on inaccurate identity data, or about $7.4 billion this year in a market projected to reach about $38 billion in 2026, according to eMarketer. There needs to be a better option.
The stakes rise further as CTV media planning shifts into programming like live sports. Once dollars are in motion, identity shapes the answers to the basic questions advertisers care about: which households they reached, how often, and whether those impressions can be measured with confidence.
A separate study from Adstra and InterMedia found that only 23% of residential IP addresses matched their intended geographic target. Each piece of research measures different parts of the problem, but they point to the same conclusion: CTV is scaling on top of an identity layer still built heavily around inference, and that design deserves a closer look.
The strongest identity moment comes early
A graph can improve a match, but it cannot change the context of how the original relationship was created. The account, profile, device relationship, and consent state all begin at the platform level. When the device relationship falls out of the process, the market ends up trying to recreate it later with less reliable signals.
Early streaming offers useful lessons. When login and profile were treated as foundational, recommendations, retention, and measurement all improved because the system had a clearer sense of the customer relationship. CTV can apply that lesson at the operating system level, where identity can support discovery, advertising, and measurement earlier in the journey.
The pieces are closer than they look
The trouble is that the process still asks too much of everyone. Publishers often need viewers to sign in app by app, viewers repeat the same steps across services, and the market still must piece together a relationship the TV already started. A consented OS-level household signal could make that cleaner. For viewers, that could mean fewer repeat ads, more relevant messages, and less friction overall.
Those improvements can help create the kind of experience consumers associate with premium environments. Research from The Trade Desk Intelligence and PA Consulting's 2025 report, The Premium Payoff, found that 85% of respondents say premium environments increase their trust of advertisers, while consumers view brands advertising in premium CTV environments as 3.2X more innovative and 2.6X more successful.
And that is why this conversation is bigger than the home screen. The home screen influences what gets watched. Identity influences how much that viewing is worth. Stronger identity does not just improve measurement, it can help create the trust and quality signals that make premium environments more valuable to both advertisers and viewers.
The Relationship Already Exists. Build on It.
The practical ask follows: platforms and OEMs should make it easier for the identity relationship that starts on the TV to carry throughout the viewing journey. Publishers should expect clearer signals from the platforms that help them reach audiences. And advertisers should value identity that holds up, because cheaper impressions lose value when the underlying household match is uncertain. The payoff is concrete. Recover even part of the $7.4 billion now lost to inaccurate identity data and publishers can earn more for the same audiences, advertisers can pay only for households they actually reached, measurement becomes trustworthy enough to justify bigger budgets, and viewers stop seeing the same ad ten times in one night.
The first decade of CTV was about bringing streaming to the TV. The next will be shaped by how well the industry builds upon the relationship the TV already creates. The work ahead is real, but the direction is clear: build identity around where that relationship begins. That's how CTV becomes a more trusted and valuable version of television.
