Editorial Team Despite all the talk about precision targeting, a recent IAB report just dropped a bomb: only 42% of us actually feel good about measuring ROI on our Connected TV (CTV) advertising. That level of doubt, even among top-tier marketing teams, shows just how deep the measurement challenges run and makes you wonder if we really have any clue what our CTV spend is doing.
Key Takeaways
- We’re overstating CTV reach. Nielsen data shows the real incremental reach is 10-15% lower than what platforms claim because they aren’t deduplicating devices properly.
- Bad attribution models are a huge problem. If they don’t get household viewing right, they can misattribute up to 30% of conversions, which completely messes up your budget.
- A whopping 60% of advertisers can’t get their data straight across different CTV platforms. This means fragmented reports and no real picture of what’s working.
- We’re still stuck on last-touch attribution for CTV, a mistake that undervalues all the brand-building work happening at the top of the funnel and pushes money to the wrong places.
45% of CTV Ad Spend Is Attributed to the Wrong Device or User
That 45% figure from an eMarketer study gets right to the heart of the problem: the absolute mess of device graphs and user ID. When you’re running campaigns across a dozen services like Hulu, Peacock, and YouTube TV, every platform has its own secret sauce for identifying who’s who in a household. This gives you a totally broken picture of the customer journey, so you can’t confidently say the ad someone saw on the living room TV led to the purchase their kid made on a smartphone minutes later. In practice, we’re in a constant fight with identity resolution, trying to tape together data from all these walled gardens. Without some kind of universal ID or at least a decent, privacy-first way to match devices, marketers are mostly just guessing which ads actually led to a conversion. This fundamentally misinterprets where your budget is having an impact, going way beyond just a few lost impressions.
Only 38% of Agencies Can Provide Unified Cross-Platform CTV Measurement
An IAB survey on the state of CTV recently found that only 38% of agencies can actually deliver unified cross-platform measurement. This number shows a massive gap in what the industry can do. It’s one thing to run the ads. It’s another thing entirely to integrate those CTV performance numbers with data from search, social, and even linear TV. Most agencies are still structured in silos. The CTV team is in Innovid or Adform, the search team lives in Google Ads, and the social team never leaves the Meta Business Suite. The real breakdown happens when you try to pull all that data into one coherent view for a true apples-to-apples comparison. I’ve seen it a dozen times: a brand’s CTV campaigns have fantastic video completion rates, but when you try to connect that to an actual sales lift, the data’s a ghost because the measurement systems don’t sync up. The problem here is more organizational and methodological than it is technological. Agencies must hire data scientists and integration people who can actually connect these sources instead of just handing clients a stack of platform-specific PDF reports.
62% of Marketers Still Primarily Use Last-Touch Attribution for CTV Campaigns
Statista reports that 62% of us are still using last-touch for CTV campaigns. That’s a genuinely alarming number. Using last-touch as your main model for CTV is a basic error, even if it has some utility in simpler digital channels. CTV is almost always a top- or mid-funnel play that’s great for building awareness and driving consideration. An ad watched on a huge screen in the living room isn’t meant to get an immediate click and purchase like a search ad would. It’s building brand affinity and planting the seed for a later purchase. By giving all the credit to the final interaction before a conversion, you’re effectively ignoring the huge contribution of your CTV budget. This leads directly to bad budget decisions, with marketers pulling money out of perfectly good CTV campaigns because the immediate, last-touch ROI looks weaker than channels that naturally own that final click. We’ve got to move to multi-touch attribution models that can assign credit properly across the whole journey, giving CTV its due for influencing decisions early on. Tools like Attribution App or setting up custom data clean rooms are becoming non-negotiable for getting this right.
Only 25% of Advertisers Are Confident in Their Ability to Measure Incremental Reach from CTV
A recent Nielsen report found only a quarter of advertisers are confident they can measure incremental reach from CTV. This reveals a huge blind spot: are we actually reaching new people, or just hitting the same ones over and over? CTV’s big promise is its surgical targeting, which should expand a brand’s reach beyond what linear TV can do. But if your CTV ads are just hitting the same households that already see your linear spots or click your digital ads, the extra value is pretty low. The real work is in deduplication, figuring out how to identify unique households across linear, CTV, and digital without just trusting panel data that might not match your audience. Platform reports will give you a “reach” number, but it’s based only on their own little world and completely ignores overlap with your other media buys. This job requires serious audience measurement tools and, often, a third-party verification partner who can give you an honest look at your actual incremental reach. Without it, brands are just lighting money on fire by over-saturating audiences they already own instead of finding new ones.
Challenging the Conventional Wisdom: The Myth of the “Unified CTV Dashboard”
Everyone is chasing the dream of a single, “unified dashboard” that solves all these problems. I think that’s a pipe dream, and a dangerous one at that. The CTV world, with its walled-off data, different measurement rules, and fierce platform competition, makes a truly unified dashboard practically impossible. Even when a vendor swears they have the solution, it usually means you have to compromise on something big, like data granularity, real-time reporting, or the ability to customize your attribution logic. Instead of waiting for some magic-bullet dashboard, a smarter move is to focus on your own data plumbing and analytics. That means you should be investing in a data warehouse like Google BigQuery or Amazon Redshift, and then putting business intelligence tools like Microsoft Power BI or Tableau on top to build your own reports. The point isn’t to get one pre-built dashboard. It’s to build the capability to pull data from everywhere into one place where your own team can run the numbers and get real answers. This approach accepts the fragmented reality of the market and lets brands create their own measurement system that fits their business, rather than waiting for a solution that’s never coming.
The mess of CTV measurement isn’t going away. It’s just changing shape. To actually understand performance and spend efficiently, marketers have to get past simple metrics and build a real measurement strategy that accounts for cross-device paths, incremental reach, and smarter attribution. Understanding these details is key for anyone trying to optimize their paid media, especially as the automation of Google and Meta Ads gets more complex. And going forward, AI agents will be absolutely necessary for fixing reporting accuracy and working through this difficult measurement field.
What is cross-device attribution in CTV?
It’s about connecting an ad someone saw on their TV to something they did later on a different device, like their phone or laptop. This is really hard and depends on being able to tie all those devices back to one household or person, a process called identity resolution.
Why is incremental reach difficult to measure in CTV?
It’s tough because there’s no single ID that works across all the streaming apps and linear TV. To figure out who you’re *really* reaching for the first time, you have to perform some pretty complex deduplication to sort out the audience overlap between your different channels.
What are the limitations of last-touch attribution for CTV?
Last-touch gives 100% of the credit to the very last click before a sale. That model completely misses the point of CTV, which is usually about building awareness at the start of the customer journey. You end up undervaluing your CTV campaigns and cutting budgets that are actually working.
How can marketers improve CTV data integration?
The best approach is to pull all your data from the various platforms into a central spot, like a data warehouse or data lake. You can use APIs to automate this data extraction and then use business intelligence (BI) tools to build your own reports for analysis, creating a single source of truth.
What role do data clean rooms play in CTV measurement?
Think of them as a secure sandbox. You can put your first-party customer data in one, and a publisher puts their audience data in, and the clean room lets you see where they overlap without either side having to share raw, personally identifiable user data. It’s a privacy-safe way to do better targeting, cross-platform measurement, and advanced attribution.
