As CTV continues to attract more viewers—and more media dollars—it’s become clear that audience accuracy is essential to effective advertising. At the core of that identity layer sits the IP address: the connective tissue between content delivery and the household receiving it. But how accurate are those linkages?
As traditional identity signals like third-party cookies and mobile ad IDs fade, IP addresses have taken on greater responsibility, especially in CTV, where they serve as the default match key for targeting and measurement. Yet unlike cookies or emails, IPs are inherently fluid. They shift with device changes, network resets, and household behaviors. That makes it difficult to pin down.
Advertisers investing heavily in CTV have to ask themselves how confident they are in the ability of IP data to connect them with the audiences they seek.
Coalition for Innovative Media Measurement (CIMM) and GoAddressable commissioned Truthset to answer that question by conducting the industry’s first large-scale IP linkage accuracy study. The findings highlight a complex, nuanced challenge that the industry must address together.
The Study
Truthset benchmarked nearly one billion IP records from six major data providers against an accurate database of IP-address-to-household assignments provided by two leading Internet Service Providers (ISPs) and one leading Multichannel Video Programming Distributor (MVPD). This ISP data served as ground truth, reflecting verified IP assignments at the household level during specific timeframes.
The study focused on two common linkage types:
- IP-to-postal address: to connect digital signals to households
- IP-to-email address: to tie online behaviors to individuals
The goal wasn’t to evaluate any one vendor’s method, but rather to understand the overall state of IP accuracy and the factors influencing it.
The Findings
The results provide a clearer picture of the CTV identity landscape and offer direction for where the industry can improve:
- IP-to-postal linkages are accurate, on average, 13% of the time.
- IP-to-email linkages are accurate, on average, 16% of the time.
- Data providers agree only 6.4% of the time on IP-to-postal linkages and 2.8% of the time on IP-to-email linkages, with the large majority of linkages over the same 90-day period unique to each provider.
- There are 3x more IPs-per-postal than there should be, compared to ISP reports.
- IP timestamp reporting is inconsistent across providers, with conflicting definitions of “first seen,” “last seen,” or “versioned,” undermining comparability.
- When (time of year) and where (state-level geography) both impact IP linkage accuracy.
- IPv6s are under-represented by 72% in large datasets, with only 15% of commercial files capturing them.
Download the complete findings for free here for even more granular data on linkage accuracy over time, state-by-state fluctuations, IP-to-email distribution variances, and much more.
The Impact: From Accuracy to ROI
When identity signals are misaligned, it affects outcomes. A simplified model from the study illustrates this cascading effect. The value of a $1 media investment could fall as low as $0.09 per dollar spent if IP-to-household linkage accuracy is low, because it reduces the effectiveness of subsequent postal, email, and demographic match layers.
That doesn’t mean IP data isn’t useful. But like all powerful tools, it benefits from calibration. The study’s findings suggest that improving consistency across providers and linkages could unlock substantial gains in targeting accuracy and media efficiency.
The Path Forward
Rather than raising red flags, these findings highlight an opportunity to treat IP accuracy with the same rigor applied to viewability, fraud prevention, and campaign measurement. And there’s momentum to do so.
At Truthset, we’re already seeing marketers, platforms, and data providers use Data Rated Audiences—audience segments scored by accuracy tier—to bring more transparency into their planning and buying. With these ratings, brands can choose to optimize for precision, scale, or a balance depending on their goals. It won’t eliminate risk, but it can make risk visible and manageable.
And for the broader ecosystem, two recommendations stand out:
- Standardize timestamp reporting for IP-based data. Alignment on definitions and update cadence will reduce confusion and improve interoperability.
- Increase transparency around IPv6 treatment to ensure today’s most common IP protocol is fully considered in activation and measurement.
Collaboration, Not Competition
These are shared challenges, and the good news is that collaboration is already underway. By working across silos—brands, data providers, publishers, platforms, and measurement firms—the industry can co-create standards that enhance trust, accuracy, and performance for everyone.
CTV is one of the most dynamic areas of digital advertising. With better data practices around IPs, it can also become one of the most efficient.
Understanding where data comes from—and how it links together—has always been key to audience accuracy. This study offers new visibility into one of the most important identifiers in today’s ecosystem. Now, with that insight in hand, the next step is collective action.
Accuracy isn’t just about getting things “right.” It’s about making smarter decisions with the best information available. And that’s something we can all agree on.
Download the full report below for additional data, findings, and takeaways.
The Challenges of Using IP Address for Audience Identification and Measurement