Imagine sending a birthday gift to someone based on a home address that you’re only 13% sure is accurate or an email address that you’re only 16% sure is correct.1 Chances are the gift ends up at the wrong house or email address.
That’s what happens when IP-based targeting assumes accuracy. In other words, when IP-based targeting relies on shaky accuracy, the likelihood that an advertiser is actually reaching the intended audience becomes incredibly small.
Many advertisers tend to rely on IP addresses and device IDs to identify and tailor ads to their audiences and then measure the impact of those ads. IP addresses are crucial in the digital and streaming ecosystem because they play a large role in connecting devices in a household, validating bids, and protecting against fraud. And while IP addresses are an important piece of the overall identity puzzle, when it comes to using IP addresses as a single source of truth for identifying households and the associated audience attributes they have, players in the streaming ecosystem need to better assess their reliability. Here’s why.
IP addresses are like hotel rooms: temporary stays,
not forever homes
Say the internet starts lagging during a big video call or buffering during the championship game – what are most people going to do? Reset their router. But many users don’t realize that any time that router gets reset or unplugged, it causes the IP address to change. Internet providers also mix things up for security reasons or due to software updates, rotating IP addresses between households. So, the IP you have today might belong to your neighbor tomorrow. Throw in a VPN for remote work or privacy, and suddenly that IP could point to a completely different city. And when it comes to location, IP addresses usually only get as specific as the city level, making them especially unreliable in crowded neighborhoods or near borders.
This unreliability isn’t just theoretical, it shows up in the data. According to a recent study by Truthset on the reliability of IP addresses, commissioned by CIMM and Go Addressable, IP addresses can be unreliable for several reasons. Truthset found the match between IP address and email address across identity providers was only accurate 16% of the time and the match between IP address and postal address was accurate just 13% of the time.1
If an identity provider has information on someone based on their postal address or email address but then uses their IP address as a means to target them on CTV, chances are they are not targeting the right household.
This should raise a major flag for anyone relying on IP addresses as a household identifier for audience targeting, frequency capping, and measurement. Many solutions are simply not delivering the accuracy advertisers assume they are.
IP addresses are reliable when verified at the source
The way many IP addresses are used today for targeting is often based on inferred linkages to other datasets such as email or postal address. These probabilistic connections often break down because IP addresses themselves lack consistency.
When the source of IP address data comes from deterministic sources, IP-based targeting is as good as any identifier used for targeting. Internet service providers, like Comcast, are deterministic sources because they generate, assign, and manage IP addresses for users. Some data providers have direct connections to deterministic sources in this way, making them viable for use in targeting. The best practice for using IP addresses as identifiers is to connect to a source of truth to optimize targeting accuracy, frequency, and outcomes for advertisers.
Reliable identity solutions are shaping the future of advertising
IP addresses are a foundational part of the internet and will continue to play a large role in targeting and performance measurement on CTV. Increasingly, other more stable identifiers are being used for targeting.
For example, At FreeWheel, IP addresses play only a small role in audience targeting—just 17% of the time—because our Identity Network prioritizes more reliable identifiers. Most targeting within Buyer Cloud (formerly known as Beeswax) leverages stronger signals, ensuring accuracy and scale.2 Streaming video advertisers are shifting toward deterministic methods for targeting because they deliver better precision, are privacy-centric, and reduce wasted impressions.
Many of the leading identity solutions today incorporate proprietary first-party IDs and partner with ISPs for verified linkages to connect devices and ultimately identify users. More accurate graphs have deterministic data at their core and fill in the gaps with probabilistic data.
The FreeWheel Identity Network was built with the challenging nature of IP addresses in mind, prioritizing reliable identifiers and enabling scalable integration for ID owners. Our graph natively incorporates Comcast’s deterministic IP-resolution data, verified household addresses from trusted data providers, and first-party IDs from streaming subscriptions, giving us direct access to a verified IP address–to-email address linkage and addressability with proprietary IDs.
Start with the right questions
There is no single solution to streaming addressability. It takes interoperability and several durable inputs to get there, as there isn’t one company that has every data input across all households.
Advertisers should make sure to ask their tech partners the right questions about data and identity solutions – including what signals are prioritized to find and reach their target audience and how those signals allow for matching back to an attributed action or conversion.
Publishers should partner with the right audience providers to ensure audiences accurately scale across varied inventory and share deterministic identity signals wherever possible to deliver outcomes for their advertiser clients.
Working with a partner that utilizes deterministic data to reliably attribute IP address to a household allows advertisers to better ensure they can reach and measure their audiences with confidence is a great place to start.
If you want to learn more about FreeWheel’s identity solutions, contact us here.
Sources
1. CIMM, Go Addressable & Truthset, “The Challenges of Using IP Address for Audience Identification and Measurement,” November 2025.
2. FreeWheel, internal data, 2025.
