When Truthset revealed that IP-to-postal matches are accurate only 13% of the time, it exposed an uncomfortable truth about connected TV advertising: The industry has been optimizing for the wrong metric.
Somewhere along the way, we started equating scale with truth. Match rates became the scoreboard, and accuracy stopped being the goal.
When Truthset Founder and CEO Scott McKinley said, “Advertisers ask for maximum match rates because that’s how they deliver scale,” he was describing a system designed to fail. The higher the match rate, the looser the definition of a “match.” And the looser the definition, the less reliable the audience.
If CTV is ever going to live up to its promise of being truly addressable, accuracy has to matter again.
The root problem: scale over substance
Many CTV identity graphs pull an IP address from a bid request, treat it as a household and stitch in MAIDs and emails from data partners.
But in many cases, that IP address doesn’t belong to a household at all; it belongs to an ad server. In other words, the server is mapped to a street address. Add third-party segments on top of that, and all you have is noise.
As an industry, we treat identity like a commodity when it’s actually critical infrastructure. If you have a high, impressive match rate but the foundation is unstable, everything built on top of it collapses, including targeting, frequency management, attribution and marketing impact.
How to fix the incentive structure
The heart of the issue is a misaligned incentive structure that favors maximum segment size over precision. If you inflate match rates by accepting weak signals, outdated data and probabilistic stitching, you get bigger numbers on paper and weaker performance in reality.
But there are clear principles the industry can adopt to make CTV truly addressable:
- Regularly refresh household-to-IP associations. IP addresses change frequently, so household-to-IP associations need to be refreshed constantly. Once a month is not enough.
- Prioritize accuracy over volume – always. Multiple independent signals should agree before an identity match is made. Weak identity resolution inflates scale at the expense of truth.
- Limit bidstream data usage. Bidstream data is attractive because it increases your audience size. However, it’s also known for data-quality issues, inaccuracy and a lack of standardization.
- Make first-party data the primary source of truth. When identity is anchored in real consumer behavior, like on-site sessions and transactions, accuracy in targeting and measurement increases dramatically.
Accuracy matters more than ever
CTV has evolved from an upper-funnel-only channel to a measurable performance driver, but none of that promise is achievable if the foundation is flawed.
Brands can’t measure incremental lift if impressions weren’t delivered to the right households. They can’t optimize frequency if they’re effectively targeting ghosts. And they can’t justify budget shifts toward CTV if identity resolution is unreliable.
When identity is grounded in truth, brands can prove CTV drove incremental sales, not just exposure. Advertising then becomes a scientific process, rather than an act of faith.
Make accuracy nonnegotiable
CTV has the potential to be the most accountable media channel, but only if we stop celebrating inflated match rates and start demanding fidelity.
It’s time for the industry to decide what it wants CTV to become.
If we want this channel to fulfill its promise, accuracy can’t be treated as a nice-to-have; it has to be the operating system and the foundation upon which real performance is built.
