If your CTV plan depends on CRM emails for “precision,” the hard constraint is this: the most common bridge (IP-to-email) is often wrong. AdExchanger cites an average 16% IP-to-email match rate. That’s not a rounding error. That’s a measurement and governance problem wearing a targeting costume.
And the timing matters. U.S. CTV ad spend is projected to hit $37.95B in 2026 (StackAdapt). When budgets get that big, “directional” identity stops being an academic debate and starts showing up as pipeline arguments, attribution fights, and a lot of quiet distrust in dashboards.
Here’s the tension critics keep pointing to: email is a person-level identifier. CTV, in practice, behaves like a household medium—shared screens, shared accounts, shared IPs. That mismatch is why email-based IDs can look clean in a deck and messy in the real world.
The core mismatch: CTV is household behavior, email is person identity
CTV impressions don’t reliably belong to one person. A living-room screen is the opposite of a single-user browser session. So when an identity strategy tries to carry a person-level key (hashed email) into a one-to-many environment, it inherits ambiguity immediately.
Critics frame it simply: household viewing behavior doesn’t map cleanly to person-level email identity. Even if an email address is “deterministic,” it may not represent the actual viewer behind the impression. In other words, determinism doesn’t guarantee relevance.
This is where teams get tripped up operationally. The workflow looks reasonable—take CRM emails, hash them, send them to a CTV partner, get an “addressable” audience. But the underlying join can be doing something different than what the org thinks it’s doing: not “email-to-viewer,” but “email-to-household-ish proxy.” Close enough for reach. Risky for any claim of one-to-one performance.
And once the org starts using that audience for suppression, sequencing, or frequency control, the cost of being wrong goes up. Quietly. A mis-join doesn’t throw an error. It just changes who sees the ad.
Match quality is the real bottleneck (not audience size)
Most criticism isn’t “email IDs are useless.” It’s that the match method often isn’t what buyers assume. Email-based CTV activation frequently depends on IP-to-email (or similar) linkages. And that’s where the 16% number becomes a flashing warning sign.
If the average IP-to-email match rate is 16% (AdExchanger), then an email-based audience might be “present” in the system but not actually findable at scale in CTV inventory. That has two downstream effects ops teams feel immediately:
- Addressability collapses: the audience you planned against won’t fully show up in delivery, so spend shifts into lookalikes, modeled expansions, or broad household proxies (sometimes without making that explicit).
- Attribution gets noisier: because exposure isn’t as controlled as the plan implies, any lift analysis is fighting contamination from off-target impressions and inconsistent reach.
But the data quality critique goes deeper than match math. AdExchanger’s reporting also surfaces skepticism about data sourcing and consent provenance—especially in environments like FAST where publishers often don’t have direct email logins. If an ID is assembled from third-party sources, the “first-party, consent-based” story weakens, and so does confidence in the graph.
That’s the operator takeaway: email-based doesn’t automatically mean first-party. The label is not the lineage.
The part nobody can audit: ID-level reporting and incrementality
The most practical critique is also the least glamorous: measurement is often opaque. AdExchanger notes that some platforms don’t clearly tie demand or performance reporting to specific IDs, which makes it hard to prove the incremental impact of email-based IDs alone in CTV.
That’s a real problem for Marketing Ops. Without ID-level transparency, there’s no clean way to answer basic questions that should be routine:
- Did the email-based segment outperform a household baseline, or did the platform’s modeled expansion do the work?
- Was frequency actually managed at the identity level, or did it drift across devices/apps?
- Did conversion lift come from incremental reach, or from hammering the same high-intent households?
Even strong teams end up over-reading platform dashboards because the alternative—proper holdouts, clean-room matching, and controlled experiments—requires tooling, time, and cooperation across partners. And if the vendor can’t show what portion of delivery was truly “on-ID,” the experiment can’t be falsified. It becomes vibes with confidence intervals.
There’s another wrinkle critics point out: some ID vendors argue they can’t verify underlying data for each token at scale because of volume and processing constraints. That’s not an excuse, but it is a constraint. It also explains why the market can feel like a black box: the system is built to transact, not to be audited.
One practical move: treat email as an input signal, not the key
If you only change one thing, change this: stop treating email as the identity solution. Treat it as one deterministic input among several, and make the vendor prove where it’s used.
Even skeptics generally concede that email-based identifiers can help in CTV when they’re verified, hashed, privacy-managed, and combined with other signals—CTV device IDs, IP, and identity graphs such as UID2 or RampID (as referenced in the industry discussion summarized by AdExchanger). That’s the more defensible framing: multi-signal identity with documented joins.
And yes, there are cases where one-to-one precision matters more. The AdExchanger summary points to verticals like pharmaceuticals, where tighter targeting can reduce inappropriate placements. That’s a real “when this is worth it” clause: if the cost of the wrong viewer is high, paying for better identity plumbing can be rational.
The opposite is also true. If the goal is broad incremental reach and attention at a household level, pushing everything through a person-level email ID can add complexity without adding lift. More joins, more failure modes, more arguments at readout.
Circle back to the 16% match-rate critique. It’s not saying “never use email.” It’s saying: don’t build your CTV strategy on a bridge that collapses under load. In 2026, with nearly $38B projected into the channel, the winning posture isn’t faith in IDs. It’s insistence on transparency, a baseline to beat, and measurement that can survive an audit.
