Ask almost any advertiser running Connected TV and you’ll hear the same frustration.
“We’re seeing the ad everywhere.”
“Our frequency feels too high.”
“We thought CTV was supposed to be controlled.”
On the surface, it sounds like a failure of planning or execution. In reality, it’s a misunderstanding of how frequency works in a system that was never built to behave like social or display.
CTV frequency isn’t broken. It’s distributed.
Frequency Was Easy When Everything Lived in One Place
In channels like paid social, frequency is simple because control is centralized.
One platform.
One identity graph.
One delivery system.
The platform knows who saw the ad, how often, and when to stop. Frequency caps are enforced in real time because the same system that serves the ad tracks the exposure.
CTV does not work that way.
In CTV, No One Owns the Whole Viewer
A single household can encounter your ad across:
Multiple streaming apps
Multiple publishers
Multiple devices
Multiple operating systems
Each of those environments sees the viewer differently.
A Roku household is not the same as a Samsung household.
A Hulu viewer is not the same as a Pluto viewer.
A FAST channel session is not the same as an AVOD session.
Even when identity resolution exists, it’s probabilistic and delayed.
That means frequency is managed within silos, not across the entire ecosystem.
What Frequency Caps Actually Do in CTV
When you set a frequency cap in a DSP, you are not setting a universal rule.
You are setting a constraint within the context the DSP can see.
The DSP can limit how often it bids for a household it recognizes. It cannot prevent another DSP, another publisher, or another device from serving the same ad elsewhere.
This is why frequency often looks fine in-platform but inflated when viewed holistically.
Both things can be true at the same time.
Why This Is a Feature, Not a Bug
CTV inherits more from linear television than digital media.
In linear, advertisers never had true frequency control. They had reach curves, GRPs, and probability.
CTV brings more control than linear, but it does not promise perfection.
What it offers instead is distribution across contexts, which can be a strength when used intentionally.
Seeing the same brand message across different environments can reinforce recall. The problem arises when creative is not designed for repetition or sequencing.
Where Measurement Fits In
Because frequency cannot be perfectly enforced, it must be interpreted.
This is where measurement partners step in:
Estimating household-level exposure
Modeling overlap across publishers
Identifying diminishing returns
Separating effective repetition from waste
The goal is not to eliminate frequency. It’s to understand when it stops working.
The Real Question Advertisers Should Ask
Instead of asking, “Why is frequency so high?”
The better question is, “What is this frequency doing?”
Is recall improving?
Is site traffic lifting?
Are conversions increasing after exposure windows?
If outcomes improve, repetition may be doing its job. If not, the issue isn’t frequency itself. It’s message, sequencing, or audience alignment.
The Bottom Line
CTV frequency feels broken because it doesn’t behave like digital platforms trained us to expect.
It operates across a fragmented, probabilistic ecosystem where control is shared and visibility is partial.
Once you understand that, you stop chasing perfect caps and start designing campaigns that work within reality.
That’s when frequency becomes a lever, not a liability.