In short: CTV frequency cannot be managed reliably from a single publisher report. Teams need a shared identity approach, a campaign-level exposure view and explicit rules for redistributing spend when overlapping supply pushes households beyond the intended range.
Key takeaways
- Plan frequency at campaign level, then inspect publisher-level contribution.
- Separate deterministic household matches from modeled reach estimates.
- Use frequency ranges and decision thresholds instead of treating one cap as exact truth.
Map where duplication can occur
List each publisher, exchange and buying path, then identify where the same inventory or household may be reached more than once. Supply-path transparency is a prerequisite for understanding whether frequency comes from intentional reinforcement or accidental overlap.
Use identity with clear boundaries
Household graphs and device signals differ in coverage and confidence. Document which markets and publishers support deterministic matches, where modeling is used and which exposures remain unlinked. The plan should not imply person-level precision when only household-level evidence exists.
Optimize toward effective reach
Review the distribution of impressions by household range, not only the average frequency. Shift budget when additional impressions are concentrated among already saturated households while eligible audiences remain unreached.
Frequently asked questions
What is a good CTV frequency cap?
There is no universal cap. The useful range depends on campaign duration, creative rotation, audience size and the confidence of cross-publisher identity signals.
Why is average frequency insufficient?
An average can hide a distribution where many households receive one impression while a smaller group receives far too many. Distribution ranges reveal that imbalance.
