Frequency Capping Across Netflix, Prime Video and Hulu

Why frequency capping matters more here than on skippable formats

On a skippable, feed-based video ad, an over-frequent viewer has an escape valve — they scroll past or skip. Non-skippable streaming and connected-TV video removes that option, which means an uncapped or poorly-capped frequency problem is far more visible and far more likely to damage brand perception, not just waste budget. A viewer who sees the same 30-second ad four times in one evening across Prime Video and Hulu has no way to opt out of the fifth impression the way they would with an in-feed ad, which is exactly why cross-publisher capping is a bigger deal on this inventory than almost anywhere else in a media plan.

There's a cost dimension too, separate from the viewer-experience one. Every impression served to a household that's already well past a useful frequency level is a dollar that could have reached a new household instead — on a CPM-billed format, over-frequency isn't free even when it isn't actively annoying anyone, it's simply diminishing returns bought at full price.

How the cap actually gets applied

Amazon's authenticated graph — built from account sign-ins, Fire TV device registrations and Prime Video streaming sessions, reaching over 90% of US households deterministically, with match rates above 80% specifically for streaming TV campaigns — is what makes a genuine cross-publisher cap possible at all. Set correctly, a campaign spanning Netflix, Prime Video and Hulu can enforce something like "no more than four impressions per household per week" across all three at once, rather than four per platform, which would allow up to twelve. The setting has to be applied at the campaign or household-frequency level that spans supply sources, not configured separately inside each platform's own line item, or the cross-publisher benefit doesn't materialize.

A worked example

A campaign running $8,000 on Netflix, $10,000 on Prime Video and $6,000 on Hulu, each independently capped at four impressions per household per week, could theoretically serve one household up to twelve impressions across the three platforms in a week — a frequency level that would read as harassment on non-skippable video, not persuasion. The same campaign, with a single cross-publisher cap of four impressions per household per week applied through the authenticated graph, holds every household to that stated limit regardless of how many of the three platforms happen to reach them. The difference between those two configurations is entirely a setup choice, not a platform limitation.

What to do when frequency looks wrong

If completion rate drops over the course of a campaign flight with no change in creative, check frequency distribution before anything else — a rising share of households seeing the fifth or sixth impression in a week is a common, quiet cause of declining engagement that looks like creative fatigue but is actually a capping configuration problem. If a cross-publisher cap is set but frequency reporting still shows households well above it, confirm the cap is genuinely applied at the household level across supply sources rather than per-platform — a misconfigured setting that looks correct in the campaign builder is one of the more common troubleshooting dead ends in a streaming plan.

If frequency capping seems to be working but reach is much lower than budget suggests it should be, the cap itself may simply be too tight for the available ad-supported audience in that supply mix — a cap set for a large, mature audience pool can artificially constrain delivery on a newer or smaller supply source like Netflix, where the reachable household count within a given month may not support the same frequency ceiling.

The common mistake

The mistake is assuming a frequency cap set inside one platform's line item automatically governs the whole campaign. It doesn't, unless the cap is specifically configured to work across supply sources using Amazon's authenticated graph. reMKTR checks cross-publisher frequency settings as a standard step in every multi-platform streaming buy, precisely because the failure mode here is silent — nothing in a standard delivery report flags an uncapped household the way a budget overspend alert would, and the first sign of a problem is often a completion-rate dip that gets misdiagnosed as a creative issue.

The related mistake is setting a cap once and never revisiting it as a plan's supply-source mix changes. A cap tuned for a two-platform campaign running Prime Video and Hulu needs re-checking the moment Netflix or ESPN gets added — the same numeric cap that felt right for two supply sources can quietly under-constrain frequency once a third or fourth is layered on, if it isn't confirmed to still be applying at the household level across all of them.