Connected TV (CTV) Ad Completion Rates: What Counts as Good, and Why Non-Skippable Isn't the Same as Watched

Pull up a Connected TV ad report and the completion rate will usually look excellent — often 90% or higher, sometimes close to universal. For anyone used to YouTube or social video benchmarks, where a 50% completion rate can be considered solid, that number looks like a huge win. It mostly isn't telling you what it appears to be telling you, and treating it as a straightforward quality signal is one of the more common mistakes in CTV campaign planning.

Why CTV completion rates are high almost by default

Most CTV ad inventory is structurally non-skippable — a viewer streaming a show on a connected TV app generally can't fast-forward or skip the ad break the way a YouTube viewer can hit "skip ad" after five seconds. That single structural difference explains almost the entire gap between CTV and social completion benchmarks. A high completion rate on CTV mostly reflects the format's constraints, not necessarily viewer interest, which means it's a much weaker signal of creative quality than the same number would be on a platform where leaving is one tap away.

What a high completion rate can hide

Non-skippable doesn't mean attentive. CTV is watched in a living room, often on a second screen setup where a phone is open at the same time, often with someone getting up during a break, often as background while doing something else entirely. A "completed" ad in this context can mean anything from fully attended to fully ignored while technically playing in front of an unoccupied room. This is the same distinction that shows up between attention and engagement on social platforms, just pushed to an extreme — on CTV, the passive metric (completion) is almost guaranteed by the format, so it stops functioning as a proxy for the thing you actually care about, which is whether the message registered.

What actually predicts whether a CTV ad is landing

Because completion is close to a given, the more useful signals shift toward outcome and recall metrics — brand lift studies, site visits or search volume following a flight, and in retail contexts, direct attribution to purchases. These are slower and more expensive to measure than a completion percentage, which is exactly why completion rate remains the most-quoted number in CTV reporting even though it's one of the least diagnostic ones available.

What this means for how CTV creative should actually be built

If viewers may only be partially attending for a meaningful share of the ad's runtime, front-loading brand and message becomes more important than it is on a skippable format, where a slower build can still pay off because a viewer has to actively choose to leave. On CTV, the more realistic assumption is that attention is intermittent throughout the ad rather than continuously present until a viewer deliberately exits — which argues for repeating the core message and branding more than once across the spot, rather than saving the payoff for a big finish that a distracted viewer might simply miss.

Don't rely on hook mechanics built for skippable formats. A slow-build structure designed to survive the first five seconds of a skippable pre-roll is solving a problem CTV doesn't really have — the risk on CTV isn't an early skip, it's being tuned out mid-play.

Repeat the brand and the offer, not just the story. A viewer who tunes back in at second twenty should still be able to tell what's being advertised.

Testing creative before the media buy, not after the flight report

Because completion rate won't tell you much, and outcome data only arrives after a flight has already run and spent real budget, the more useful check happens earlier — at the edit itself. Axony analyzes a finished cut and produces a predicted, second-by-second attention and retention curve, which is a more specific read on where a spot is likely to lose engaged attention than a completion percentage that was always going to be high regardless of how the ad was built.