Sound-on creative: designing for the minority who listen

Every creative team already knows the baseline rule: design for sound off first. Captions carry the dialogue; visual cause and effect carry the hook; the mechanic has to read before the audio track has done anything at all. The rule is correct. It is also not the whole brief, because it treats sound as a nice-to-have layer instead of asking what the minority of viewers who do have sound on are actually experiencing, and whether that experience is worth the production budget it costs to get right.

The minority-with-sound-on audience is not evenly distributed, and that unevenness is the useful part. Sound is far more likely to be on in a handful of predictable contexts: a viewer scrolling at home rather than in a public space, a viewer already engaged enough to have tapped to expand a video, a viewer on a platform surface where autoplay audio is the default rather than the exception. So the sound-on segment skews toward people already closer to a decision, not casual scrollers. Treating audio as an afterthought means deprioritising the creative experience of your more engaged prospects in favour of the passive scroller who was never going to convert on this exposure anyway.

Sound-on creative: what changes when audio is a first-class layer

The mistake most portfolios make is not omitting sound. It is bolting sound onto a visual cut that someone storyboarded silent. A stinger dropped on a UI tap, a generic music bed under a gameplay clip: neither counts as design, both are decoration. Creative built with sound as a first-class layer starts somewhere else, with the mechanic's own audio feedback, a satisfying merge click, a combo chime, a failure buzz, belonging to the hook itself and tested for whether it reinforces or contradicts what the visual is claiming. A playable ad that promises "oddly satisfying" through visuals but offers a sound-on viewer no audio payoff is leaving its strongest asset unused for exactly the segment most likely to act on it.

Voice and dialogue are the other place this matters. Narrative-driven genres, particularly puzzle titles with renovation or story layers, increasingly lean on short voiced lines in creative. For a sound-off viewer that line does nothing beyond its caption. For a sound-on viewer the delivery and the comic timing do work that text cannot replicate, and testing separately against both conditions tends to surface a real gap between how a script reads and how a script lands.

Practical iteration, not a redesign

None of this argues for building two entirely separate creative pipelines. It argues for a second pass, once the sound-off cut locks, that asks specifically what changes for someone who can hear it. Music choice. Mix levels on mechanic feedback sounds. Whether a voiced line adds anything a caption did not already say. Testing should isolate the audio variable deliberately rather than assuming autoplay data alone answers the question, since autoplay engagement metrics conflate viewers who never turned sound on with viewers who did and thought nothing of what they heard.

The sound-off rule exists because it stops most impressions going to waste, but that isn't permission to stop thinking about the rest. A portfolio that nails the silent scroll and ignores the engaged, sound-on viewer is optimising for reach at the expense of the segment most likely to convert.

Related archive reading

These articles provide related context and remain subject to their stated review status.

• Vertical Video Sound Design for Mobile Game Ads
• Sound Design for Mobile Video Ads Most Players Never Hear