For most brands, the honest answer is two or three out of five. The rest rely on voiceover, music, or sound effects to deliver a meaningful part of the message. Some of them are structured in a way that the core promise is never communicated visually at all — the product appears, the brand is shown, but what the product does and why someone should care is carried entirely by the audio.
This is a performance problem, not just an accessibility problem.
How People Actually Watch Social Video
A significant portion of social media consumption happens without sound. The specific numbers vary by platform and context, but the pattern is consistent: mobile users in public spaces, users who have been trained by years of autoplay to expect silent video, users who simply prefer to browse without audio. Estimates suggest that anywhere from 40 to 85 percent of social video is watched without sound depending on the platform and placement.
TikTok users are different from Meta users on this dimension — TikTok is a more audio-centric platform, and a higher percentage of TikTok viewing happens with sound on. But the gap is not as large as brands often assume, and TikTok's own data documents a 32% completion lift when captions are added to paid ads. That number is significant because completion rate is the primary distribution signal on TikTok. A 32% improvement in completion rate is not a marginal gain. It is the difference between an ad that gets distributed broadly and one that does not.
What Sound-Off Viability Actually Requires
An ad is sound-off viable when a viewer who watches it without audio receives the complete core message. This does not mean the ad has to work identically with and without sound. It means the fundamental argument — here is the problem, here is how we solve it, here is what you should do — can be followed without audio.
There are three ways to achieve this. The first is through captions, which transcribe the voiceover or dialogue so the verbal message remains accessible. The second is through on-screen text that communicates key messages independently of the audio track. The third is through visual storytelling that does not require audio to be comprehensible — product demos, before-and-after sequences, and visual problem-solution narratives that work as silent film.
The most effective approach is usually a combination. Captions for any verbal content. On-screen text for key claims or calls to action. Visual sequencing that makes sense without either.
The Algorithm Measures This
Both Meta and TikTok measure watch completion as a core quality signal for paid creative. Ads that lose viewers partway through get less distribution than ads that hold attention to the end. Sound-off viewership affects this metric directly.
An ad viewed without sound that loses the plot at the thirty-second mark because a critical product explanation was delivered verbally will show lower completion rates than it would if captions were present. Lower completion rates lead to lower distribution. Lower distribution leads to higher effective CPMs, because the ad is working harder to reach fewer people.
This is the mechanism through which sound-off viability becomes a performance variable rather than just an accessibility consideration. The algorithm is measuring the behavior of all viewers, including the ones watching without sound.
The Caption Execution Problem
Most brands that add captions to their paid social ads do so as an afterthought. The captions are auto-generated, not reviewed for accuracy, styled inconsistently with the video, and placed in ways that compete with the visual content for attention.
This is better than no captions. But it is not the same as building sound-off viability into the creative from the start.
The difference between adequate and good caption execution is: captions that are accurate and synchronized, styled to feel like part of the creative rather than appended to it, positioned where they support rather than compete with the key visuals, and written in a way that makes sense as standalone text, not just as transcription.
When captions are built into the creative rather than added afterward, the production is typically simpler and the result is better.
The Practical Test
Watch your next piece of paid social creative on mute. If you cannot follow the core message, neither can a meaningful percentage of your audience. That is not an accessibility gap. It is a performance gap.
The fix is usually straightforward: add accurate, well-styled captions, review for any key messages that are audio-only and either add on-screen text equivalents or restructure the visual sequence to carry them. Most of the time, this is a two-hour production task. The performance upside — documented at up to 32% completion lift — makes it one of the highest-return creative improvements available.