So be careful when you create and design for silence.

Should we publish video on social platforms with or without sound? That was one of the big debates last year, especially with regards to publishing on Facebook and Snapchat.

In a 2016 industry survey, Digiday found that 85 percent of Facebook video is watched without sound:

The news shouldn’t come as much of a surprise, as Facebook has built a video ecosystem that does not require users to turn the volume up — and publishers have been more than happy to play ball. Most users’ news feeds are now inundated with short videos that feature text or captions narrating what’s being shown on screen. While most of these videos feature narration or some form of background music, the intent is to make it easy for people to consume the information presented in the videos without needing to turn the sound on.

As a counter-point, Snapchat’s Imran Khan took a shot at the practice. “‘Basically when you’re buying advertising without sound,’ Khan said, ‘you’re not really buying video, you’re buying moving banner.’” And perhaps he’s right to say so: more than two-thirds of Snapchat videos, according to the platform’s internal metrics, are viewed with the sound on.

So which platform has the right approach? Could it be … both of them?

For one, videos are being viewed in different contexts in each platform. The key is consent: In each platform, has the user given consent for the platform to deliver content that will meaningfully change their current experience?

On Facebook, most videos are encountered within some kind of feed, whether it is on the main News Feed or on an individual’s or publisher’s Facebook page. A user may not have navigated to those pages with the explicit purpose of watching a video — and no one likes to be surprised by the sound of an autoplay video — so Facebook is probably right to set videos to play without sound by default.

On the other hand, Snapchat’s videos require a form of opting-in. Whether it is clicking through to see a friend’s story or clicking into a publisher’s section in Discover, there is clear intent to watch a video. That video may include ads which the user did not explicitly opt into, but the user would not be surprised to encounter them. It should also be noted that Snapchat follows the audio setting on your phone anyway — if you had set media to mute, the platform won’t override that setting.

Why does all this matter? There’s a strong case to be made that, in many cases, video is more powerful with sound.

Case Study: United Airlines

If you have been on the internet this week at all, you surely already know about the United Airlines debacle. Video captured by passengers of Dr. Dao’s forced removal instantly became a major talking point across social media, internet forums, and cable news.

I’ve spoken to several people about the video, and what has jumped out at me in the hours and days that followed was the varying reactions to what happened from people who 1) only read about the event, 2) either only saw stills from the video or watched without sound, and 3) watched the video with the sound on.

For people who only read about the video, their reactions felt dispassionate. They seemed to me to be the most likely people to raise questions about Dr. Dao’s motives and intentions. It doesn’t help when many of the articles referencing the event used United’s terminology of “re-accommodating customers.” But even if readers came out against United’s handling of the situation, there was little outrage. It was just another airline incident to them.

For people who only saw imagery or muted video of what happened, there was certainly some level of outrage — I counted myself in this bucket at first. For the first few hours after it was initially reported, I was working on a project, and only saw stills of Dr. Dao slumped over and of him getting back on the plane. It was certainly enough to see the incident for what it was — an egregious abuse of power.

But it wasn’t until I watched — and listened to — the full video, and heard Dr. Dao’s screams as they dragged him away, that it really hit home for me just how shocking this was. I couldn’t put myself in his situation without that kind of immersion. So when platforms or publishers choose to remove audio by default, I think it’s very possible that they are limiting the full impact of their stories.

So how can publishers solve for this?

The answer probably isn’t to revise settings wholesale to enable sound by default — otherwise they would be violating the principle of consent discussed earlier. Instead, publishers should be more mindful, on a case-by-case basis, of whether audio is needed for their storytelling. In those cases where sound meaningfully adds to the experience, find ways to encourage users to enable audio, whether through upfront on the video itself, or as recommendations in the post text.

This way, users will have the information they need, as well as the power to make informed decisions on not just what they watch, but how they watch it.