select all

Video Is the Web’s Future and It’s a Wonderful Mess

When users began fleeing Myspace for Facebook roughly a decade ago, the general consensus was that Facebook’s big advantage, besides its exclusivity, was design. Facebook helped usher in a sort of “white picket fence” era of the web, in which uniformity reigned, and profile “hacks” were unnecessary and impossible. Gone were the Blingees, the auto-playing music embeds, the Comic Sans and worse fonts. In their place were profiles that looked structurally and visually identical.

The early web aesthetic, best typified by Geocities and later by Myspace, prized owner customization over audience usability. The most important thing about your little corner of the web was that it satisfied you, rather than your visitors. People need to hear this MIDI version of Yellowcard’s “Ocean Avenue” when they visit my page — who cares if it makes for a terrible user experience?

The uniformity of Facebook profiles (and Twitter profiles, and those of most modern social networks) solved this problem by taking all that customization away. You couldn’t tweak your profile’s layout, or add auto-play music, or even add GIFs. Cover photos do a bit to spice things up, but not much. And, despite some growing pains, the homogeneity of Web 2.0 means web surfers rarely have to worry about loading a page and getting served bad fonts, low-res images, and tags. It’s a much cleaner experience for the user, especially on the smaller and slower mobile browsers that have become increasingly common — and it’s a much more profitable experience for the companies that own these social networks, which can segment audiences and sell ads much more easily.

But we’re entering an age that’s moving away from text and static images and toward multimedia — video specifically — that threatens to upend web aesthetics yet again. Unless they want to accept the substantial costs associated with processing video, platforms are powerless in compelling users to adhere to any sort of style guide or uniformity. Online video is the new Myspace, and it’s chaotic and cluttered and garish, and I love it.

If it’s not already, video — rather than text — will become the lion’s share of the content you consume and share online relatively soon. It began with YouTube, but Facebook and Twitter are now also pushing native video platforms, and Snapchat’s far-reaching Discover section has proved a boon to publishers as well. Livestreaming services like Periscope and Twitch are becoming more en vogue. In a white paper published last year, Cisco predicted that “It would take an individual over 5 million years to watch the amount of video that will cross global IP networks each month in 2019.”

Globally,” the report says, “consumer internet video traffic will be 80 percent of all consumer Internet traffic in 2019, up from 64 percent in 2014.” One of the issues that web video poses is that it requires substantially more bandwidth than text or photos. The cost of streaming video for both providers and consumers is considerable. Video is generally what eats at your smartphone’s data plan, and has become a nuisance to the point that T-Mobile offered to not include Netflix and Hulu traffic when calculating data usage.

Rendering video — processing footage and post-production effects into a singular format — is similarly cumbersome on a technical level, and almost always happens on a user’s computer, not on a platform’s servers. Consequently, a video file with a number of elements, such as text overlays, graphics, and audio, is compressed into a single, uneditable file before it even leaves the computer. This gives users relatively high amounts of aesthetic control compared to a medium like text, where Facebook and Twitter don’t even let users bold or italicize passages.

The Myspace-ification of web video is most obvious on livestreaming services like Twitch, which has turned into a cacophony of visual elements as a result of the ad-hoc economy that surrounds it. That economy features escalating calls to action encouraging viewers to follow, subscribe, and donate. Subscribing for $4.99 a month gets viewers special perks, like custom emoticons to use in a stream’s chat room, while donating is like throwing money into a tip jar — you can do it as often as you’d like.

Let’s break down this screenshot of a Call of Duty stream.

On the left side is a stream of social-media mentions being piped in, above a banner of rotating usernames on other platforms.

On the right side, working top to bottom, is a video feed of the streamer, xMinks, playing the game. Below that is a grid of various economic metrics that commonly appear in the feeds of top streamers: the top donor of the day (customized to refer to them as “top llama”) and the amount that they donated, the name of the last person to subscribe, the name of the most recent donor and their contribution, and a goal for new subscriptions — currently six-tenths of the way there — leading to an incentive (in this case, a “wheel spin,” though I’m not clear on what that means specifically).

At the bottom is a banner featuring other social metrics: the name of the most recent follower (an act different from subscribing; it’s free), a counter of the number of new followers that day, and the name of the top monthly donor. On the left side of the banner is the name of a sponsor; in this case, it’s Astro, a gaming-headset manufacturer.

This is all in addition to the HUD elements that are actually part of the game: a ticker of kills, a minimap, a scoreboard, and an ammo count. This is a ton of information to throw at a viewer, and it’s always visually customized in a way that is specific to the streamer’s personal brand.

Here is a Minecraft streamer that includes many of the UI elements mentioned above as well as some new ones.

In the lower-left corner is a fund-raising thermometer. In this case, the streamer is raising funds for a Logitech C920 USB HD. At the top-center is a ticker displaying the current music track being played — it’s Avicii — an added aural customization impossible to convey in screenshots (much like the auto-play music of yore).

At the bottom of this stream for Counter-Strike are ads for stream sponsors, which are often gaming-equipment manufacturers and sellers like G2A and the aforementioned Astro.

This stream features another common Twitch feature. That penguin in the upper-left is a pop-up animation that appears whenever a new user follows the channel. The penguin is a custom flourish much like every other HUD element on Twitch streams. Some streams, in addition to in-game chat, also feature the text of the Twitch chat room running inside the frame.

Marcus “djWheat” Graham, Twitch’s head of programming, explained to me over the phone that all of these stream HUD elements — webcam picture-in-picture, chat-room functionality, follower pop-ups, etc. — came about organically over time as livestreaming game play gained popularity, cobbled-together kludges that echo the build-it-yourself culture of the early web. “At the end of the day,” he said, “Twitch has really not been the innovator of this at all; it’s just been the innovation of the users who have said, ‘You know what? This is great, it’s cool, I think I might be able to do it better.’ And once one user does it, there tends to be this trickle-down effect.”

If we think of Twitch as the Myspace of livestreaming — with a clear basic template that offers near-infinite visual customization — then what is livestreaming’s Facebook? Right now, there isn’t one. Apps like Periscope feature an analogous form, a continuous stream of comments and textual information laid directly onto the video stream, though it looks identical between users and slightly cleaner because Periscope is restricted to mobile devices. There is no way to pipe in video through anything other than the device’s camera.

There have been attempts at introducing first-party uniformity to online video. The most well-known instance being YouTube annotations, which were created with the best of intentions. In June of 2008, at the feature’s launch, Google said that annotations could be used “to add background information about the video, create stories with multiple possibilities (viewers click to choose the next scene), or link to related YouTube videos, channels, or search results from within a video.” And while some annotations were used in this way, the feature was often abused, creating a glut of text boxes and invisible link bubbles. Annotations were flat-out ugly. A TechCrunch headline six months after the feature’s introduction read, “Adding YouTube Annotations Just Got Easy. I’ve Got A Bad Feeling About This …”

In early 2015, YouTube introduced “Cards,” which it positioned as a successor to annotations. Unsurprisingly, their usage is more tightly controlled — they don’t appear automatically like annotations do (users have to click for more info), and their overlaid positioning is uniform. There are obvious reasons for reining in the usage of annotations — a cleaner look, and a more uniform mode of viewer interactivity will theoretically drive more clicks to ads, to channels, etc. If you know how to interact with one video, you know how to interact with them all.

But YouTube’s annotation system doesn’t really work for livestreaming, because the data contained within annotations is static. On livestreaming services like Twitch, variables are constantly changing and data points are always streaming in: new comments, new followers, new donations. That’s a heavier load to bear in terms of both technical processing and audience understanding.

I asked Twitch if they were thinking about first-party broadcasting tools that might allow for more uniformity. On some level, Twitch already offers apps on Xbox and PlayStation consoles that lack the customization options of a roll-your-own third-party solution. “Have we talked about having our own broadcast software? Of course, it’s come up many, many, many times,” Graham says. “Have we talked about building some of our own tools? It has, but at the same time, Twitch really supports and gets behind the sort of ecosystem that has been created around Twitch streams and all of the folks who have used their expertise to make tools that benefit our streamers.” The focus, as of now, is on letting users do their thing. That could change overnight, depending on the whims of the audience or Twitch’s parent company, Amazon — Twitter famously nerfed its third party in order to exercise greater platform control — but there is little sense that Twitch is inching toward the same idea in terms of video.

Graham isn’t concerned about users fleeing for cleaner pastures. “I think that more than anything, people want more flexibility and customization,” he said. He also noted that the service’s main demographic — the 15–35 range — is used to what might be incorrectly referred to as “information overload” at this point. “We use the internet every day, we’re distracted every 30 seconds by what we do. I would actually potentially argue that in this day and age, our user base is used to having a plethora of information fly out from their monitor and right into their brain.” There’s a learning curve to figuring out the HUD elements and interaction loops of a Twitch stream, but the learning curve isn’t particularly steep for digital natives. The system is not nearly as confusing as it might seem at first glance, similar to how a financial news channel like CNBC might fill its frame with market metrics. For livestreamers, those metrics aren’t related to money, they’re related to interaction.

As video — and livestreaming in particular — grows in popularity on the web, we can expect to see more of this: people becoming their own professional broadcasting operations, warping and tweaking the aesthetic of their stream to fit their brand in a way similar to a cable news channel, and piling loads of extraneous information into the frame. This is exciting! The idea that users want a tidy, uniform experience across a service is mostly an idea clung to by technologists — the average social-media user doesn’t care about cleanliness. If they did, we wouldn’t be seeing an astonishing amount of compression rot in the multimedia passed around on Facebook and Instagram and Twitter and Tumblr.

Twitch is, as of now, the best indication yet that the web is ebbing back toward Myspace on the Myspace-Facebook spectrum. The reasons for this are both technological — rendering and processing video is expensive — and cultural. As more and more people come of age using the web and using technology, uniformity in design and aesthetic isn’t as necessary. Facebook emerged as a service friendly to people who had never used a social network before, and that population is rapidly dwindling. We’re moving toward visual cacophony because we now have the ability to parse that mess easily. That beautiful mess is something to look forward to.

The Future of Video Is a Wonderful Mess