This is the first in a series of posts I’ve got planned on the state of video on the modern web. Video is becoming central to the internet experience on more and more sites and it’s important to understand it.
Today’s post - why running a dedicated video site on the internet is nearly impossible.
Twitch is in Trouble
Amazon purchased Twitch for $1 billion in 2014. In many ways, the investment was an incredibly smart move. Twitch’s userbase has grown by a lot since 2014, just as Amazon thought it would. Its cultural relevance has exploded as well - livestreaming is increasingly mainstream. It seems like every few months there are new records being set on Twitch for the biggest livestream with the most viewers. And because of the Amazon purchase, Twitch has a giant supporting it with the knowledge of how to make money scaling a tech business.
There’s only one hiccup to all this great news for Twitch - they don’t make any money. They aren’t even close to being profitable.
Twitch somewhat infamously loses boatloads of money each year. They laid off staff in March 2023, October 2023, and January 2024 as part of cost-cutting efforts, and staff fear there might be another round of layoffs late in 2024. Twitch is struggling in both directions of the profitability equation. It’s incredibly expensive to maintain hundreds of thousands of simultaneous livestreams to millions of viewers, and they’ve also struggled to monetize that huge volume of traffic, with revenue flat over the last several years. They’re cutting costs in increasingly desperate ways, like not renewing the license rights to custom emotes. Any way you look at it, Twitch is struggling financially.
Objectively, this is a pretty weird result! Twitch, again, is hugely successful in terms of audience size and cultural relevance. They’re getting the traffic. They regularly create new superstar streamers with millions of fans. They have better platform features than other sites like Kick or Youtube Live. And they’re not a plucky startup any longer - they have Amazon’s muscle behind them. Amazon knows how to make money and they know how to scale. Amazon owns the largest and most sophisticated cloud computing infrastructure in the world, Amazon Web Services. Amazon even provides huge subsidies - anyone with Amazon Prime can gift a free Twitch Prime worth $5 to a streamer. And yet even with all these advantages, Twitch still can’t make any money! If Twitch can’t then how can anybody hope to?
All Video Struggles
The straightforward answer to that last question - how anyone can hope to make money in video online? - is that they mostly can’t and don’t.
YouTube is the 800-pound gorilla of online video, and we’ll come back to them shortly. But it’s pretty remarkable how little competition they have. Imagine you’re a video content creator - not a livestreamer or a shortform vertical video person, just regular well-produced 20 minute videos. What are your options other than uploading to YouTube? Could you even name another site?
There have been attempts. Vimeo, DailyMotion, Vevo all tried. But they all failed spectacularly. Vevo eventually deleted their app and standalone service to become a syndicator of YouTube music channels. Vimeo abandoned competing with YouTube to become a SaaS1 company. DailyMotion still technically exists in zombie form, but don’t visit the site, it’s very sad. The fact of the matter is that nobody competes with YouTube.2 It might be the case that YouTube enters new markets to compete with another company - they’ve entered livestreaming to compete with Twitch and created YouTube Shorts to compete with TikTok. But in their core business, YouTube stands alone. It’s simply too difficult to make money in the online video industry.
This extends out in a number of directions. Snapchat is still losing money. Streaming site Kick.com only exists because it’s owned and heavily subsidized by Stake, an online casino. The infamous pivot to video was driven by bad metrics and ended up as a terrible business decision for nearly every site that tried it. Even for a massively successful world-conquering app like TikTok, profitability is elusive. TikTok’s parent company ByteDance doesn’t release specific numbers on whether or not TikTok itself is profitable, but independent analysts believe it lost several billion dollars last year. It’s telling that they’re so focused on TikTok Shop and the live selling experience - they seem to think pursuing profitability purely through video itself is a bad bet.3
Even among major streaming services (admittedly a different business than Youtube or TikTok) it’s uncommon to see profits. Netflix is regularly profitable, but other services struggle greatly. HBO Max and Disney Plus both hover on the barest edge of profitability, having posted their first ever profitable periods in the last year after years of huge losses. As for the rest - ESPN+, Paramount, Amazon Prime, Peacock, Hulu, CNN+, AMC+, Apple TV - all money losers. These are major efforts from well-funded giant corporations with experience in the industry, and most of them fail.
Even YouTube famously took more than a decade to generate any profit at all, despite utterly dominating the industry. Any study of online video keeps coming back to the same conclusion. Unless you’re the biggest game in town - YouTube or Netflix - turning a profit in online video is almost impossible.
Why is Video So Brutal?
There are two real models to cover when we talk about online video: user-created and streaming.
The user-created space is where YouTube, TikTok and Twitch all live. Anybody can upload a video or start a livestream on these sites at any time, and people go to these sites to watch independent creators. And this space is especially brutal in terms of costs.
On YouTube, more than 500 hours of video are uploaded to the site every single minute. And that’s not a current number - that was the state of YouTube’s business back in 2019. It wouldn’t surprise me at all if that number is an order of magnitude higher today. The cost to dealing with that - around 5,000 hours of video uploaded every single minute - is enormous. Simply hosting and organizing those videos in a coherent way is a staggering technical challenge, even if many of them never get watched.
And that’s just the cost of hosting. An even more significant money pit is the cost of serving all that content with low latency and high resolution. These bandwidth costs are the real killer for any online video service. No video service that I’m aware of releases their bandwidth expense numbers, but everyone agrees it’s high. And then you have to deal with the cost of content acquisition. For streaming services, this is how much it costs to produce a hit show. For YouTube and Twitch, it’s how much they pay out to content creators. YouTube gives 55 cents of every ad-generated dollar to creators, whereas Twitch gives streamers 70% of revenue.
The basic math of trying to host an enormous amount of content, serve it instantly and at high quality to users all around the world, and pay to retain that content is grim. And beyond that basic math, there may be a natural monopoly effect for user-generated content.
Why would talented video creators post on any other site than YouTube when that’s where the audience is? And why would the audience use any other site, when all the creators are on YouTube? The only successful attempts to create new challengers have come from new formats - livestreaming and shortform vertical video - but YouTube is quickly gaining ground with copycat services Shorts and Live. YouTube is what Ben Thompson would call an Aggregator - the dominant gatekeeper for a two-sided market (viewers and creators) where both sides depend on the gatekeeper for access to the other side of the market. Viewers rely on YouTube to find creators and creators rely on YouTube for viewers, and one of the central insights of Thompson’s ‘Aggregation Theory’ is that Aggregators are often natural monopolists.
How to Break the Paradigm
Luckily, I think there are some ways to break out of this situation. There are companies testing new models, and we’ll explore them in part two of the series next week. Stay tuned!
Software-as-a-Service, aka selling technology software to other businesses
The only real competitors come in countries where YouTube is banned - RuTube in Russia, Youku in China, etc.
This is the core reason ByteDance is so profitable in China - livestreaming on the Chinese web is tightly tied to selling goods, almost like an internet version of HSN or QVC.
I can give some insight working for major streaming providers like Amazon Prime, BBC iPlayer and DAZN..
You gotta separate 2 type models - live and vod. Also 2 distribution models low concurrency with huge archive/channels versus high concurrency with small number of assets/channels.
Also 2 more variables funding( ads v subscription ) and coverage ( encoding profiles , DRM.and device support ).
Very fun field . I believe India cricket league achieved 100million concurrency recently and I deal regularly with 10,000,000 plus petabyte sized archives.
Looking forward to follow up articles. Reach out if you need any support
An interesting topic, and I look forward to reading the follow-ups. I feel like the subtitle is misleading; you don't actually explain why here, you just lay out the problem.