Even in 2025 every other streaming site is unwatchable because of video buffering, videos not loading, and other hiccups in service that make it annoying as hell to use.
99% of websites in general suck, and the ones that don't get millions and billions of users.
Arrogant programmers think their products are great and they are geniuses, but in general 99.9% of their work is glitchy unusable trash and going nowhere.
Say what you want about youtube, it simply works better than anything else.
This. Just look at Nebula in comparison - it's positioning itself as a competitor and yet their developers didn't even think supporting landscape mode on Android tablets was important for a VIDEO app. It's so riddled with bugs that it's mostly unwatchable.
It's hard to take these "YouTube is popular due to monopoly" arguments seriously when the competition can't even get the basics right.
> It's hard to take these "YouTube is popular due to monopoly" arguments seriously when the competition can't even get the basics right.
I feel the same about a lot of online shopping. In Germany people often moan about Amazon and while it's has it's share of issues, the competition is often so bad. Really slow processes that feel like someone adopted a "submit order via fax" process slightly for the web, horrible web sites, sometimes next-to-non-existent customer service. No wonder the alternatives aren't taking off as they fuck up the basics before we even get to the point of starting to compete.
Wow those were some strong words, that I’m not necessarily disagree with.
I think one company that would be a counter argument is twitch, it’s main business is definitely live stream but it also hosts a lot of clips generated from the various streams, the user base is happy with the product I think.
Not as an argument, but a side note: this is an effect of webdev being a ball of mud with wires sticking out of it. They couldn’t make a bad video player if <video> was good enough as is. Same for forms, layouts, etc. The huge upside of desktop frameworks around 2000 was that they just worked as designed, and if you weren’t skilled, that was a big barrier for making bad ui (not a barrier for true idiots, but still). Web though, it is basically a recipe for making something so idiotic that you couldn’t even imagine, and it invites, when not requires, you to try.
Youtube may work better than “everything else”, except for basically any mainstream porn site. It is just so trash compared to pornhub or xvideos that according to your theory these sites could just start serving sfw content and destroy youtube. Idk about that.
I find YouTube's UI performance quite bad but at least the performance of the videos themselves is good.
By UI performance I mean I click to go to the homepage and it loads a blank homepage with placeholder video thumbnails then spins for 5-10 seconds. I have fibre.
Also, you open a video and want to click fullscreen? Oops, [ ] button just-displaced due to reflow.
You press up arrow to up the volume? Oops, the seek bar had blue focus frame and you skipped five seconds instead.
You wanted to skip five seconds with a right arrow but missed and pressed numpad 0? Oops, you jumped to 0:00 and there’s no way back, not in this 5 hour stream record.
Buttons load for extra ten seconds, there’s a visible delay between click and actual pause, non-16:9 videos break the layout below the player, moving a song in a playlist jerks as hell or crashes the tab, and so on and so forth.
The usual “you are product not a client” is incoming, but I think that youtube ui team is just trash. There’s no prioritization of these issues, right, but having these in the first place on a site like youtube means you’re just incompetent.
> By UI performance I mean I click to go to the homepage and it loads a blank homepage with placeholder video thumbnails then spins for 5-10 seconds. I have fibre.
I’ve never had this issue with YouTube ever and I also have fibre. I never had this issue with YouTube when I didnt have fibre.
I pointed out that I have fibre to preempt anything about it being designed for rich people in rich countries with good internet or whatever. The point was that it isn't an internet issue or a connectivity issue, it is a YouTube bug.
I am glad you havent had the bug. It is an annoying one. Your not having it does not make it any less of a bug for me though, unfortunately.
Yes, this is all true, but the monopoly part is YouTube was supported by Google search profits for most of its existence, which its competitors don't have. The good part is YouTube is a really good service (ignoring their handling of gun related videos), the bad part is there are no viable competitors. If you like (legal) content they want to limit, block or otherwise hobble, you don't have any other great option.
So YouTube is subsidized by Alphabets profits. Basically everyone in the world is better off apart from Alphabet shareholders and maybe a few giant corporations slightly smaller than Alphabet that want to compete if they were able to sell a worse product for more money.
Exactly. It's bizarre to me that there's this idea of "fairness" that we have to purposely degrade a great product just so others can create a bad (but now relatively less bad) product, to stick it to Alphabet. Wasn't anti-monopoly policy supposed to benefit the consumer and not the corporations that are just behind the #1 product?
Antitrust/competition law was originally aimed at improving competition because anticompetitive conduct harmed competitors. Over the 20th century it evolved. Today it is squarely aimed at benefit to consumers long term. But that can be long: if anticompetitive subsidising makes things cheaper but kills competitors then (1) there will be low investment in the market, leading to less consumer surplus in the future and (2) there is the potential to abuse the resulting market power in the future.
I don't know that anyone wants to make YouTube worse to improve competition. The allegation is that YouTube is subsidised by other Google ventures so as to drive other competitors out and that it abuses its market power to harm competition. That doesn't ultimately benefit consumers in the long term.
But I have no idea if any of that is actually true. Maybe YouTube is just better designed and developed and has natural network effects so Google would win even if they didnt do anything anticompetitive.
The idea "to ban" ISP to put servers is not a very good one. (as mentioned in the article). It makes the internet work better and the ISP have lower costs that if there are a free-ish market they usually pass on the consumer. Not every "problem" needs more and more regulation.
ISP's are not forced to host these servers, they gladly do, because it's cheaper to colocate a caching server than to clog the uplinks and pay for that traffic. It's a win-win solution.
Big CDN providers also have their caching servers in ISPs. That's why building a CDN is a high-capex business. And if Youtube didn't have the money to pay for that traffic/server/code they woudln't have been successful.
We had a competing product that we sold to a big media group around 2010. The economy of this kind of platforms is super skewed towards put humongous amount of money beforehand, try to achieve network effect and try to find a way to monetize it. Google might've been the only company that could've done it, because it's a 3 sided market:
- uploaders/creators
- advertisers
- viewers
They not only put the money, but they shared the revenue in a meaningful way and that's why their patience and huuuge costs gave them the lead they enjoy now.
I am not sure that if you add all expences from the beginning they are net positive on this investment (if we measure the free cashflow generated after all capex has been paid off).
It's also important to recognise that content providers and CDNs adding private peering or hosting within ISPs doesn't diminish public peering. It actually frees up capacity on other transit routes.
ISPs don't QoS some companies to give them better service, the only difference is that in-demand companies tend to invest in capacity in partnership with ISPs. But ultimately, in most cases those investing generally use the same capacity that everyone else can use. The only company who doesn't resell their CDN capacity is Netflix. The others, Google and Amazon, dog food their own products. If you want to use the same systems as Prime or Google Video then you absolutely can. Other streaming providers use public CDN capacity just like anyone else.
Does YouTube get a favourable rate for capacity over other users? Yes and no. If Google doesn't charge YouTube then it's losing profit on the compute to sustain YT. But YT still has to make a profit and YT carries the cost burden of a great deal of legacy crap that a new entrant wouldn't. What Google has built is a miracle of engineering, to be able to get videos from relative nobodies to the other side of the in the world within minutes, at relatively high quality. While also allowing millions of kids to watch someone play Minecraft.
I respect what they've built. Would it be good to have diversity? In some ways yes, but in other ways choice sucks. Fragmentation of places to view content is something that gets increasingly complained about in the streaming world.
Is YouTube greedy? I don't think so. Building and maintaining what they've built is hard. As everyone else whose tried it knows. Just riding on their coat tails and leaching on their servers isn't sustainable. Ad blocking and saying Google deserves it isn't sustainable. In the extreme, if we burned down Google and said we wanted that model to end, the world would be a poorer place for it IMHO.
Context: 24 years in media, a decade in streaming for big companies, no affiliation with Google.
Yes, I agree... But it's even more complicated, because peering exchanges are only available in meaningful numbers in Europe and some countries around it. Also paths can be QoSed depending on many parameters, costs and deals.
We built a software that optimized routing based on cost depending on the 95th percentile usage, channel quality and some other parameters.
In Asia, South America (really shitty place for internet infra) and North America the conditions are different.
But Youtube did indeed built a great CDN, coupled with control of the end user video player made the best video on demand platform by far.
It's also worth noting that in most cases on-net caches (in ISPs) aren't as common as people think. It's mostly private peering at public data centres. Google doesn't have servers in all ISPs, this was more common in the past than today.
Because YouTube was created before copyright enforcement got strong. YouTube grew on pirated bootlegs until it got big enough it didn’t need them. You can’t replicate that again the ladder has been pulled after the launch and nobody else can do that.
Isn't TikTok more like the Instagram of video? I seem to remember early Youtube had a lot of digitised content, like private videos and TV show ripped fragments. With TikTok I see more of a Twitter/Instagram vlog and opinion platform. Youtube's success these days is, somewhat surprisingly, long-form content, often even exceeding an hour.
I'm too old to see just how popular TikTok is and where they could take that with the (supposedly petty young) demographic they have.
piracy probably wasn't the reason they won back then, but I think the point was that the change in IP enforcement since then might be the reason they can maintain their lead now
They had a rough start until googie acquired them. Lot of traffic, little revenue. It was the infusion of billions of dollars at the time and in the landscape it existed in that made youtube successful.
You would need a few times that to make it work all over again. Either that or some sort of decentralized framework with decades of volunteer man hours and hundreds of millions of dollars of volunteer hardware and network power to displace/replace it.
I remember going to daily motion spcifically to watch stuff that got DMCA'd out of youtube, the problem being I had no reason to stay there, the content was too sparse, even considering the amount of illegitimate content published
Vimeo decided awhile ago that they weren’t going to be a “shareable video” platform. That, combined with their pretty user-hostile UI, has limited their growth.
My guess is that they looked at the costs of hosting trillions of hours of video and decided that only a corporate giant like Google would ultimately be able to afford it.
One perspective is thinking about what would happen if YouTube is a separate entity from Alphabet/Google and more importantly adsense and search.
I think a lot of people that turned a hobby into a full time content creator job on YouTube will find themselves with much less ad revenue. Adsense is going to start charging a third-party company for services, which YouTube would be at that point, and those costs are likely to eat into any adsense revenue creators make, across the board.
There would also be the question of what search will power YouTube and if that can be physically separated from Google. There are likely economies of scale with how Google organizes data for search behind the curtain. That could be lost and increase YT operational costs or be another service YT needs to purchase.
If you could opt-out of Google ads and just be distributed/indexed by YouTube, then you'd be paying for hosting/delivery/indexing. Given that the economies of scale are spread among many users, the bigger streamers who this would benefit would then make the platform worse for everyone else.
Youtube is spreading the burden of carrying all that content, from utter crap that no one watches, deep archive and onwards to Mr Beast, etc. There's a huge volume of content that Google hosts that's costing more than it earns them.
Your comment I think drives my point even further. No content creator is paying for YT, neither is any end user[0]. So in this theoretical world where YT is separate from Google, now YT has to possibly pay Google for storage and one way to subsidize that would be to deduct it from creator's adsense revenue and/or limit the free content one can post in some way, like plans with different tiers of GB.
In some ways in this theoretical world, the small/new upcoming creators would have a larger chasm to cross into profitability if they move from a free plan into a paid hosting/delivery plan before becoming profitable. Unless a small/new creator gets massive quickly or goes viral they will have a much longer time before adsense can fund the storage. This might mean that the nonsense content goes away because churning out volumes of content to have more "surface area" for people to discover a channel can't be profitable.
[0] There may be premium content subscription options where some users pay a creator but I would imagine that is a minority of creators.
Youtube is a monopoly because it's not a very good business to be in and it basically lives off Google subsidization. It has plenty of openings for competition and none have too much forward movement.
Revenue. From earnings release for 2024 Q3[1]: "YouTube's total ads and subscription revenues surpassed $50 billion over the past four quarters". 2024 Q4 says: "Together, Cloud and YouTube exited 2024 at an annual revenue run rate of $110 billion."
Why would you sabotage YouTube?! Imo it is the best social internet app ever created, at least the one I spent the most time on. You have so much diverse content and everything is free....kind of.
But at this point I think the only way to compete with YouTube is decentralized P2P video hosting product because there is no way anyone can afford hundreds of millions of dollars for centralized video hosting product. TikTok was able to pull it off tho but remember that it started as a short form video service and its parent company was beefy enough to invest billions into user and content acquisition.
Aren't there companies that provide the "caching server for free" as a service?
If I understand the author right, the big companies are allowed to set up caching servers at ISPs.
Isn't this basically a CDN? If you spin up your own screaming start-up you would first go with akamai or whatever and if you reach sufficient scale you set up your own agreements with ISPs.
Is the blog basically arguing for making it illegal to cut out the middle man here?
One very important factor is creators. Unless you got people to create on your platform, it isn't going anywhere. As far as I know, Youtube has best ad-revenue split of all platforms
I think the ad revenue is only attractive for new entrants and serves as motivator. Allow to go full time for low pay earlier and invest in the channel. At the point where you need return you have to have sponsor deals and/or merch and other streams.
I think this is a key aspect. Advertising is what brings in the money after all, to keep the service running. And arguably this market is 3-sided: advertisers, content creators and viewers.
And as any multi-sided market there is a huge problem of bootstrapping. Advertisers will not be interested until there is sufficient viewers. Viewers not interested until there is enough content/creators. But creators not interested until there is enough viewers/advertisers.
One would need a really smart plan and excellent execution getting into a place where such a model starts to be sustainable or feeding into itself. The economies of scale regarding video hosting might mean that a break even point requires millions of monthly users.
I used to follow the peertube project years ago. Unfortunately it never rocketstarted despite the fact, that P2P for video streaming is by far the most efficient way to distribute content.
I've always thought of how weird it is that YouTube even exists, as much as I love it. I have tried many times to figure out how it could exist independently of Google or some other tech giant, and watched many competitors try and fail.
I'm not sure YouTube can exist outside of being a monopoly. I'd actually argue YouTube is the strongest evidence in existence in favor of monopolies, far better than anything Thiel has suggested.
I want to be wrong about this but the evidence suggests it's so.
I have to guess it's some kind of uber situation, where the core business is only possible (let alone profitable) at enormous scale. Free video hosting for everyone, accessible to everyone at any time, is a heavy lift that doesn't make sense without videos that people watch millions of times.
The real reason isn't bandwidth cost, it's visibility in the algorithm. You want a game trailer on your site: you can host it at a better bit rate for a great experience or use a competitor to. But if you do that, you get less views and thus less recommendations in the YouTube algorithm on your version posted there. So you out in a YouTube embed instead.
Same with steam or other dominate online stores with recommendation algorithms: you can market your game with ads linking a store with a lower cut. Less sales and views on steam from that means you don't take off in their algorithm. Game flops. Buy ads pointing to steam store and it would have kicked off the self stoking cycle and done well in this hypothetical. The other store can't compete with rate alone even in cases where you are driving the traffic, because you are giving up driving even more traffic at at the other store through the augmentation of the algorithm.
It benefits from being bundled with other Google businesses like the ad network and all the data they collect from search. And there’s network effects of creators and users. But given how big it is, the platform should really be regulated like the communication utility and public square that it is.
The platform shouldn't be regulated just because it is big.
Copyright laws, and unfair enforcement via the platform with no recourse (e.g., the "DMCA"-esque rules that cirvumvent actual DMCA laws, if nothing else), needs to change to make it a fairer place.
Simple - youtube doesn't suck!
Even in 2025 every other streaming site is unwatchable because of video buffering, videos not loading, and other hiccups in service that make it annoying as hell to use.
99% of websites in general suck, and the ones that don't get millions and billions of users.
Arrogant programmers think their products are great and they are geniuses, but in general 99.9% of their work is glitchy unusable trash and going nowhere.
Say what you want about youtube, it simply works better than anything else.
This. Just look at Nebula in comparison - it's positioning itself as a competitor and yet their developers didn't even think supporting landscape mode on Android tablets was important for a VIDEO app. It's so riddled with bugs that it's mostly unwatchable.
It's hard to take these "YouTube is popular due to monopoly" arguments seriously when the competition can't even get the basics right.
I don't see Nebula as a competitor to YouTube. They're both streaming platforms and have some of same content.
Nebula isn't free. Only invited creators are on there. Seems like they're clearly focusing on a niche.
I also had issues with Nebula at first. It's been steadily getting better. When did you try it?
I think now it probably performs more reliably than Youtube + Firefox on my Android.
> It's hard to take these "YouTube is popular due to monopoly" arguments seriously when the competition can't even get the basics right.
I feel the same about a lot of online shopping. In Germany people often moan about Amazon and while it's has it's share of issues, the competition is often so bad. Really slow processes that feel like someone adopted a "submit order via fax" process slightly for the web, horrible web sites, sometimes next-to-non-existent customer service. No wonder the alternatives aren't taking off as they fuck up the basics before we even get to the point of starting to compete.
Wow those were some strong words, that I’m not necessarily disagree with. I think one company that would be a counter argument is twitch, it’s main business is definitely live stream but it also hosts a lot of clips generated from the various streams, the user base is happy with the product I think.
Not as an argument, but a side note: this is an effect of webdev being a ball of mud with wires sticking out of it. They couldn’t make a bad video player if <video> was good enough as is. Same for forms, layouts, etc. The huge upside of desktop frameworks around 2000 was that they just worked as designed, and if you weren’t skilled, that was a big barrier for making bad ui (not a barrier for true idiots, but still). Web though, it is basically a recipe for making something so idiotic that you couldn’t even imagine, and it invites, when not requires, you to try.
Youtube may work better than “everything else”, except for basically any mainstream porn site. It is just so trash compared to pornhub or xvideos that according to your theory these sites could just start serving sfw content and destroy youtube. Idk about that.
I find YouTube's UI performance quite bad but at least the performance of the videos themselves is good.
By UI performance I mean I click to go to the homepage and it loads a blank homepage with placeholder video thumbnails then spins for 5-10 seconds. I have fibre.
Also, you open a video and want to click fullscreen? Oops, [ ] button just-displaced due to reflow.
You press up arrow to up the volume? Oops, the seek bar had blue focus frame and you skipped five seconds instead.
You wanted to skip five seconds with a right arrow but missed and pressed numpad 0? Oops, you jumped to 0:00 and there’s no way back, not in this 5 hour stream record.
Buttons load for extra ten seconds, there’s a visible delay between click and actual pause, non-16:9 videos break the layout below the player, moving a song in a playlist jerks as hell or crashes the tab, and so on and so forth.
The usual “you are product not a client” is incoming, but I think that youtube ui team is just trash. There’s no prioritization of these issues, right, but having these in the first place on a site like youtube means you’re just incompetent.
> By UI performance I mean I click to go to the homepage and it loads a blank homepage with placeholder video thumbnails then spins for 5-10 seconds. I have fibre.
I’ve never had this issue with YouTube ever and I also have fibre. I never had this issue with YouTube when I didnt have fibre.
I pointed out that I have fibre to preempt anything about it being designed for rich people in rich countries with good internet or whatever. The point was that it isn't an internet issue or a connectivity issue, it is a YouTube bug.
I am glad you havent had the bug. It is an annoying one. Your not having it does not make it any less of a bug for me though, unfortunately.
Yes, this is all true, but the monopoly part is YouTube was supported by Google search profits for most of its existence, which its competitors don't have. The good part is YouTube is a really good service (ignoring their handling of gun related videos), the bad part is there are no viable competitors. If you like (legal) content they want to limit, block or otherwise hobble, you don't have any other great option.
So YouTube is subsidized by Alphabets profits. Basically everyone in the world is better off apart from Alphabet shareholders and maybe a few giant corporations slightly smaller than Alphabet that want to compete if they were able to sell a worse product for more money.
If their competitors were any good, then google would have given them the money.
Exactly. It's bizarre to me that there's this idea of "fairness" that we have to purposely degrade a great product just so others can create a bad (but now relatively less bad) product, to stick it to Alphabet. Wasn't anti-monopoly policy supposed to benefit the consumer and not the corporations that are just behind the #1 product?
Antitrust/competition law was originally aimed at improving competition because anticompetitive conduct harmed competitors. Over the 20th century it evolved. Today it is squarely aimed at benefit to consumers long term. But that can be long: if anticompetitive subsidising makes things cheaper but kills competitors then (1) there will be low investment in the market, leading to less consumer surplus in the future and (2) there is the potential to abuse the resulting market power in the future.
I don't know that anyone wants to make YouTube worse to improve competition. The allegation is that YouTube is subsidised by other Google ventures so as to drive other competitors out and that it abuses its market power to harm competition. That doesn't ultimately benefit consumers in the long term.
But I have no idea if any of that is actually true. Maybe YouTube is just better designed and developed and has natural network effects so Google would win even if they didnt do anything anticompetitive.
The idea "to ban" ISP to put servers is not a very good one. (as mentioned in the article). It makes the internet work better and the ISP have lower costs that if there are a free-ish market they usually pass on the consumer. Not every "problem" needs more and more regulation.
ISP's are not forced to host these servers, they gladly do, because it's cheaper to colocate a caching server than to clog the uplinks and pay for that traffic. It's a win-win solution.
Big CDN providers also have their caching servers in ISPs. That's why building a CDN is a high-capex business. And if Youtube didn't have the money to pay for that traffic/server/code they woudln't have been successful.
We had a competing product that we sold to a big media group around 2010. The economy of this kind of platforms is super skewed towards put humongous amount of money beforehand, try to achieve network effect and try to find a way to monetize it. Google might've been the only company that could've done it, because it's a 3 sided market:
- uploaders/creators
- advertisers
- viewers
They not only put the money, but they shared the revenue in a meaningful way and that's why their patience and huuuge costs gave them the lead they enjoy now.
I am not sure that if you add all expences from the beginning they are net positive on this investment (if we measure the free cashflow generated after all capex has been paid off).
It's also important to recognise that content providers and CDNs adding private peering or hosting within ISPs doesn't diminish public peering. It actually frees up capacity on other transit routes.
ISPs don't QoS some companies to give them better service, the only difference is that in-demand companies tend to invest in capacity in partnership with ISPs. But ultimately, in most cases those investing generally use the same capacity that everyone else can use. The only company who doesn't resell their CDN capacity is Netflix. The others, Google and Amazon, dog food their own products. If you want to use the same systems as Prime or Google Video then you absolutely can. Other streaming providers use public CDN capacity just like anyone else.
Does YouTube get a favourable rate for capacity over other users? Yes and no. If Google doesn't charge YouTube then it's losing profit on the compute to sustain YT. But YT still has to make a profit and YT carries the cost burden of a great deal of legacy crap that a new entrant wouldn't. What Google has built is a miracle of engineering, to be able to get videos from relative nobodies to the other side of the in the world within minutes, at relatively high quality. While also allowing millions of kids to watch someone play Minecraft.
I respect what they've built. Would it be good to have diversity? In some ways yes, but in other ways choice sucks. Fragmentation of places to view content is something that gets increasingly complained about in the streaming world.
Is YouTube greedy? I don't think so. Building and maintaining what they've built is hard. As everyone else whose tried it knows. Just riding on their coat tails and leaching on their servers isn't sustainable. Ad blocking and saying Google deserves it isn't sustainable. In the extreme, if we burned down Google and said we wanted that model to end, the world would be a poorer place for it IMHO.
Context: 24 years in media, a decade in streaming for big companies, no affiliation with Google.
Yes, I agree... But it's even more complicated, because peering exchanges are only available in meaningful numbers in Europe and some countries around it. Also paths can be QoSed depending on many parameters, costs and deals. We built a software that optimized routing based on cost depending on the 95th percentile usage, channel quality and some other parameters. In Asia, South America (really shitty place for internet infra) and North America the conditions are different.
But Youtube did indeed built a great CDN, coupled with control of the end user video player made the best video on demand platform by far.
It's also worth noting that in most cases on-net caches (in ISPs) aren't as common as people think. It's mostly private peering at public data centres. Google doesn't have servers in all ISPs, this was more common in the past than today.
Because YouTube was created before copyright enforcement got strong. YouTube grew on pirated bootlegs until it got big enough it didn’t need them. You can’t replicate that again the ladder has been pulled after the launch and nobody else can do that.
I argue TikTok did / is doing the same path.
Isn't TikTok more like the Instagram of video? I seem to remember early Youtube had a lot of digitised content, like private videos and TV show ripped fragments. With TikTok I see more of a Twitter/Instagram vlog and opinion platform. Youtube's success these days is, somewhat surprisingly, long-form content, often even exceeding an hour.
I'm too old to see just how popular TikTok is and where they could take that with the (supposedly petty young) demographic they have.
Everybody pirated, only YouTube succeeded. This can't be the reason.
piracy probably wasn't the reason they won back then, but I think the point was that the change in IP enforcement since then might be the reason they can maintain their lead now
They had a rough start until googie acquired them. Lot of traffic, little revenue. It was the infusion of billions of dollars at the time and in the landscape it existed in that made youtube successful.
You would need a few times that to make it work all over again. Either that or some sort of decentralized framework with decades of volunteer man hours and hundreds of millions of dollars of volunteer hardware and network power to displace/replace it.
I wonder why the likes of Vimeo and DailyMotion from a similar era never seemed to really make it.
I remember going to daily motion spcifically to watch stuff that got DMCA'd out of youtube, the problem being I had no reason to stay there, the content was too sparse, even considering the amount of illegitimate content published
Vimeo decided awhile ago that they weren’t going to be a “shareable video” platform. That, combined with their pretty user-hostile UI, has limited their growth.
My guess is that they looked at the costs of hosting trillions of hours of video and decided that only a corporate giant like Google would ultimately be able to afford it.
Their software experience was terrible as well. Real player type stuff. You only went there when you had no other option.
One perspective is thinking about what would happen if YouTube is a separate entity from Alphabet/Google and more importantly adsense and search.
I think a lot of people that turned a hobby into a full time content creator job on YouTube will find themselves with much less ad revenue. Adsense is going to start charging a third-party company for services, which YouTube would be at that point, and those costs are likely to eat into any adsense revenue creators make, across the board.
There would also be the question of what search will power YouTube and if that can be physically separated from Google. There are likely economies of scale with how Google organizes data for search behind the curtain. That could be lost and increase YT operational costs or be another service YT needs to purchase.
If you could opt-out of Google ads and just be distributed/indexed by YouTube, then you'd be paying for hosting/delivery/indexing. Given that the economies of scale are spread among many users, the bigger streamers who this would benefit would then make the platform worse for everyone else.
Youtube is spreading the burden of carrying all that content, from utter crap that no one watches, deep archive and onwards to Mr Beast, etc. There's a huge volume of content that Google hosts that's costing more than it earns them.
Your comment I think drives my point even further. No content creator is paying for YT, neither is any end user[0]. So in this theoretical world where YT is separate from Google, now YT has to possibly pay Google for storage and one way to subsidize that would be to deduct it from creator's adsense revenue and/or limit the free content one can post in some way, like plans with different tiers of GB.
In some ways in this theoretical world, the small/new upcoming creators would have a larger chasm to cross into profitability if they move from a free plan into a paid hosting/delivery plan before becoming profitable. Unless a small/new creator gets massive quickly or goes viral they will have a much longer time before adsense can fund the storage. This might mean that the nonsense content goes away because churning out volumes of content to have more "surface area" for people to discover a channel can't be profitable.
[0] There may be premium content subscription options where some users pay a creator but I would imagine that is a minority of creators.
Youtube is a monopoly because it's not a very good business to be in and it basically lives off Google subsidization. It has plenty of openings for competition and none have too much forward movement.
YouTube made $50 billion last year - I wouldn’t call it subsidized
revenue or profit?
Revenue. From earnings release for 2024 Q3[1]: "YouTube's total ads and subscription revenues surpassed $50 billion over the past four quarters". 2024 Q4 says: "Together, Cloud and YouTube exited 2024 at an annual revenue run rate of $110 billion."
[1] https://abc.xyz/investor/
Last I checked, Alphabet still doesn't break down profit/operating costs for YouTube.
YouTube has been profitable for a long time now...
I don't think we know this - Alphabet publishes YouTube revenue, but not profit. Though I assume they are and have been for a while
Source? Alphabet’s 10-K does not account for all of YouTube’s expenses as is if it were a standalone business. It doesn’t even seem possible.
Why would you sabotage YouTube?! Imo it is the best social internet app ever created, at least the one I spent the most time on. You have so much diverse content and everything is free....kind of.
But at this point I think the only way to compete with YouTube is decentralized P2P video hosting product because there is no way anyone can afford hundreds of millions of dollars for centralized video hosting product. TikTok was able to pull it off tho but remember that it started as a short form video service and its parent company was beefy enough to invest billions into user and content acquisition.
Aren't there companies that provide the "caching server for free" as a service?
If I understand the author right, the big companies are allowed to set up caching servers at ISPs.
Isn't this basically a CDN? If you spin up your own screaming start-up you would first go with akamai or whatever and if you reach sufficient scale you set up your own agreements with ISPs.
Is the blog basically arguing for making it illegal to cut out the middle man here?
Most providers who do cheap/free CDN have exclusions for serving video because the economics of it suck.
- Vimeo is niche, but still around (and not bad).
- YouTube isn't the dominant video player in every country! Niconico is super popular in Japan.
> YouTube isn't the dominant video player in every country! Niconico is super popular in Japan.
This doesn’t make sense because Niconico might be popular in Japan, but it is absolutely dominated by YouTube there and has been for a long time now.
One very important factor is creators. Unless you got people to create on your platform, it isn't going anywhere. As far as I know, Youtube has best ad-revenue split of all platforms
I think the ad revenue is only attractive for new entrants and serves as motivator. Allow to go full time for low pay earlier and invest in the channel. At the point where you need return you have to have sponsor deals and/or merch and other streams.
If you don't burn out before that.
I'd say it is the two-sided market for advertising.
I think this is a key aspect. Advertising is what brings in the money after all, to keep the service running. And arguably this market is 3-sided: advertisers, content creators and viewers. And as any multi-sided market there is a huge problem of bootstrapping. Advertisers will not be interested until there is sufficient viewers. Viewers not interested until there is enough content/creators. But creators not interested until there is enough viewers/advertisers. One would need a really smart plan and excellent execution getting into a place where such a model starts to be sustainable or feeding into itself. The economies of scale regarding video hosting might mean that a break even point requires millions of monthly users.
I used to follow the peertube project years ago. Unfortunately it never rocketstarted despite the fact, that P2P for video streaming is by far the most efficient way to distribute content.
I've always thought of how weird it is that YouTube even exists, as much as I love it. I have tried many times to figure out how it could exist independently of Google or some other tech giant, and watched many competitors try and fail.
I'm not sure YouTube can exist outside of being a monopoly. I'd actually argue YouTube is the strongest evidence in existence in favor of monopolies, far better than anything Thiel has suggested.
I want to be wrong about this but the evidence suggests it's so.
I have to guess it's some kind of uber situation, where the core business is only possible (let alone profitable) at enormous scale. Free video hosting for everyone, accessible to everyone at any time, is a heavy lift that doesn't make sense without videos that people watch millions of times.
> strongest evidence in existence in favor of monopolies
i dont think it is evidence that monopoly is good.
The real reason isn't bandwidth cost, it's visibility in the algorithm. You want a game trailer on your site: you can host it at a better bit rate for a great experience or use a competitor to. But if you do that, you get less views and thus less recommendations in the YouTube algorithm on your version posted there. So you out in a YouTube embed instead.
Same with steam or other dominate online stores with recommendation algorithms: you can market your game with ads linking a store with a lower cut. Less sales and views on steam from that means you don't take off in their algorithm. Game flops. Buy ads pointing to steam store and it would have kicked off the self stoking cycle and done well in this hypothetical. The other store can't compete with rate alone even in cases where you are driving the traffic, because you are giving up driving even more traffic at at the other store through the augmentation of the algorithm.
It benefits from being bundled with other Google businesses like the ad network and all the data they collect from search. And there’s network effects of creators and users. But given how big it is, the platform should really be regulated like the communication utility and public square that it is.
The platform shouldn't be regulated just because it is big.
Copyright laws, and unfair enforcement via the platform with no recourse (e.g., the "DMCA"-esque rules that cirvumvent actual DMCA laws, if nothing else), needs to change to make it a fairer place.