I don't think it's bandwidth costs that are the big problem here. People want a competitor that offers what YouTube offers: a way to get paid for uploading videos.
I think this service is fundamentally very hard to produce. For all its faults with monetization, it's not clear it can be done better (beyond toy scale). It's clearly not profitable to have humans match videos with advertisers, and computers are inevitably going to screw up, like in this instance. edit: it seems this case is just a basic policy issue, though that policy is a consequence of trying to avoid attempting to solve this problem for the long tail of small time uploaders. I just assumed it was another false positive based on the title.
Keeping advertisers and uploaders happy seems like an almost impossible task. Google may not be doing very well at it, but that doesn't mean it's easy to do better.
People want a competitor that offers what YouTube offers: a way to get paid for uploading videos.
The problem is actually bandwidth costs; unless you're at the scale of Google (where bandwidth and storage is really cheap for you), advertising alone wouldn't even pay for the service itself, let alone pay the uploaders.
Do we have to have distribution tied to ads / monetization? It’s not true for the text content, yet here we are with the video, YouTube havving consumed it all.
I imagine a world where creators pay someone for video distribution, but also plug-in advertising from third party ad agencies that have learned to match ads to content due to advantages in machine learning. Same as text today.
It was tried before (remember Vessel[0]?) but never caught on, even with a massive amounts of money they raised and were willing to spend. Most competitors believe that YouTube's proposition is similar to Netflix. You get quality content and people will come. But that's not how YouTube gets its audience.
In minds of general population, YouTube is a search engine. The was majority believe that YouTube aggregates the content all around the Web. Why wouldn't they? The content shows up there minutes after it originates somewhere else. It's almost impossible to prevent it from gaining any traction on YouTube. Another company just tried to go against them. You can read how it went here https://vid.me
I do hope that Peertube could figure that out one way or another. Paid content could boost the project a lot.
It wouldn't even have to be perfect like a lot of publishers usually demand, most youtube videos would be fine if the payment would simply authorize the user to stream and download the video as they wish.
I think the key is to take the advertisers out of the loop, removing them removes almost all of the issues that YT has experienced. I have some ideas how to do that, it’s the costs of storage and bandwidth that I can’t figure out right now.
Youtube ads have a cost of between $.10 and $.30 per view, according to googling "youtube ad cost"
An arbitrary youtube video [1] is 56.69 megabytes (7m36s of 720p MP4). AWS outbound bandwidth, which I'm told is outrageously expensive, costs $0.050 per GB, so a single view costs $0.0028 to serve.
Seems to me bandwidth isn't an insurmountable problem - if you can get the mix and volume of viewers, creators and advertisers on board, and make $.10-$.30 per view, and you can keep your overheads for engineering, support and moderation low.
They are but I think that content creators want to have all of the tools wrapped up in one platform verses having to manage multiple platforms and accounts.
On the contrary; a lot of professional Youtube content creators prefer to diversify their sources of income, to reduce the blow when one of the inscrutable corporations decides to halve your income.
A 10 minute HD video is about 122mb. If that video is watched 1 million times you have to provide 122TB of Bandwidth. That’s just one quarter of the average pewdiepie video.
Comments
There is a massive chance to disrupt this market right now. But bandwidth costs alone make it difficult to challenge google in this space.
I don't think it's bandwidth costs that are the big problem here. People want a competitor that offers what YouTube offers: a way to get paid for uploading videos.
I think this service is fundamentally very hard to produce. For all its faults with monetization, it's not clear it can be done better (beyond toy scale). It's clearly not profitable to have humans match videos with advertisers, and computers are inevitably going to screw up, like in this instance. edit: it seems this case is just a basic policy issue, though that policy is a consequence of trying to avoid attempting to solve this problem for the long tail of small time uploaders. I just assumed it was another false positive based on the title.
Keeping advertisers and uploaders happy seems like an almost impossible task. Google may not be doing very well at it, but that doesn't mean it's easy to do better.
The problem is actually bandwidth costs; unless you're at the scale of Google (where bandwidth and storage is really cheap for you), advertising alone wouldn't even pay for the service itself, let alone pay the uploaders.
Do we have to have distribution tied to ads / monetization? It’s not true for the text content, yet here we are with the video, YouTube havving consumed it all.
I imagine a world where creators pay someone for video distribution, but also plug-in advertising from third party ad agencies that have learned to match ads to content due to advantages in machine learning. Same as text today.
It was tried before (remember Vessel[0]?) but never caught on, even with a massive amounts of money they raised and were willing to spend. Most competitors believe that YouTube's proposition is similar to Netflix. You get quality content and people will come. But that's not how YouTube gets its audience.
In minds of general population, YouTube is a search engine. The was majority believe that YouTube aggregates the content all around the Web. Why wouldn't they? The content shows up there minutes after it originates somewhere else. It's almost impossible to prevent it from gaining any traction on YouTube. Another company just tried to go against them. You can read how it went here https://vid.me
[0] https://en.wikipedia.org/wiki/Vessel_(website)
I do hope that Peertube could figure that out one way or another. Paid content could boost the project a lot.
It wouldn't even have to be perfect like a lot of publishers usually demand, most youtube videos would be fine if the payment would simply authorize the user to stream and download the video as they wish.
I think the key is to take the advertisers out of the loop, removing them removes almost all of the issues that YT has experienced. I have some ideas how to do that, it’s the costs of storage and bandwidth that I can’t figure out right now.
it seems it s a huge problem - but there is ipfs and torrenting and platforms like d.tube which may actually be onto something interesting.
These sound like the words of a time-traveler from the CBS board room 75 years ago!
Youtube ads have a cost of between $.10 and $.30 per view, according to googling "youtube ad cost"
An arbitrary youtube video [1] is 56.69 megabytes (7m36s of 720p MP4). AWS outbound bandwidth, which I'm told is outrageously expensive, costs $0.050 per GB, so a single view costs $0.0028 to serve.
Seems to me bandwidth isn't an insurmountable problem - if you can get the mix and volume of viewers, creators and advertisers on board, and make $.10-$.30 per view, and you can keep your overheads for engineering, support and moderation low.
[1] https://www.youtube.com/watch?v=NjKgPLeJ79Q
There is something insane with your numbers. Probably the .1/3 per-view (per 1K views maybe) but too lazy to find out.
You always forget the huge(?) amount of un/low monetizable content/traffic/countries.
Patreon is doing well by allowing creators to monetize without doing the actual content hosting.
They are still growing exponentially.
They are but I think that content creators want to have all of the tools wrapped up in one platform verses having to manage multiple platforms and accounts.
On the contrary; a lot of professional Youtube content creators prefer to diversify their sources of income, to reduce the blow when one of the inscrutable corporations decides to halve your income.
This post is about a problem with unpopular videos. I don't think the bandwidth costs are that high.
A 10 minute HD video is about 122mb. If that video is watched 1 million times you have to provide 122TB of Bandwidth. That’s just one quarter of the average pewdiepie video.
Facebook could do it if they wanted to. They already have the scale.
YouTuba