1. Costs
>> Since the majority of Google’s costs for the service are pure variable costs of bandwidth and storage, and since they’ve already reached the point at which no greater economies of scale remain, the costs of the business will continue to grow on a linear basis.
Costs of bandwidth and storage will continue to decrease, perhaps geometrically, in the foreseeable future, and so is Youtube operating cost.
2. Monetization
There must be other ways to monetize the eyeballs that have not been fleshed out in full yet. It's still very early days. The author assumes that YouTube has mostly run out of ways to monetize through ads (although he did point out a subscription model, esp. for marketers.)
I suspect bandwidth costs are decreasing at a rate that's faster than the increase in resolution of display devices though, and there's no point pushing more pixels than the best display has.
I suspect you're right, and this is a double-whammy when you consider that excluding file sharing, there's no significant driver for internet bandwidth above a small fraction of (even USA) broadband installations.
You can stream HD video in 2mbit, and virtually all of the content on YouTube is drastically lower native bitrate than this (and transcoded to 300-384kbps).
The elephant in the room is that YouTube should have given Google the "first mover advantage" and control of the legal feature-length TV episode (and perhaps movie) space, giving them the revenue of the popular, high-rate advertising to bear the load of the "uninteresting and unwatchable" (mind you that unwatched media has no bandwidth costs associated with it, reducing it to a storage expense alone). Hulu blindsided Google and now Google's best bet is to pull a Microsoft and iterate and hope for their competitor to slip. This is complicated by the fact that Hulu is backed by the content producers themselves.
The thing that hackers (pg included) fail to understand about video online is that there is no first mover advantage. There's no network effects. There's little if any brand loyalty. It's about IP ownership of popular content.
Who cares if YouTube is big? Without the rights to the most popular content, they're stuck trying to make money out of dog weddings and tearful tirades about Britney with 13 hours of additional low quality crap uploaded every minute. The content owners who setup Hulu are under no pressure from them whatsoever.
Google paid $1.65 billion for YouTube. It has never made a cent and its running costs are now equivalent to a small airline. Really, this is a no brainer. Shareholders should be asking for someone's head on a platter.
Isn't this the great flaw in the author's theory?
' It seems safe to assume that YouTube’s traffic will continue to grow, with no clear ceiling in sight. Since the majority of Google’s costs for the service are pure variable costs of bandwidth and storage...'
If traffic grows, and bandwidth and storage are the highest costs, then revenues should increase while costs decrease as bandwidth and storage costs come down.
He says later that youtube is big enough that they've maxed out their economies of scale (hence costs won't come down, which seems like a reasonable argument).
The real flaws with this article are that user content is growing geometrically compared to paid content (citation? where does he get this from), along with the fact that the author is the CEO of fliqz.com - a video streaming company. Good job reporting conflict of interest, business insider.
These are two different issues. Economies of scale are cost advantages that businesses obtain by expanding. Web Startup X starts out in a rented office with their content sitting on externally hosted servers, since they're too small to buy their own space/servers. At some point, they have enough users and cash that it becomes much more cost effective to build a datacenter and host everything in-house.
Youtube has gone through many of these stages, culminating in the Google purchase, and are now the big fish, with no more economies of scale left. They already have datacenters worldwide, and global name recognition. No matter how much they continue to grow, they will not become more profitable due to growth alone.
However, the other issue here is the exponentially rising value of computing technology, especially in bandwidth and storage space. It's why you paid $3500 for a 5MB hard drive in 1981 when you can get 200,000x the space for 2% of that price today. This fact, which has nothing to do with economies of scale, is why Youtube's profit should increase to >0 someday. The question is whether Google can (or will) wait that long.
This argument only makes sense if you believe Google's ability to generate money from YouTube will actually get worse as YouTube becomes more popular.
Why? Because bandwidth is getting cheaper. It won't be long before running YouTube costs half as much per visitor as it does today, so even if their cash potential per visitor stays the same, they will eventually break even.
On top of this, Google has some of the lowest bandwidth costs on Earth. They have massive peering and lots of their own wires; to suppose that they pay anything near what a startup video company would pay for bandwidth is ludicrous. But if you want to make YouTube look unsustainable, it helps to subtract a profitable network business from the bottom line.
Yes, YouTube might cost a lot for a few years, but if it retains its dominance for another three to five years, it will become ridiculously profitable.
"they’ve already reached the point at which no greater economies of scale remain, the costs of the business will continue to grow on a linear basis." This is wrong on several levels.
YouTube helps Google reach even larger economies of scale. Granted YouTube is still going to end up as a net loss, but things like heath care become cheaper per person as the company grows. However, over the next 10 years hardware and bandwidth is going to get even cheaper. While moving to HD will probably eat most of that benefit, I still think YouTube will break in the not so distant future.
Your forgetting about overhead. They spend K + Nx so cost per person is x + K/N which drops as N increases. Granted for companies the size of Google K is small relative to Nx but it's still a large number.
PS: Management tends to work the other way, so cost per person is ~x + Log(N) but generally speaking it's a net win to increase the size of the company.
Where is the actual data for these articles? Does anyone have a link to the actual figures Credit Suisse put together, with explanations as to where the figures came from?
OK, here goes some back-of-the envelope math (risky business in this group, so looking forward to hearing what I'm doing wrong..). Where I have made assumptions, I have erred on the side of 'expensive', i.e. this should be an upper bound estimate:
75,000,000,000 streams/year
400 kbps (average bw)
average stream length: 300 s (stream length, not video length, highball estimate)
-> BW/year = 1.13e9 GB/year
0.1 $/GB (BW cost, based on AWS S3 prices, again a highball)
-> BW cost ~= $113m / year
Only 1/3 of the estimated amount, and I'm not taking account of any clever stuff YT do to lower the costs. What gives? Is this a peak BW issue?
"To arrive at the estimated $360 million bandwidth tab for YouTube, the analysts assumed the site will receive 375 million unique visitors in 2009 and that a maximum of 20% of those users are on the site at any given time."
There's no way 20% of visitors are streaming video simultaneously. I doubt it's even 1%. Kongregate maxes out at about 0.5% of its monthly uniques connected to chat at once.
I've actually long thought that Google's YouTube acquisition wasn't purely to get into the online video market. It was all about the huge corpus of video that they acquired.
Let me explain. Google's general game plan is this: massive compute clusters + massive datasets + Phd CS brains trying to do something cool. My bet is that Google thought that having the world's largest video dataset was well worth the $2.5 billion they spent to buy YouTube. It's still worth $0.5 billion a year for them to keep maintaining it.
I don't think that they are going to be panicking about monetizing YouTube any time soon. That's not the reason they bought the company.
"I may not agree with what you have to say, but I will defend to the death your right to misattribute it to Voltaire."
-- Voltaire
(Seriously, though, YouTube as a company may or may not prosper, but YouTube-style streaming online video isn't going anywhere. So for most of us, the whole issue is moot.)
Marginal Rev is a fascinating blog, and it this comment is by the guy who posted an article that rose to the front page, explaining "not only is this article fascinating, but there's a lot of other fascinating stuff at location X on the internet".
Wait a second here! This guy is saying Google can't monetize a video site made up almost entirely of illegal content with an enormous user-base consisting mostly of people with an intelligence level and attitude roughly equal to a toddler on meth?! That's just crazy talk! YouTube will make tons of money, just you wait! Just like Facebook! I mean, there must be some way to monetize millions of users no matter what!
YouTube never made sense. It was geared to be bought out from the start.
I like watching videos on youtube as much as the next person, but according to the numbers ... youtube is in the red (never mind that Google will keep pumping money into it for the time being).
Serving up content costs money. Period. And youtube serves up a lot of content. The only reason Google bought youtube and continues to run it at a loss is because they hope to monetize it somewhere in the future.
This kind of strategy isn't new (cough XBox cough) and it may or may not pay off somewhere along the line.
Google's acquisition of youtube was a gamble and an attempt to move into yet another market. Will this pay off? Who knows, but one interesting aspect is that Google isn't identifying itself branding-wise with youtube. If they flush youtube down the toilet tomorrow it will have an almost minor impact on the Google brand.
The mathematics of most sites, especially the super large ones like Youtube and Facebook, simply do not add up. They are putting all their chips down on some profitable future that I just don't see materializing. Giving aways tons of content, functionality and bandwith is simply not a sustainable operation. In my opinion, the future of the internet will contain a lot of toll booths or perhaps we will find a better way. These large monolithic web services just does not seem the way to go.
One could easily have said the same thing of Google's search business itself, which gives away tons of functionality and bandwidth.
The real question is whether online advertising will ever bring CPMs in line with TV or print advertising. Personally, when I hear about YouTube not being able to place ads on most of their user-generated videos, it makes me immediately stop and think: is this perhaps just a huge market hole, and a big opportunity for anyone willing to pay for ads on those videos? Are the eyeballs on user-generated videos really worth less than those on professionally produced ones, or is it just that most of the big advertising spenders are nervous about associating themselves with lower quality or potentially offensive content?
Further, are online eyeballs really worth that much less than TV eyeballs?
I still think it's possible that reduced online CPMs may come up naturally over time as more of the entertainment market shifts online. If this is the case there's not much Google can do to help the situation, but it may be worth holding out a while longer to see if things normalize a little bit.
Then again, it's also definitely possible that either a) print/TV advertising rates are a lot higher than they "should" be, and the only reason TV has been able to sustain itself on ads is that advertisers were getting ripped off but couldn't figure out a way to prove it to themselves, or b) online eyeballs are inherently worth less, perhaps due to demographic issues. If either of those is true, free video on the Internet may be fundamentally unsustainable until bandwidth costs come down substantially.
I previously mentioned that if video was just a 9mb file on someone's website, like a HTML page or GIF, it would kill YouTube. I'm not sure about the bandwidth requirements for the more popular sites, though.
video/image hosting is not the point, it's community that counts here: youtube is just a vehicle to deliver your video to more eyeballs, because poking around youtube is easier for the joe user, rather than poking around the 'net. also, as others have pointed out, bandwidth is cheaper for google rather than for the private VPS that hosts your site.
I've seen enough of these stories recently that part of me wonders if Google is deliberately trying to spread the idea that youtube isn't, and can never be, profitable. I can see a few ways in which it would benefit them to have the division perceived as such.
1. Costs >> Since the majority of Google’s costs for the service are pure variable costs of bandwidth and storage, and since they’ve already reached the point at which no greater economies of scale remain, the costs of the business will continue to grow on a linear basis.
Costs of bandwidth and storage will continue to decrease, perhaps geometrically, in the foreseeable future, and so is Youtube operating cost.
2. Monetization There must be other ways to monetize the eyeballs that have not been fleshed out in full yet. It's still very early days. The author assumes that YouTube has mostly run out of ways to monetize through ads (although he did point out a subscription model, esp. for marketers.)