Cheaper bandwidth or bust: How Google saved YouTube | Hacker News Reader

Cheaper bandwidth or bust: How Google saved YouTube

arstechnica.com

147 pointssmacktoward55 comments

I still remember the loud howls of mockery when Google paid $1.65 billion for YouTube in 2006.

Mark Cuban said: "Only a moron would buy YouTube" (due to the potential legal liability) and "There is a reason they haven't yet gone public, they haven't sold. It's because they are going to be toasted" and "Would Google be crazy to buy Youtube. No doubt about it. Moronic would be an understatement of a lifetime."

http://blogmaverick.com/2006/10/07/some-thoughts-on-youtube-...

It was almost universally ridiculed as being far too expensive. Analysis was done in every other article about how Google would never be able to pay for the acquisition with YouTube as a business.

Eight years later, it's more than clear that YouTube was a steal for Google at $1.65 billion. It's worth 10+ times that today.

I think it was quite reasonable to doubt youtube had a future.

The bandwidth costs still must be enormous and back then more so. Only a few companies had the clout to negotiate favorable prices for bandwidth.

Even now I am surprised that they are either making so much from ads or pay so little in bandwidth that they can do what Vimeo charges for free.

Google has a lot of settlement free peering, so the bandwidth itself is going to be close to free, although the infrastructure to support it would not be. They also don't have to worry about being present in as many peering locations as most networks of comparable capacity, as their services aren't that latency sensitive, which also helps reduce their infrastructure costs.

> Eight years later, it's more than clear that YouTube was a steal for Google at $1.65 billion. It's worth 10+ times that today.

A 10x valuation still puts it at less than WhatsApp. If WhatsApp is worth $19b, then I'd wager that Youtube should be worth significantly more.

EDIT: I'll expand with a few dot points

- Youtube has an unimaginably huge content library

- Youtube has a solid monetization strategy

- Barrier to entry is significantly higher than a messaging app

- Youtube is even more synonymous with video streaming than WhatsApp is with messaging (anecdotally, my grandparents use Youtube. They don't even have smartphones, let alone WhatsApp).

The problem with having a monetization strategy is then people can see how much money you make. If you aren't making any money and not even trying to make money, the cash investors and buyers throw at you is Monopoly money printed on valuations and total active users.

Most startups would drop in valuation significantly if they made even one penny.

>Most startups would drop in valuation significantly if they made even one penny.

This kind of stuff sounds EXACTLY like what people were saying during the first dot-com bubble. Did you mean that ironically perhaps?

With this and the startup generator posted earlier, I am convinced I am re-living 1998, only at the age of 39 rather than 22...

Not ironic, no. It's just the truth. Do you think WhatsApp or Instagram would be worth billions if they had to make money? With the absolute minuscule amount of profit they might make, their valuation would be worth maybe hundreds of thousands. Not billions.

The reason they're valued so highly isn't because of how much money they would make, but how much value they would bring to an already profitable company. Those valuations aren't based on the idea of profits, but on the idea of being bought out by a bigger player. If they started making money on their own, that value would drop massively.

WhatsApp is not worth $19 billion. That was a defense acquisition, not an RoI based acquisition. Facebook was protecting themselves from losing $19 billion or more of business to WhatsApp or WhatsApp in the hands of a competitor (like google).

I think this might be a bit of a single cause fallacy just like the other poster. Only Facebook knows why they payed what they did, but it was certainly more complex than just preventing Google from buying them.

The aggregate benefit of not having WhatsApp in a competitor's hand + the benefits to their core business of owning it + the lifetime profitability of WhatsApp + some other factors I'm probably forgetting <= a value of 19B in mostly Facebook stock.

> some other factors I'm probably forgetting

The verified phone numbers of 800 million people [1] and the ability to act as a gateway drug to other Facebook services [2]?

[1] https://en.wikipedia.org/wiki/WhatsApp

[2] https://en.wikipedia.org/wiki/WhatsApp#Acquisition_by_Facebo...

A company is worth exactly what someone is willing to pay for it, nothing less, nothing more. That's the best evaluation that exists.

The motivation for thart valuation is still relevant though.

I specifically went with 10+ just because beyond 10x I feel that it becomes increasingly difficult to be certain. I too believe it's greater than 10, but how much so ... I have no idea.

YouTube has revenue about ~20% less than Netflix, and is supposedly profitable according to articles from late 2013. Netflix is worth $33 billion, and has something like 57 million paying subscribers.

I believe the paying subscriber base, ie a superior business, gives Netflix a valuation edge, along with higher sales. Netflix is also rapidly building large international subscriber base that will be incredibly valuable over time, to go with their custom content (DareDevil, House of Cards, etc).

10x for YouTube would be $16.5 billion. YouTube isn't worth dramatically more than that range, the ceiling is clearly Netflix. So $15 - $25 billion is fair imo as a broad range.

What makes the barrier to entry higher for video than messaging? I would expect the opposite, by a good margin. For messaging, you get strong network effects, and it's useless unless the people you want to contact are signed up for the service. For video, you can upload anywhere, send or post the link, and they can watch it. Obviously that isn't actually happening, so I'm curious what I've missed. Or is it just because YouTube is pretty good?

Yep. The niches carved out by vimeo and liveleak are interesting in their own right.

But I mostly wanted to say, youtube is pretty fucking great.

There are network effects among video, although not as strong. How do people find what they want to watch? Both the search feature and "related videos" are important here.

How much of that is due to Google helping them out though? Could Google have launched a competitor for say, $0.5BN? So long another big player didn't buy them up and do well, we could easily be laughing how they wanted 1.6 and Google only needed a third of that to win.

But I suppose it doesn't matter in the big scheme of things. Why take the risk of someone buying YT and maybe getting more traction? A billion dollars doesn't matter enough, long term.

Or look at Groupon, which hasn't really done better than Google's $6BN offer - that didn't work out spectacularly for Groupon. Imagine if Google used a few billion to get people using its deals site. They could spend millions a day subsidizing deals for a year or two just to build traction. Yet they still offered 6BN.

Google did have a competitor called Google Video which was failing horrendously.

One of the huge downsides of Google Video from a consumer side was that videos took 3 - 5 days to upload as they were proactively being checked for copyright infringement. Google probably never could have gotten away with the laissez faire attitude towards copyright infringement that a young startup like Youtube which is why it made sense for them to buy it, regardless of the price.

"Only a moron would buy YouTube" - Marc Cuban

Let's make sure that quote is never forgotten.

why?

Yes money cant buy you a brain of your own. Poor Cuban.

Says the man who would go on to invest in "I Want to Draw a Cat For You!"

Hey, leave me out of this.

I love your work, but I don't think Cuban was playing the hard-nosed businessman when he invested; rather he saw something in your idea that just might turn a profit.

I think it's rather the same with Google and YT. Google saw that video was going to be big on the Web but their Google Video experiment was suboptimal in implementation They saw that YT did it better and saw a future of integrating that technology with Google's other services, and it paid off.

I find Mark Cuban's take back then — and now, he still believes YouTube is a loser as recently as 2015 — particularly funny because selling Broadcast.com to Yahoo for $5.7 billion is how he made his billions. YouTube is soaring whereas Yahoo's music and video offerings are almost non-existent. Maybe he was biased by how bad his own acquisition went for Yahoo.

Or maybe he thought that he had suckered Yahoo every bit as much as he was thinking YouTube had suckered Google.

I didn't think it was a good idea either. I though google would get bogged down by DMCA request and having to pay the big name content makers. I also though once it became only "legal" content everyone would just move on to the next video sharing site.

If I remember correctly, it was a way for the investors to cash out on their Google value indirectly. The same VCs (Sequoia I think) that funded Google had funded Youtube.

It's been a long time so I may be remembering incorrectly. OK, I just looked up the deal and it was done in stock and not cash. Sequoia invested $9M and received $516M as part of the deal.

I did see an article saying Sequoia planned to sell their Youtube acquired Google shares in 2007 (before the lock-out period for employees expires) but I'm not sure if the sale went through.

http://www.nytimes.com/2006/10/10/technology/10payday.html?_...

The fact that Google buoyed Youtube's sinking boat by plugging the leaks with money does not mean that Youtube shouldn't have collapsed under it's own weight and been allowed to die. Maybe then, we would've gotten a free, open, p2p alternative out of the ashes. Now we have DMCA takedowns, region-locked videos, and unskippable commercials.

The sooner someone disrupts youtube, the better the world will be.

Let's not forget YouTube can also arbitrarily decide what video sizes competing browsers are allowed to receive, or what DRM system they should have implemented.

Good luck shipping something that doesn't work with YouTube. I'd say their position of power is greater than Netflix.

Killing youtube is not solution for that, you can wish Facebook etc will create genuine competition.

DMCA takedowns and region locks are a legal/social problem, not a technological one. Look at what happened to PopcornTime.

You can still self-host your own videos, but then it's hard to monetize them to pay for the bandwidth. Remember that this is a "two sided market": you have to bring viewers and video producers to it.

And the commercials are blocked successfully by AdBlock.

DMCA takedowns and region locks are a legal/social problem, not a technological one.

That seems like oversimplifying matters. The way Google automatically identifies infringing content is very much a technological issue, as well as the vast problems that it causes when it's wrong.

Given that legally they must remove infringing content, what other choice do they have but to automate the identification process? And how often is it wrong?

The DMCA specifically removes any obligation they would have to actively police user-submitted content, so it's simply not true that they legally must do so.

That's right, but the scale at which infringing content is uploaded to YouTube can make it possible for somebody to make a case that the DMCA doesn't quite cut it. I don't think anyone wants that.

Because even if you assume DMCA is sufficient, at this scale what is the alternative? The content owners would essentially have to implement their own content ID like system to find infringing content (which they already do to some extent), but rather than having hundreds of different stakeholders implement a hundred different systems (or the same system hundreds of times) externally, it's simply more efficient to centralize it at the point where the content is stored. Not to mention that the more likely outcome would be a massive lawsuit rather than content creators bothering to implement this system on their own.

Also, if YouTube wants to make money, let's face it, they need content people really want, which is what the big content creators own. Pretty sure they have this content ID system to keep those folks happy.

The issues are still less technical and more legal or business-related.

> The content owners would essentially have to implement their own content ID like system to find infringing content (which they already do to some extent), but rather than having hundreds of different stakeholders implement a hundred different systems (or the same system hundreds of times) externally, it's simply more efficient to centralize it at the point where the content is stored.

There's a big difference between the claim that content owners would do something similar anyway to provide a basis for DMCA takedown notices and that it is more efficient for Google to do it for them and the claim that it is legally mandatory for Google to do it.

> Also, if YouTube wants to make money, let's face it, they need content people really want, which is what the big content creators own. Pretty sure they have this content ID system to keep those folks happy.

There's also a big difference between the claim that it is beneficial to YouTube in terms of positive relations with people it wants to do business with and the claim that it is legally mandatory for them to do it.

Ah I see, my original comment said "Legally they must...". Yes, that was not correctly worded.

If you agree they have to automate the process, then haven't you just agreed it's a technological problem?

I'd say it's a legal and business problem, only the solution happens to be technological because of the scale involved.

Also, downloading of Youtube content is still not possible officially.

It is almost like another big company who refused, for years, to implement a PDF export function into their flagship word-processor product.

Refused?? Adobe straight up threatened to go antitrust on their ass if they did: http://arstechnica.com/uncategorized/2006/06/6973-2/

> Maybe then, we would've gotten a free, open, p2p alternative out of the ashes.

P2P is not the answer. It may work for the 1% (popular videos), but for the 99% it does not provide a viable alternative for on-demand content... I don't want to wait 5 hours for a 3 year old how-to video to stream from a single seed with limited upload capacity :(.

E.g. Look at torrents. Popular ones are readily available, but obscure or old torrents will be lucky to have a single seed.

What percentage of the bandwidth is consumed by popular videos, and what percentage by the long tail? P2P can alleviate a big chunk of the popular content bandwidth. Not sure how much that is, though.

In a P2P world, if you're uploading a long tail video of your cat that only your family will watch, maybe it's up to you to "seed that torrent" - or more realistically, keep your machine running so that your peer can serve the content to theirs. It's the move from desktops to mobile and laptop devices that run only sporadically that is the real problem for P2P.

Someone can't "disrupt youtube" until costs come down to the point that Google's search/ad cash cow isn't subsidizing YouTube anymore.

At some point, hosting video yourself on an aws instance will be feasible. Until then, we're probably stuck with YouTube.

The fact that you keep saying "you can't" is why you're not creating a startup to do exactly that.

TPB was floating this idea without funding almost a decade ago. it didn't take off because Cohen's Bittorrent Sync was implied to be open, and the project died in anticipation of the better (standard/trusted) technology.

Pretty interesting stuff. I do wonder how much internal collaboration there is on topics like HTML5 video & WebM. The benefits of owning end-to-end both browser and server for driving media codec development seem enormous.

It makes Google's decision to launch Chrome in 2008 much more of an obvious call.

I had a coworker who used to work for a direct competitor of Youtube. He left that competitor (as they were struggling) as Youtube was taking off. His older company had been founded by a Hollywood exec/veteran, meaning he made sure to keep out copyrighted material.

The way he put it, he said Youtube got big Because Youtube initially allowed copyrighted material.

This led Youtube gaining more videos -> more traffic -> more videos -> repeat.

Almost every day I watch something on Youtube that breaks copyright law and think that society is better off as a result, so I think this is a positive outcome. I wouldn't sacrifice all the old adverts, rare blues music etc. just so that big corps can have a better bargaining position over their copyrights.

This sort of thinking only encourages more breaking of copyright law, the vast majority of which is of content that is neither rare nor in danger of being lost. I don't think society is better off for that, as it simply exacerbates this specific freerider problem. "Big corps" (and small creators) are investing their resources into providing content that you like? What are you giving back in return?

Possibly, the content is system recognizes actual owner and add revenue share is going to him/her and not there uploader

Seems a bit whitewashed by not making any mention of all the youtube internal emails turned up in viacom's discovery with youtube staff participating in, aiding the infringement (including, e.g. removing report functionality), and saying the infringement was integral to the success of their business; But it mentioned viacom's not-completely-clean hands. (Though there was nothing unlawful about viacom uploading their own stuff to youtube...)

> (Though there was nothing unlawful about viacom uploading their own stuff to youtube...)

I dunno, sounds kinda like entrapment or framing to me...

Defined saved; the site is still losing money.

Reply on news.ycombinator.com