Youtube remains unprofitable 3 years after Google acquisition
dealbook.blogs.nytimes.comHe cited research from Credit Suisse
Seriously? People are still citing that hack-job of an analysis? It didn't even take peering into account - it was ridiculous.
It's true across the tech community; someone can post a completely wrong analysis of pretty much anything and as long as it fits peoples' expectations, they'll keep citing it even if it's wrong, and it's incredibly aggravating.
Some common examples include Windows having a lower TCO than Linux, Theora being better than H.264, and of course practically anything in the realm of politics.
I think this is getting more and more common because of so-called "tech" bloggers not fact-checking their stories and generally overhyping social media (twitter/facebook/youtube etc..) because of their relative simplicity compared to something like the intracacies of bandwidth economics.
Unlike traditional journalists (though clearly not the NYT!) these guys don't have black books of knowledgable people to call up and ask on a certain issue.
Social media makes everyone's opinions and views a lot more equal (when really they shouldn't be) and these stories eventually bubble up to traditional media, which have the view that since the whole "blogosphere" is talking about it as fact; it must be fact.
Not sure why this guy is being downvoted. He's right.
video search isn't big today, but it'll probably be huge in the future, may be as big as text search if not bigger. same can be said about video advertising. at that time, who do you think is going to have tons of inventory at their disposal? Google
it is just part of a bigger strategy - text search, book search, audio/video search etc.
youtube is a long term investment. as long as G has money to keep it running/improving, they have nothing to worry about.
may be 5-10 years from now, the same ppl would say that the best investment that google made, is youtube.
Video search isn't big today because there is exactly ONE use case for it: [heroes season 2]. Text search works perfectly fine for that -- you type in [heroes season 2], you will find Heroes, season 2. It is probably one of the best single domains for searching -- canonical or near canonical names for everything, with disambiguating numbers that users actually know, and built-in metadata which links all your canonical names together in a pattern users actually understand! Brilliant!
OK, so we've got Heroes season 2. Now we have the problem: do people want to pay money for it? Answer: well, no, not if they can get it free on Youtube. Piracy is Youtube's killer app. Unlike the iPod (piracy is the iPod's killer app, too, unless you think 20-somethings are filling 8 GB iPods at $1 a song), Youtube doesn't have plausible deniability or a convenient way to extract money from the pirates.
Um, what? You really think the only use case for video search is finding particular clips with known titles? There's an entire industry around stock photos/videos that right now relies almost entirely on manual tagging, you don't think that field's ripe for search? Or the petabytes of video being streamed from security cameras, you don't think anyone out there wants to search that for faces or license plates or whatever? Video search is going to be absolutely killer, and Google's sitting on top of the largest, most varied training corpus in the world, thanks to YouTube.
what if I am searching for a particular scene (say a 2 min scene), in a particular episode of heroes, in a particular season?
How can you automatically detect that from a video? And if instead of automatically detecting it, you just let the users tag the video, why does text search not work? Finally, even if they manage to get a useful video search working, why do they need Youtube for it, when they can crawl the web for video?
Google is working on this- One theory is that part of why they are working on Goog411 is to increase their voice reco engine, so that they can apply it to Video and Audio search.
Google hired several engineers away from Nuance (one of the leaders in voice recognition), and they're working hard on growing their development there.
How can you automatically detect that from a video?
that is the point I'm trying to make. it is not possible today, I guess it will be possible in future.
how much can you tag? and who will do all the work of tagging? it has to be automated, like text search. video search is going to be much more difficult than text, and G has tons of videos to mine and refine algos.
One thing you can search is closed captioning, like SnapStream does:
YouTube has also added a closed captioning feature, but I'm not sure how it is done. I assume this is also working towards video search.
Important point about the cost of the original acquisition:
"Even so, it’s important to remember that Google paid for YouTube in stock, not cash, that represented a tiny fraction of the company’s total market capitalization. (The deal was initially valued at $1.65 billion.) And immediately after the merger was announced, Google’s shares rose, which in some sense seemed to pay for the deal on its own."
This probably doesn't surprise anyone but their current business model is terrible. Why would anyone click on a text based ad which distracts them from the video they want to watch. I don't see why anyone would voluntarily do that. At some point google is going to have to bite the bullet and have video ads. Even if they are on after a video is shown they would be far more effective and desirable for advertisers.
Because, if youtube is ever profitable, they will be sued. over and over until the end of time.
much better to just miss breaking even. Shovel all the user data into doubleclick, and all the video data into crazy data mining algorithms.
On the surface, your argument has intuitive merit. However, you don't sue a business unit (Youtube), you sue a company (Google). Google is profitable.
On the surface, your argument has intuitive merit. However, youtube isn't a business unit. it's a subsidiary.
http://googleblog.blogspot.com/2007/01/look-ahead-at-google-... http://news.cnet.com/8301-10784_3-6153267-7.html
edit: sorry, that was sorta snarky. and things may have changed since the viacom lawsuit. (it was a subsidiary in late 2008)
Touche! Didn't do my homework on this one.
it is not direct profits but it is indirect profits and strategic intentions as the reason google bought it. google wants you to use its services, the more you do the more you depend on them. video, voice, data, etc. they make profit as a whole, so you cant look at one unit and say it doesnt bring in profit. the longer you associate google with your video, voice and data needs they win.
Actually, the more you use Google's free services the more you cost them. It's the non-techies who are clicking their ads which are making money for Google so, for all their technical marvels, Google's revenue source(s) are pretty narrowly defined.
I can almost picture Steve Ballmer slapping his cheeks with a mocking "Oh, no! Google's dominating yet another money-sucking cost pit! Whatever shall we do?!?!" before bursting out in laughter.
I think Ballmer stopped laughing at google sometime between them crawling out of their garage 10 years ago and them approaching MSFTs market cap today.
Until GOOG approaches MSFT's revenues, I don't think Microsoft will worry. If there's a 'war' between those two companies, Microsoft is probably handling it as one of attrition.
Google has one strong source of revenue (search advertising) and has been thrashing about for years in an attempt to build at least a second source without notable success.
Microsoft has many different sources of revenue and continues to build on each. To take out Google, all they have to do is have Bing, Yahoo and the rest patiently chip away at GOOG's search advertising dominance by a few percentage points each year and that revenue will dry up. Then, without that revenue, all of Google's vanity projects will collapse in a heap.
Google's many projects remind me of the Flying Lawnmower: http://www.youtube.com/watch?v=kNWfqVWC2KI
You can make anything fly if you have a strong enough engine shoving it and Google's engine is its search advertising revenue. Once that revenue decreases enough, their lawnmower will fall to earth, and Microsoft is surely happy to see them tack on new expensive doodads like YouTube and Chrome O/S to hurry that day forward.
Somehow the anecdotes of Ballmer throwing chairs and threatening to "fucking kill" Google run counter to your speculation. Either Microsoft is getting worked up over a non-threat or Google is more dangerous than you think.
Have you seen youtube comments? They probably are coming from the non-techies.
...and yet, populated with hordes of non-techies, Google still hasn't been able to move YouTube from substantial loss to minimal profit (even when the world was in the middle of a "Web 2.0" tech boom!).
Now that we're in a bust, what does that tell you about their future prospects?
exactly. how many of MS products are profitable? every successful company has a handful of ultra profitable/successful products, and a bunch of average ones, some that take years to start making profit (as will probably be the case in youtube) and many products that may never make any profit. google is just making sure they get into as many online/mobile biz as possible.
Google may be playing up the notion that YouTube is very unprofitable. Making VCs think that online video is a bad business will help reduce new competition. And, unlike eBay's Skype acquisition, there's probably little shareholder pressure for Google to prove that it made a good choice.
My general calculation on YouTube is this:
-At Google Search, you do a search and get a text ad. Each text ad is about 400 bytes, each search page (with your results) is around 5k. Serving up the webpage is done from another site.
-At Google YouTube, you do a video search and get a text ad alongside it. Each text ad is 400 bytes, each search result you look at (eg, video) is about 2MB.
So, are you going to make as much, after paying for bandwidth, with a factor of 400 (2,000,000 / 5000) difference in bandwidth?
It's been a while since I learned something new from a nytimes tech article.