Slashdot Mirror
YouTube Gets 1.8 Billion Logged-in Viewers Monthly (engadget.com)
On stage today at Radio City Music Hall, YouTube CEO Susan Wojcicki made a surprising revelation: the service gets 1.8 billion logged-in viewers every month. And that doesn't include people who aren't logged in -- which means the actual number of people watching YouTube is definitely much higher. From a report: Last June, the service had 1.5 billion logged-in watchers. On TVs alone, people are now watching 150 million hours of YouTube every day. The latest figures are yet another sign that YouTube's reach is staggering, something that Wojcicki wanted to make crystal clear for the audience of advertisers and potential partners at its annual BrandCast event.
It'd have zero viewers.
1.25 billion bots inflating view counts, spamming comments and uploading bogus content.
I try to avoid Google tracking as much as possible, as such when using YouTube I can easily see what effects their algorithms have on recommended videos as it is stark contrast with defaults recommendations for when they know nothing about you.
Almost universally, they suggest extreme versions of content you just searched for. For example, searching anything related to climate gets you to doomsday, the end is next year videos. Searching anything related to religion lands you on most radical versions of that content. Searching anything political lands you into directly into conspiracy videos.
Everyone I know has tried Youtube at least once and usually more. Bandwidth wise it probably has some of the heaviest usage on our Internet usage aside from Torrenting. YouTube is basically the go to place for virtually anyone who wants to view a quick video for laughs, repairs or even documentaries. I probably watch more YouTube than I do regular TV now.
Put another way, the average is 0.1 hours per day per viewer, or 0.7 hours per week per viewer. Either their reach is greatly exaggerated or they don't have nearly as good engagement as network TV. For comparison, an "average" american watches 5 hours of TV each day (http://www.nydailynews.com/life-style/average-american-watches-5-hours-tv-day-article-1.1711954). Yeah, that's probably dropped a good bit, but I suspect it is still quit a bit more than 0.1 hours per day...
Adblock Plus + http://someonewhocares.org/hos... make sure of that. If they can't connect connect to their CDNs they can't serve ads. Simple. I only see the "real face" of the Internet when someone asks me to fix their computer but that doesn't happen much anymore. Thank deity_of_choice.
youtube-dl downloads?
I get bored sometimes and watch Russian Car Crash videos of the month...
I wonder how much of their ad revenue they give back to providers of that content. I'm not necessarily talking about up-loaders, but content creators. Because without them there would be no audience. I watch music videos mostly. I add stuff to my playlist so I can click and fall asleep to it. I'm constantly running across videos that have been taken down. If youtube was fairly treating these content providers rather than trying to see what they can get away with the service would be more valuable, the people who actually create the content would be happy and this might have a viable future.
I don't believe in karma, I just call it like I see it.
What's the advantage of being a logged in viewer? Yet another reimplementation of bookmarks?
There was a comment from an anonymous Google/YouTube employee admitting that their recommendation algorithm has "gotten away" from their control. Its job is to increase viewership and it does that by suggesting videos that its systems consider most relevant. Unfortunately, it's is a mathematical judgement, not a values judgement. This is why so many videos being suggested are extreme: they get lots of views (which suggests to the algo that it's a "good video") and but they're often distasteful.
I can't say for sure how Google's tracking works, but I know how mine works, and we can be pretty sure that Google's is better than mine. Mine would "recognize" you pretty well.
Basically, you'd be lumped in with all of the other people who use that exact same VM, on the ISP, in the same area, and visit the same sites - basically just you. :) Here's a bit about how it works.
First we have your IP address/24, which puts you in a group of a couple hundred people. We ignore the last octet of the IP because that changes. We also record IP/20.
Next up we have the User Agent, which looks like this:
Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/41.0.2228.0 Safari/537.36
Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/41.0.2227.1 Safari/537.36
Mozilla/5.0 (Windows NT 6.4; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/41.0.2225.0 Safari/537.36
So if you're on Windows 7, that's going to separate you from Windows 8 and Windows 10 users, etc. Note also the very fine-grained Chrome build number. Not just version 41, but version 41.0.2227.1 for one person, someone else is on 41.0.2225.0, etc.
Here's the gold mine mine right here. The Accept header. Check out the most recent person trying to log in to the site I checked:
Accept: image/gif, image/jpeg, image/pjpeg, application/x-ms-application,
application/vnd.ms-xpsdocument, application/xaml+xml,
application/x-ms-xbap, application/x-shockwave-flash,
application/x-silverlight-2-b2, application/x-silverlight,
application/vnd.ms-excel, application/vnd.ms-powerpoint,
application/msword, */*
I can distinguish between him and someone with the EXACT same computer image, but who hasn't installed Silverlight. There are about 50 different applications that can show up in the Accept header. The exact mix of applications you have installed on that VM really help distinguish between you and the only other person in your neighborhood, on your ISP, who behaves similarly to you, using a fresh VM and not logging in.
Of course you're talking about being tracked by the company who made your browser, so they see you visit Slashdot twice a day, and whatever other sites you visit. that, combined with everything else above, distinguishes you from pretty much anyone else.
That's not even getting into cookies, Flash cookies, local storage, super cookies, etc.
I may have been a little unclear about one point. You might ask how does that identify me. It doesn't give them your name of course, but Google doesn't care about your name.
They put you in a group, and profile that group. The smaller the group, the more specific the profile, the better that profile represents exactly you.
Imagine an "advertiser might say "Mac users tend to buy _____, so show these ads to Mac users". Then we start to get more specific, "Mac users in North East Chicago. How about "Mac users in Northeast Chicago who have a Microsoft Office 2010 installed on their Mac Pro", what ads should they see?
Your group is something like "Windows 7 users in Northeast Chicago who visit Slashdot and Breitbart multiple times per day and have a Microsoft Office 2010 installed, don't have Java but do have Flash 8.06b, and use DSL and use Chrome 41.0473.83.1 and don't log in, and still run 1024X768 resolution and launch a freshly installed copy of the VM every day and have ...". They've profiled that group of people. There are about 1 members of that group. That group is you, and maybe one other person like you.
I use these techniques to catch bad guys trying to log in using your user name and password. If someone on a Mac in California claims to be you, but just a few hours ago I saw you still in Chicago on your Windows 7 VM, I'm going to be awfully suspicious of the dude in California who claims to be you.
The problem is that many of these views amount to "OMG, what is this shit, anyway?
People need to be way more aggressive on the thumbs down button after sizing up the moral car wreck.
I also find the "not interested" button works wonders after some initial persistence.
Google would fix something in the larger scheme if the viewing population normalized their rating habits to a 66% baseline rejection rate. Everyone should just post "thumbs down for tasteless recommendation" in the comments thread, until these comments drown out everything else. Google would learn fast were the adverse signal even mildly organized.
the service gets 1.8 billion logged-in viewers every month. And that doesn't include people who aren't logged in
Thanks for the tautology, genius.
If the YT Algorithm is so smart, then why does it keep suggesting the same 200 videos from the same 20 channels over and over again?
YT boasts of the millions of hours of content uploaded each and every day. Where IS this stuff? There's no way to search for it. It's effectively invisible.
The YT Algorithm is hardly smart. If you watch ONE video from a prolific YouTuber, your entire recommendations will be poisoned with crap from that same channel for weeks.