Blog Post

YouTube Not Built on Big Media’s Back?

So maybe YouTube really is about the long tail, the little guy, and the lonely girl. Videos removed at the request of copyright owners accounted for just 5.93 percent of the site’s 6,725 most-viewed videos over the last three months, according to a study by video-tracking site Vidmeter (see summary, full report).

Four times a day, Vidmeter checks YouTube for the 100 most viewed videos of all time, the month, the week, and the day. For the purposes of the study, it compiled a list of URLs where videos had been replaced with a takedown notice alert. The disappeared videos accounted for 9.23 percent of top videos and 5.93 percent of 94,187,203 views of these videos — not anywhere near the majority.

Viacom famously demanded YouTube to take down 100,000 copyrighted clips, and later claimed 160,000 clips seen 1.5 billion times had been pirated when it sued Google for more than $1 billion dollars. According to Vidmeter, just 72 of Viacom’s taken down clips had made it to the most-viewed videos list — 1.07 percent of the top videos accounting for 2.37 percent of views. The most-affected copyright holder was reportedly actually Time Warner, with 93 clips.

In many cases, according to Vidmeter founder Bri Holt, multiple companies claimed to own the copyright on a video. Viacom, Sony, and the RIAA all took it upon themselves to demand removal of the music video “Girlfriend” by Avril Lavigne.

Presumably Viacom took responsibility because the clip aired on MTV or VH1, Sony because Lavigne is signed to Sony BMG, and the RIAA because it likes to take action against music piracy. “Even these major copyright holder are confused as to who should do what with this!” commented Holt in a phone interview Tuesday.

21 Responses to “YouTube Not Built on Big Media’s Back?”

  1. It is surprising how few copyrighted clips are out there on youtube and it goes to show the justification of apple in removing its DRM. If music could be sold for less than the equilvalent of a days pay, as they do on itunes, then people would buy their products.

  2. Liz,

    I agree totally. I would really like a lot more detailed usage data about that remaining 90%. How much of the non-copyright content is viewed by significant numbers or are we talking about small personal groups. Groups that would be private if the wide open internet were not the medium. Family & Friends in other words.

    Can family and friends be considered a long tail niche is my main question. I’m not sure it should be. But, I’m certainly ready to change my opinion based on better data.

  3. Liz Gannes

    This study is hardly the final word, but I do think it’s meaningful. Ideally this would be for all of YouTube and for all of the last year, prior to the Google purchase when you could find nearly anything you wanted on YouTube.

    A bigger sample would bring in things like the Daily Show, where lots of bits and pieces get posted all the time, but the number of views per clip wouldn’t approach something timeless like SNL’s Dick in a Box.

    As for concerns about ‘the long tail,’ I do think YouTube still addresses the needs of people who are trying to find an audience, however small it might be. doug, I don’t think this necessarily means that the rest of the clips on the site are solely personal, because personal to me means private. I would love to know how many clips have a total of one view, though.

  4. jlewin


    Since you’re hyping this study – does it meant that you think the long tail isn’t relevant to YouTube?

    Some questions you should be asking:

    Is this a representative sampling?

    Did Vidmeter capture a significant portion of the videos on YouTube?

    Do you really think the removed video list represents the traffic generated by copyrighted content?

    The ultimate question, though, is whether or not content creators should be able to control who tacks adds onto their work, – the Vidmeter info seems irrelevant.

  5. Haven’t read the Vidmeter report yet, but the conclusions seems potentially skewed:
    1. By definition, if a video is taken down then it won’t get as many hits, and viewers will go to other sites to view them. So much of the data could be comparing hits on videos that are up for months compared to videos up for days (before they are taken down).
    2. Only a small percentage of overall big media videos are requested to be taken down. There is plenty of big media in the top 100 for any given day that is not requested to be taken down, and it seems like those videos would fall into the “non big media” bucket in these stats?

  6. The breakdown of video views between the studios is fascinating. Viacom the acknowledged follower in the rush to the internets, has nearly double the views of any other studio. I’m going to go through the full report to verify a few facts. But, it seems the laggard has the most popular content (by views).

    This data is one of the most telling indictments of their bungled web strategy. It also tells the tale of a sleeping giant. Well, maybe not giant, but it looks like this kid is twice as big as the other media companies with the YouTube audience.

    Big media’s 9.23/5.93% of content is nowhere near the size they thought they were (or pretended to be). But as far as long tail, I’m not sure if it applies unless self consumption is a part of the tail. YouTube seems to be mostly personal stuff once you discount the few vlog celebrities. The Long Tail seemed to be a graphic description of a chain of niche markets. Can the cam kid content on YouTube, posted for embedding in profile/blog pages be considered niche media? At what point do you say, this is not about niche markets, but personal media?

    Now I know some will say that as long as you have a group, no matter how small, consuming the content it’s a niche market. I understand that the point of the long tail is to advocate the value of the small in aggregate. But it just seems to me that sharing sites like YouTube and Flickr are not about feeding niche markets. It’s personal media (photo albums and home movies) that you share with your friends and due to the technology used a lot of other people can see it if they happen to run across it.

    I can see, and totally agree with the long tail as a model for content businesses. The examples of Netflix and Amazon are dead on. Those business models are what the entertainment industry needs desperately. But to apply the long tail in online social environments is like saying the phone system is a long tail because no matter how obscure I may be, there is somebody out there interested in a conversation with someone just like me. (A very small niche indeed.) The long tail doesn’t work as a useful model for personal sharing nearly as well as it works for niche selling.

    You can look at everything as a long tail if you stare at it long enough and are willing to call any number greater than one a niche.