What is a Video View Count on YouTube, Metacafe, Revver, DailyMotion, etc.?
Getting the statistics of how many times your video was viewed can be quite confusing if you do not have a solid definition on what counts as a view: a full view, or just a partial one, or at least the half of it, and what about different IP addresses? If you wish to monitor the success of any of your shared video communications you need to put up with the lack of a transparent and industrially standardized video metrics system currently making both content makers’ and advertisers’ life chaotic. Andrew Baron (one of the RocketBoom guys) is referring to a video analytics company called TubeMogul video study of some of the well-known video sites (both video social networking and video search).
According to TubeMogul, Yahoo Video and Metacafe get the ‘Highly Stringent’ plus YouTube the ‘Very Stringent’ qualifications, Revver and AOL Uncut is considered to be the ‘Least Stringent.’ Oh yes, and Google Video is simply ‘Stringent’ and DailyMotion is ‘Moderate.’ But the Metacafe data is a bit obscure as there was no private setting for the videos: “Metacafe informed us that they count a view if 50% of the video was viewed or at least 30 seconds. However, from our testing, it appears that even multiple full-length views are not counted in full.”
While Metacafe or YouTube only record 1 count per IP address for a fully watched video, the rest of the video sites in the study ignore IP address identification for a full view (YouTube only counts it if it is embedded). As for (less than) half video views, Google, YouTube, MySpace and Yahoo Video do not count it at all, whereas Revver, AOL Uncut, Daily Motion, and Metacafe give it a count. Is it artificial inflation to get some grasp in the growing competition? Andrew is saying “Why would Revver be so slack on a count? The only reason I can think of is hyperbole.” But why pick out Revver, if AOL Uncut shows the same approach towards video analytics? True, Revver got nice publicity about sharing revenues recently, while Uncut is just uncut.
Here’s the summary:
TubeMogul carried out the study back in June (between 6/5/07 and 6/12/07)
I’m the CEO of Aweli, and we have technological solutions to aid in the definition of a "view" of video. Doing so is something I’ve been talking about and working on since late 2003. It has been a core focus, and is the basis of practically everything we do.
If you’re interested in discussing the topic of viewing statistics further, please feel free to contact me.
Very interesting. Makes it very difficult to judge how successful a Youtube etc. viral marketing has been if most viewers watch 99% of a video but some sites still don’t count it.
Thats actually good news. I am one4 of the thousands directors at youtube and i always wondered how many real people were vieuwing my videos. Although my channel started from boredom, im now havin g a lot of fun with it as an outlet and a lot of my movies carry a message. So for me it is important to know if that message reaches the target. A lot of trackers give you a count when someone jusrt clicks on your site, and even when he or she clkicks it away emediatly, its still a count, meanign you know nothing about what it did. This post says it clearly, Even if there were only 18 vieuws, they were at least 18 real vieuws:) Still i have one question, Last sunday i was awarded to be nuber 25 on the list of most vieuwed “regiseurs΅ Wich is the dutch word for directors> since i only have about 2000 vieuws a week, that seems to be to good to be true. So one sugestion i had was that its not 25 of all directors, but only the dutch one’s, or is it really true that there is not so many original material on youtube? Ofcourse i like to have al lot of vieuwers, But im not doing it for the viewcount alone. So im still a bit sceptic about what youtubes criteria for videocounts are. By the way, im dutch but my videos are mostly in english…