Music is My Hot Hot Sex
agency, TBWA/Chiat/Day. The agency liked it so much that they contacted Nick Haley and recruited him to create a television quality version. The commercial was shown during the World Series and propelled the song and ClarusBartel's video into the mainstream.
Although this does explain millions of views, YouTube users don't believe that this alone accounts for the unprecedented rise of Bartel's video. They point out that while Bartel's video has almost 100 million views, it has fewer than 5,000 comments and has been favorited less than 11,000 times. This appears very suspicious when compared to Evolution of Dance, which has been viewed almost 80 million times and has almost 125,000 comments and has been favorited almost 400,000 times. Many YouTube users attacked Bartel's video, forming a mobocracy of sorts, and started accusing Bartel of hacking YouTube.
YouTube users rebel by trying to lower the video's rating. Bartel sees this and disables ratings.YouTube users rebel by trying to lower the video's rating. Bartel sees this and disables ratings.
Angry YouTube users leave comments of disbelief and anger.Angry YouTube users leave comments of disbelief and anger.
Andy Baio at waxy.org interviewed Bartel shortly after YouTube refreshed its leaderboards. When asked about the possibility of gaming the YouTube system, Bartel replied, “I get constant messages accusing me of being a hacker, when I barely know how to turn a computer on and off. Some people call me a genius, because I beat the system. Everyone is free to imagine me as they choose, however they prefer.” Given the fact that he is a semi-prominent music blogger, it appears he is understating his computer skills. Luckily for us, he wasn't able to attain this amazing feat without unintentionally unveiling some valuable insider marketing tactics.
When a user searches on YouTube, the results are determined by the relevancy of the user's query with the combination of the title, tags, and description of a given video. By manipulating these variables, it is easy to craft YouTube search results. This seems to be the case with this video. Below are the tags chosen by Bartel.
cansei | de | ser | sexy | Music | Ipod | touch | commercial | hot | sex | boyfriend | music | apple | nick | haley | computer | Barack | Obama
While the first ten tags are legitimate, the remaining eight were chosen to increase the video's visibility. The related terms, 'Nick' and 'Haley', are obviously being used to draw views away from the original ad's creator. 'Boyfriend', 'Computer', 'Barack', and 'Obama' are blatantly used to get traffic from highly searched terms. This approach is seen again with previous titles of the same video. For example, a week before the video hit number one it was titled “VOTE OBAMA 08 CANSEI DE SER SEXY Music is My Hot Hot Sex.” This tactic of keyword stuffing in the title was used in conjunction with a “Vote Obama” image displayed next to the video (ClarusBartel's profile picture).
Although understanding Bartel's tactics are enlightening, exploring the fundamental flaws that he exposed are what is really interesting. On YouTube, a video is graded by comments, ratings, and views. Unfortunately, all three of these criteria have become ineffective due to misuse.
In the vast majority of cases, comments are such low quality that they add virtually no value to the video page.
Useless YouTube CommentsUseless YouTube Comments
Star-ratings have also become useless. Without a single point of view, these ratings lack a reference point and therefore provide little value. My vote of one star as a video viewer is weighed equally with the content provider's mother who rates the video as five stars. The problem is worsened by the fact that the amount of times videos have been rated is not visible in YouTube's SERPs (Search Engine Result Pages). When searching, there is no way to see if a rating is based on one vote or one million votes.
YouTube videos with the same ratings while having wildly different amount of views. The amount of times the video has been rated is not shown. Thus, the ratings are useless.YouTube videos with the same ratings while having wildly different amount of views. The amount of times the video has been rated is not shown. Thus, the ratings are useless.
This leaves the users with only the number of views to judge the quality of a video before watching it.
YouTube's confusing SERPs.Views is the only way to gauge a video's worth before watching it.YouTube's confusing SERPs.Views is the only way to gauge a video's worth before watching it.
The views, however, are the easiest to manipulate. YouTube's video views are based off of page views rather than page visits. This means that if the content provider simply refreshes his or her video over and over again, the video will start to appear more worthy of clicks on the YouTube's SERPs. Although the views by the content creator are not always counted at a one to one ratio, it is still extremely easy to boost one's own video. This is not immediately obvious to many people because the view numbers do not update in real time, but it absolutely does work.
The tendency for users to click the most viewed video in the SERPs combined with how easy it is to generate fake views could be linked to the rapid success of Bartel's video. (Note: I am not saying he did this, I am merely explaining how it could have been done given the circumstances of his video's fast success.) Imagine if a devious user chose to manipulate video rankings on YouTube. One could potentially utilize the high traffic at social media sites to provide almost limitless views from different IP addresses. Let's say an article that went popular on Digg contained a 1 pixel iframe of a video page on YouTube. Anyone who viewed that article would unintentionally add views to that video. This practice would explain how Bartel's video received so many views in such a short period of time while receiving so few comments and links. This idea is not new and has already been used with variations on MySpace (see Samy, and thank you to DangerMouse for the clarification). It is naive to think that YouTube would not also be targeted by a similar attack.
To counter this, YouTube engineers need to implement a real ranking system. Just like the SERPs on modern search engines, the SERPs on YouTube should display the most useful and related results first. To achieve this, the new ranking system should be based on the following criteria:
Intelligent User Rankings
By basing search results on these, the best videos would be returned rather than the most manipulated. Google (YouTube's owner) simply needs to do what it does best: use intelligent signals and patterns to determine algorithmically what the best results are.
At the time of posting, YouTube has already made the first step. While ClarusBartel's video is still credited with having almost 90 million views, it has been removed from the most views of all timelist. The unwashed masses of YouTube are once again calm, as the Evolution of Dance has regained its throne. Hopefully this will be a big enough lesson to persuade YouTube to implement the necessary changes. If not, we are sure to see many more videos wrongfully move up the leaderboards and detract from the value of YouTube.