In order to test the validity of the constructs in the proposed model of innate viral potential and to understand the interrelation of those constructs, I wanted collect data on the characteristics that were common in viral videos. The gathered data should allow the use of confirmatory data analysis of the variables.
Although the assessment of a video’s content quality can never be completely objective (because it is done by human beings and not by a computer), it is possible to make reasonable judgments regarding clear concepts such as happiness or sadness, on a video’s ability to stimulate the senses, or its ability to involve the viewer, if we base our judgment on the reactions of a large heterogenic test group. Generally speaking, human beings share certain attitudes and behaviors no matter what their background is, such as the need to be social (by varying degrees) or to share information. Therefore, this research uses an analysis of viewer responses on selected videos in an effort to evaluate the seven variables proposed in the model of innate viral potential.
The selection of the videos to be tested depended on three main criteria: number of views, viral status, and diversity. Since the goal was to find common success factors of viral videos, the selected videos would firstly have to be considered viral, meaning that they should have been viewed many times and attained the large number of views through sharing. Furthermore, I wanted to select a diverse set of videos.
Although “many times” is a vague quantity, I believe that videos with more than
5 million views can be considered to fit this criterion. The compliance with the second condition, viral status, depended on how each video achieved its popularity. While Lady Gaga’s “Bad Romance” for example has more than 500 million Youtube views, it is unlikely that the large number was mainly achieved through sharing. Lady Gaga was a world-famous entertainer before the video was uploaded. Therefore, many viewers were probably actively looking for the video. Even without any sharing at all, Lady Gaga’s videos would still have many views. Because the goal of this research is trying to evaluate the quality of video content that makes a video go viral, popular videos that rely on prior fame of a person or product related to the respective video should be excluded. Thirdly, the videos tested should have a wide range of purpose and of content and production quality. Both privately and commercially produced videos have gone viral regardless of genre and production quality, which is why it would be useful to research several different video genres to see if there are common success factors for videos in general.
I finally selected six videos to be included in the survey: “Gangnam Style” by Psy, “Charlie bit my finger – again!” by HDCYT, “The man your man could smell like” by Old Spice, “The Harlem Shake” by TheSunnyCoastSkate, “Abercrombie &
Fitch Gets a Brand Readjustment #FitchTheHomeless” by GKarber, and “Victim Fights Back In NSW Sydney School” by Fail Nation. These six videos represent success stories of the genres music video, home video, product commercial and nonprofit campaign; they are mutually distinct and all gained their popularity through viral sharing.
An online survey was designed using the website www.soscisurvey.de. Only one video was attached to the top of each questionnaire, so that there were six different
questionnaires in terms of video content, but all including the same questions (an exact listing of the questions and items of the survey is attached in the Appendix).
Questions 1 and 2 (3 items) asked about the participants’ gender, age, nationality and mother language. All following items except for the construct “Degree of User Involvement” were measured using 10 point scales.
The construct of “Simplicity” was measured in Question 3, consisting of 4 items measuring ease of understanding for the participant (items 1 and 2), expected ease of understanding for others (item 3) and ease of sharing (item 4); a 1 point answer constituted high difficulty, while a 10 point answer constituted extreme easiness.
Question 4 measured “Sensory Impact” by asking for the strength of the visual stimulus (item 1) and the auditory stimulus (item 2) with a 10 point answer representing a very strong stimulus.
Question 5 (emotional impact) asked the participants to rate the strength of the emotions they felt while watching the videos on a scale from 1 to 10 from a list of 16 different emotions. The “Novelty” factor was measured through the emotional response of “surprise”.
For the construct of Degree of User Involvement (question 6) the questionnaires contained a list of actions (“discussing the video with friends or family”, “sharing the video online”, “incorporating the content the viewer’s life”, etc.) and asked whether or not the video allowed for these actions to take place. The participants were asked to check all options applicable using checkboxes.
In question 7, the participants were then asked to rate their desire to perform each of the above actions on a scale from 1 to 10. This question measured the construct of “Provocative Power”.
Questions 8 and 9 asked for the participants’ opinions on how popular the video would be in the future (1, 5, 10 and 20 years later) and how big of an impact it would have on the ideas or products related to it. These questions measured how significant the videos were considered by the viewer and if they were expected to have a lasting impact. The factors of significance or lasting impact are for the purpose of this study combined in the construct of “Staying Power”.
The questionnaires were all sent out together as part of the survey. The participants were instructed to first watch the videos and then answer the questions.
The questionnaires were available to the participants in English and Chinese language.
A pretest of the questionnaire was conducted with three individuals of different linguistic backgrounds (English, German and Mandarin Chinese) so as to ensure ease of understanding and a logical order of the items asked. The feedback that those individuals gave led to minor changes in the wording of some questions and answers.
Finally the survey was conducted using a convenience sample of Facebook users. For this study a convenience sample should be an appropriate method because Youtube
does not specifically target any demographic group
(http://www.reelseo.com/youtube-demographics-tip/), meaning that people of any age and any nationality could potentially participate in sharing.
The survey was sent out via Facebook to personal friends and acquaintances and also posted in several public Facebook groups which altogether contained more than 3000 members of various nationalities and cultural backgrounds.