A Viral Video Case Study: Lessons Learned from an LGBT Tobacco Prevention Video That Went Viral Globally

Background: 

Viral videos are often discussed in public health because they lead to extraordinary levels of exposure and worldwide reach with little media spending. However, very few videos “go viral” and the exact formula to achieve this goal has received little scientific investigation.  Though there is an element of chance in whether a video is shared widely, there appear to be specific strategies that can increase a video's reach.

Program background:  CRUSH is a tobacco control program funded by the Southern Nevada Health District designed to reach and appeal to LGBT communities using nightclub nights, direct mail, and social media. In 2012, CRUSH launched the “Gay Military Marriage Proposal” (GMMP) video as part of its social media strategy. It depicted a service member’s proposal, where he determined that in order to live a long and happy life with his partner he needed to quit smoking. A blog-focused strategy was used to distribute the video and drive traffic to CRUSH’s educational website. The video was not promoted using any paid media, traditional or digital.

Evaluation Methods and Results: 

The GMMP video’s reach was evaluated using Youtube metrics, which included number of views, comments, external referrals, and referral sources. As of April, 2015, the video had received 698,436 views, 5,190 “Likes” and 1,288 comments and had been featured on over 100 external sites. 35% (n=246,733) of GMMP traffic came from YouTube search results, followed by 31% from unknown sources linked directly to the YouTube video (n=218,648), YouTube’s suggested video feature (13%, n=92,606), unknown embedded players (12%, n=82,784), and a variety of other sources. The highest number of external referrers came from Facebook (48%, n=7,286), Google (15%, n=2,264), and reddit.com (14%, n=2,125). Analysis focused on mapping the digital spread of the video to understand how it “went viral,” which can help create future public-health-oriented viral videos. Inspecting the source of traffic at the day, week, and month-level illuminated several patterns. First, social media sites such as Facebook, YouTube and Twitter contributed significantly to the overall views, but were not a significant source of initial views. Instead, 23 blogs contributed significantly to the initial spread of the video. Tumblr was also a significant source of traffic initially. Together, they accounted for the majority of traffic in the first two weeks. Afterward, social media sites dominated the traffic sources, sustaining significant traffic to the video to present day.

Conclusions:  This viral video was first distributed across blogs and Tumblr pages before it spread across social media. These blogs were the seeds for its viral distribution. Although they only represent a small fraction of the total video views, it is unlikely that the video would have “gone viral” without the earned media from these blogs upfront.   

Implications for research and/or practice:  Vral messages are possible in public health campaigns, given proper planning and their incorporation into a larger strategy. To increase the chances a video will “go viral,” campaign managers can develop a list of blogs that reach their target audience and perform outreach to blogs owners and authors.