21 October 2013 - 6:48am | posted by | 4 comments

The Science of Sharing - what makes a video go viral?

Psi - Gangnam style Psi - Gangnam style

By some considerable margin, ‘Gangnam Style’ remains the most viewed viral video of all time. So far.

As The Drum reported last year, its global success was not the work of random chance, but the careful manufacture of content married to rigid research, careful seeding and artfully plotted strategy. Videos of sneezing panda cubs or Charlie biting his brother’s finger can achieve enviable share on purely organic terms, but the Holy Grail of marketing at the moment is to be found in achieving the perfect sharing formula by design and the team at South Australia’s Ehrenberg-Bass Institute believe that they have found it.

The institute’s researchers work strictly on applying the science of marketing, and whilst many in marketing profess to possess the alchemic, Don Draper-ish, mystical Jedi art of marketing nous, the Ehrenberg-Bass team have worked for over a decade demonstrating that there is a formula to marketing; a rigid set of inputs will yield a defined set of outcomes. It can be distilled, broken down, understood and applied. And it is the subject of its latest publication, Viral Marketing - The Science of Sharing.

“According to our research over the last two years, getting big is largely about getting seen. To be seen by many, the distribution must be optimised -earned media alone will not result in huge reach,” lead researcher Karen Nelson-Field says.

“This may challenge the diehard social media marketers who believe earned reach is king. Additionally, to be truly viral (where the number of shares deviate above what would be expected given the reach), a social video needs to be emotionally arousing, at least to some extent.

As Bruce Lee said so memorably; "We need emotional content." But what is that, exactly, scientifically?

“Content creators should aim to increase the emotional appeal of their videos, with less emphasis and fewer restrictions on the creative devices they use. Creators should worry less about whether the video content contains a baby, a dog or a celebrity, and instead invest in pre-testing to ensure the material makes the viewer laugh, gasp or get goose pimples,” the study states.

A 2012 study of 7,000 articles by Berger and Milkman found that “arousal” - events that cause laughing or tears, take your breath away, make you sick in the stomach, gasp or give you goose pimples - play a key role in viral diffusion. However, this is only half of the story.

“What’s more, for a video to affect sales it needs to be remembered by many people—and to be remembered the brand needs to be prominent,” Nelson-Field says.

She adds: “It would be nice to believe that viral success is as easy as being sneezed on. Even sensible folk want to believe that following the ‘Top 10 Tips of Successful Bloggers R Us’ will lead to superstardom. The reality is, of course, that social video success is more complex.

“Deep practical insight can only come from high-quality, reliable research, and this takes time and commitment. It is a commitment to rigour and the use of multiple data sets to ensure the results are real, generalisable and applicable that set this book apart—these, and the fact that our primary aim is to research, not to sell.”

Whilst acknowledging that some viral videos and viral marketing campaigns have ticked the boxes accidentally, the researchers have narrowed down the winning formula into its specific components, identifying the importance of emotional engagement, explaining how to reach beyond the seeding platforms to connect with a broader audience across the full customer base, and how to target the vital quality reach that is vital.

One of the key concepts of the book - emotional arousal- confirms that videos that drew a “high-arousal positive” emotion from their audiences were shared more often than those that drew any other emotional response.

“If we looked exclusively at arousal, high-arousal videos were shared around twice as often, compared to videos with low-arousal content,” the book states. It also quantified the difference between those virals which are loved, rather than disliked.

“When we looked at both arousal and valence, content that elicited a high-arousal positive emotional response was shared around 70 per cent more on average than the next highest sharing arousal–valence group, high-arousal negative,” it said.

“Particular emotional responses do not guarantee that a video will go viral. Yet when a video really does go viral, it is the high-arousal videos that earn massive sharing.

“Knowing what emotional response to go for improves your potential return, but not—it would seem—your risk. This suggests that the right sort of emotional response should be best considered a necessary, but not sufficient, condition for a video to be a viral superstar.

“Focus less on creative appeal and more on emotional appeal,” the authors counsel.

“We know that emotions are important. In their high-arousal form they are an important part of high-sharing video content. But in our supercharged world, funny isn’t enough—only hilarious will do. Happiness is okay, but exhilarating is what we want. Negative high-arousal emotions can also prompt us to share, but which marketer is brave enough to risk offending their customers? Positive emotions are a much safer bet.”

The authors point out that the majority of social video is “still safely ho-hum, observing that many highly emotional videos don’t go viral. The key is identifying the specific content that leads to mass sharing.

Good emotional work can be instantly undone by careless, sloppy or too-prominent branding, and the research finds that there is a science to balancing the branding message of the viral video too. It warns that care and discretion is required when interpreting eye tracking or biometric research used for testing when the brand identifier appears.

“A drop in biometric readings can easily be misinterpreted; it may not be a negative reflection on the brand, but rather a reflection of the difference in our processing of familiar versus novel stimuli,” the book states.

When faced with familiar stimuli, such as the brand, the brain can relax, as a familiar stimulus is easy. The authors define prominence as a combination of branding execution tactics such as visual frequency, size relative to the screen and duration. to process. In contrast, when faced with novel stimuli, such as the rest of the advertisement, the brain needs to remain active. So to prompt a mind to think of your brand more often, each exposure needs to work hard for the brand that is advertising.

“It needs to be prominent, clear and directly linked to the brand. Taking a subtle approach to branding makes it harder for your customers to know who is advertising. We question the value of any marketing communication that is intentionally poorly branded. If clear branding turns out to diminish the likelihood of a video being shared, then the very value of social video as a marketing tool should be questioned.”

As observed, sneezing pandas or any random incident can become accidentally viral without seeding or planning, but the researchers at Ehrenberg-Bass point out, a virtuoso playing a violin in her bathroom may be fabulous but cannot be widely appreciated, and finds that there is strong evidence that there is still great benefit in seeding and supporting that video.

“The reality is that most videos aren’t particularly contagious, so this recommendation to buy views, to ensure wide availability, carries even more weight. The creative characteristics are still important, but a waste if the video isn’t seen,” it states.

“Getting your video viewed by a wide audience will at least make it easier to also deliver good sharing metrics. If we take a definition of viral success more narrowly to mean purely how contagious it is (how likely it is to be shared by a viewer) then the winners have some defining characteristics—notably their ability to elicit high-arousal emotional responses from their audience, typically through displaying personal triumph.

“But plenty of contagious, positive high-arousal videos still hardly get any shares, and hardly any views.

“You can have a very contagious video, but if you start with a small base, the total views will typically be small. Of the 400 branded videos investigated here, fewer than 10 ticked both boxes: high viewing and higher than expected sharing. So it seems superstars are very rare indeed. This adds more weight to the notion that investing in a larger base limits risks associated with viral marketing.”

Professor Byron Sharp, co-author of The Science of Sharing, previously published ‘How Brands Grow’, and the Institute’s methodical, empirical approach to marketing is beginning to permeate globally from its South Australia base. The team’s body of work is providing a significant archive of new research that is helping marketers who previously believed there was a nebulous art to converting that lost 50 per cent of marketing spend, into scientists who understand that the ingredients of marketing success can be identified, distilled, replicated and transmitted to customers to be sold.

“And now we know where to find them. They’re not your Facebook friend or your Twitter follower. They’re everywhere else—presumably watching television or just randomly searching the web looking for an opportunity to cheat on their favourite brand,” the book concludes.

“The serious side of this finding is that marketers who need to be careful with their budgets will have to reframe how they invest their marketing dollars. They should be wary of over-investing in the relatively small numbers of already heavy buyers who typically congregate around online brand communities. Rather, they should be investing in strategies to reach the valuable light buyers.”

It adds: “People are responding to social media in the same way they have responded to radio, television and the internet. As with all new things, the hype around the potential of social media is disproportionate to reality.”

The reality, it turns out, can be easily explained.

Comments

21 Oct 2013 - 11:13
colinbancmedia's picture

Not sure about those conclusions at all. Agree that it helps to pay to distribute content, but there are 10,000s of examples of non-sponsored content going viral. Pretty sure the Triple Rainbow Guy didn't pay for Facebook promotion! Also I would say "emotional" comes a far far second place to "interesting & unexpected".

1
1
21 Oct 2013 - 12:30
mihid4711's picture

"interesting & unexpected" is actually emotional - emotions you experience when you see something interesting and unexpected make you share that video or even watch it again...

2
1
22 Oct 2013 - 13:45
colinbancmedia's picture

@mihid4711 I'm sure you can lump pretty every video on the internet as somehow relating to an emotion. But that doesn't help reveal what goes viral.

0
0
22 Oct 2013 - 12:18
thehe19578's picture

Good effort by the ever excellent Ehrenberg Bass team & Jonah B. We really do need more empirical evidence for the assertions thrown around. However, I'm not sure (without reading the paper) the conclusions are a slam-dunk.

Whether you really hard to spread something or not, if you get lucky, you get lucky. In social diffusion, it's the social piece - the people things - that matter much more than the qualities of the thing which gets spread by the people.

Also, I wonder is this all post-hoc analysis or has it been validated as a predictive tool? If not, it's not much use, is it?

Props though for going to the effort to do this kind of analysis

0
0

Please sign in or register to comment on this article.

Latest Projects from the Profile Hub

University of Aberdeen - Graduate Business School brochure

25/11/2014
Background:  999 Design has enjoyed a long working...

We delivered a 265% increase in ROI year-on-year for Pandora UK.

25/11/2014
Pandora designs, manufactures and markets, hand-finished,...

Maximising an Industry Changing Story

24/11/2014
In November 2014, WOW air announced £99 fares (yes, that’s...

In The Snow Magazine Front Cover

20/11/2014
Front cover from a fashion shoot in Zermatt earlier this...