Chair in Social Media Monetization – Social Media, between audience and virality: the scientific point of view

The term of virality is often applied to each success on the web without anyone realizing exactly what it means. The Chair in Social Media Monetization tried to decipher this notion at a conference on December 13, which was dedicated to the model of audience in social media.

Jean-Samuel Beuscart and Kevin Mellet, researchers in socio-economics at Orange Labs, presented us four ways to define the term “virality”.

1. Virality as a success curve

Success curves on the web are characterized by an exponential rise to a peak from which the sharing of content declines gradually to oblivion. Social networking can be viewed as a series of collective relatively focusesthat succeed each other. Modeled in the form of curves, these sequences combine two factors:

• A desire to watch what others have already watched: through mimicry, the user is interested in what incites general interest.
• An attraction to novelty, which also leads to the obsolescence of the content.

By combining these components, of mimicry and novelty, the researcher Huberman has developed a method to predict the success of content from its initial audience.

However, there is a weakness in this model: from the curves’ aggregation of several effects (contagion, editorial …); it is difficult to ascertain the ideal combination that guarantees the virality of content. The video of Susan Boyle, for example, has benefited from the combined effects of person-to-person circulation and from the repeated reproduction on traditional media.

2. Pure contagion analysis

A second approach aims to identify the results of word-of-mouth, isolating the other factors. By studying the success of a set of photos on Flickr, a group of researchers has been able to conclude several trajectories:

• Some photos can gain virality after editorial exposure: for example, after being published at the home page of a media site.
• Others experience a gradual and growing success which takes much longer to emerge, and which is based on personal recommendation.
• The « hybrid » way remains the most common one and it is based on a combination of word-of-mouth and editorial exposure.

The authors of this study have demonstrated the equivalence of these two factors in the success of image viralization: if the word-of-mouth can claim to be 53% responsible for their circulation, their editorial exposure is by 47%.

3. Are individuals « virulent »?

The third approach does not concern the structure of successful content, but the people using it and their potential virality.

In the 50s, Katz and Lazasfeld outlined the supposed influence of media on people: it would happen in two stages. The « opinion leaders », sensitive to media messages, would redistribute the information in their networks. As a result, recommendations of their environment would impact some people’s voting behavior or consumption habits, much more than the media.

That concept in the world of marketing would be translated into the « S curve »: the information, when discovered by a few influencers- sensitive on innovation, is transmitted to the whole population. From this arises the importance of targeting opinion leaders, as explained Malcolm Gladwell in The Tipping Point.

This pattern, little doubted for half a century, was reviewed by J. Duncan Watts and Albert-László Barabásiqui, who tried to complicate the issue: if a network is not very concentrated, influencers cannot be highly influential. However, if a network is highly concentrated, while the influencers are very difficult to influence … the virality is therefore based on very specific conditions. It is necessary that the network is hyper-concentrated and that the influencers are easily influenced themselves … which makes those individuals appear paradoxical.

To understand the dynamics of distribution of information on the internet, the researcher Watts proposes the metaphor of fire: the wind and the dryness of the ground are just as critical as the number of outbreaks in the release of a forest fire.

Similarly, for virality, it is interesting to identify the source and the crossing points of content, but one should also take into account the intervention of external parameters: for example, the work done by conventional media or advertising, which can raise awareness to the population in order to accept innovation and share it.

4. The nature of content

Three theoreticians propose focusing on specific characteristics of viral content.

• Jonah Berger, an English researcher, explains that people primarily share « frivolous or emotional » content, not in order to achieve a certain goal, but as a way to forge links from a distance. Since it is difficult to share strong feelings from a distance, the internet facilitates the circulation of content with a strong emotional dimension. Hubert Guillaud published a report of this study on InternetActu .
• Henry Jenkins in Convergence Culture opposes to the metaphor of virality that treats internet users as passive ones. A major figure in cultural studies, the researcher explains that content is viral above all, because it is subject to reappropriation by individual users … for instance Lolcat, that represents the quintessence of viral culture, is suitable for permanent reinterpretation, or for « cultural poaching » as Michel de Certeau said.
• Bill Wasik, the inventor of the flash mob, suggests that content is not very important. The viral potential is more of a « self-referential game « : the content could be moved to the background. What is important is the dynamics of buzz, the fact of being, or at least guessing what will be the next buzz.

In conclusion, these different studies make possible the identification of viral forms through sequences of focusing public attention, but it is very difficult to identify the ways of pure contagion.

The only certainties:

• There are no individuals who are influential under all circumstances.
• The buzz seems to happen on specific types of content, but is not easily predictable.
• The buzz is always a combination of several influences including those from the media. The legendary « word of mouth » usually makes us forget that.

For further study, Jean-Samuel and Kevin Mellet have written « Le succès sur Internet repose-t-il sur la contagion ? Une analyse des recherches sur la viralité (Is success on the Internet due to contagion? An analysis of research on virality)« , an ar
ticle in the journal Tracés, n° special issue « contagions » (co-written with Thomas Beauvisage and Thomas Couronné).

Video interviews of Jean-Samuel and Kevin Mellet (in french with english subtitles) here.