/ Analysis

Can audio be viral?

Can audio be viral?

I wrote a few days ago that This American Life's latest spin-off, Serial, seemed to correspond to the definition of viral — perhaps the first occurrence of an audio stream reaching such proportions.


But what makes something go viral? And why? Why do we describe the exponential social amplification of media artefacts in these terms?


The first thorough explanation is to look at the growth curves. The viral growth curves of online content can be modelled using epidemiology models drawn from biology. In most cases, the duplication of a media's audience follows a logistic growth curve. Log growth curves are characterized by slow growth at the beginning, which accelerates, then slows down, until the curve reaches a plateau of more or less linear growth. 


For online content, the nature of this "autocatalysis" depends on several elements of activity. In the case of audio, much of this actually happens offline. 


"Have you listened to this new thing, Serial, Mary? Oh, you really should! It's captivating." 


Yet in the case at hand, the absence of comparable data makes it hard to determine what type of "epidemic" we are faced with. Below are cumulative and daily epidemic curves for Ebola in Guinea, Liberia and Sierra Leone. Three countries where "virality" has had a very differently meaning.



In determining the potential reach of viral phenomena, one needs to determine how the growth curve will behave, and what kind of ressources can and should be committed to stimulate (or in the case of viruses, contain) the epidemic. 


Audio: a virus like no other


While even the very identification of what constitutes a viral phenomenon is in itself problematic, the radio / podcast sub-category is in a world of its own. This is due, in large part to the fact that the distribution system of audio content is fragmented and that very little aggregate information is available.


On Dec. 12th, Pete Davies wrote on Medium


"In web analytics, we call this duplication. It’s a big problem for counting users […] on web and mobile: the same person can read something in a mobile web view and later on their computer; it’s very hard to relate those two sessions as belonging to the same person."


Despite these measurement difficulties, the load of media coverage and omnipresence in social media hints us at a strong viral growth curve for recent audio projects.


Without further analytics though, we cannot tell whether we sit at the bottom of the curve, at its apex delta, or have reached the more or less slow-paced, linear growth of the long tail. 


Alternative indicators


On Dec. 5th, the IBTimes reported that Serial had been mentioned a mere 285,000 times on Twitter. 



Compare this with the 600,000 tweets per minute produced during the final of the most recent Brazil World Cup (a total of 35,6M tweets were produced during the match) and we find that Serial is nothing short of an epiphenomenon.


Yet to believe that would be to miss the point.


For one, podcasting remains a relatively niche market, an extension of traditional analog radio formats for which we are witnessing, to quote NextMarket's Michael Wolf, an explosion in creative use. 


Of would-be podcast listeners, it seems, Serial has captured a significant proportion. Again, following Davies, it is difficult to estimate what market share that represents exactly. But its an undeniable fact that when a product as marginal as a podcast becomes material for headlines around the world, there is something to be interested in, and a necessity to ask questions as to whether this is the beginning, the middle or the end of something big. 


A simple search at Google search trends reveals that while podcasts were in high fashion when they were initially launched in 2005, the interest receded to reach a stable, yet much lower level from 2008 onwards. 



What is happening now, though, is an explosion in search occurrences, led notably by the sudden popularity of Serial and Alex Blumberg's Startup. There has never been, in the history of Google, more interest in podcasts than now. 


And that's something. 


We need to experiment some more


On Jan. 14th 2014, Stan Alcorn published an article entitled "Is this thing on: why audio never goes viral". Chief among the reasons of podcasts' apparent anti-virality is the fact that audio content is, for many, a complementary good. "The greatest reason is structural", says Bullseye's Jesse Thorn. “Audio usage takes place while you’re doing something else.”


Second is the fact that sound, unlike text or video, doesn't lend itself to capture as easily. It cannot be displayed as easily, or quoted to be indexed. The value is in the product itself, in taking the time to listen and decide… 


There are, for instance, very few trailers for podcasts. 


Moreover, very little current work done in radio and podcasting seems to have a viral intent. It seems like oftentimes, the topic, the editing, the distribution channels, are conceived like "radio for radio's sake". 


The problem, then, is one of brand, and, furthermore, of marketing. A fact that NPR' Eric Athas recognizes as an opportunity to experiment and delve into the world of the "viral industry" using some techniques that have made the wealth of new media empires like Buzzfeed, Upworthy and Gawker. 


While it has produced mitigated results thus far — some of the packaged sound bytes shared to SoundCloud have been played up to 8,000 times — Athas' team at NPR follows a very simple logic of identifying, repackaging and aggressively sharing audio, using and developing novel tools to do so.


Michael Wolf's Forbes piece predicted that 2014 would see an "explosion in creative use"; a prediction that proved to be not only accurate, but a gateway to a reinterpretation of podcasting's vitality and virality. 


Much potential remains unexploited. 


comments powered by Disqus