Back to our blog

16.02.16 / What is a viral? The changing nature of social media distribution

There’s a pretty standard joke between social media marketers, who have all, at some point or another, been requested to produce a ‘viral’. Cue eye-rolling. But over the last eighteen months or so, that’s changed. It is possible to ‘order a viral’.

Why? Well, a lot of it comes down to the changing definition of ‘viral’, combined with changes to the social media landscape.

So, what does ‘viral’ now mean?

Well, a viral used to refer to the ‘natural’ spread of a piece of content within connected networks. The very term has its roots in its academic origin – the spread of infectious disease within social networks (populations). The adaptation and embracing of epidemiology is most evident and accessible in Barabasi’s seminal ‘Linked’.

But now viral is coming to refer to social-first content distribution, reliant on high volume captive audiences, and (then) their shares.

Some of the characteristics of the ‘old’ and ‘new’ viral are the same – for example the prominent role of an emotional and/or behavioural hook to the content, which compels people to share with peers.

But differences are also evident – in the old ‘viral’, the origins were within niche, tightly knit communities, or even from totally disconnected individuals across a wider network. The spread was often slow, until it hit critical mass (contagion, to use disease terminology). In the new world there’s nothing accidental, or serendipitous about its discovery – it is deliberately manufactured or distributed by organisations (Buzzfeed, Upworthy, College Humour, amongst others) which are premised on serving emotionally attuned content to very sizeable social audiences. It’s 1980’s MTV alive in the Facebook age in its cultural relevance and size of audience with peer-to-peer digital distribution.

As our understanding of online networks continues to grow, we gain in understanding of how to capture and exploit the unique properties of online networks. Especially key in this has been exploitation of ‘betweenness’ of nodes in networks – or, to put it less technically – the shared affinity of you and others in your specialised networks, based on factors such as location, upbringing and jobs. (Think “ten things you only know if you’re a barista”, “You’re a child of the eighties if…” and others).

Yet many marketers miss the difference between paid social distribution (which is not viral in the ‘old’ or ‘new’ sense of the term), and earned (albeit sometimes ‘boosted’ with paid seeding) ‘new’ viral distribution from Elite, LadBible and others. Whilst the dynamics of the origins of the spread of the content are new, the patterns and reasons for its subsequent spread are the same as they always were.

Or, put another way, ‘old’ viral was about unexpected, unpredictable events. They provided exceptions to the norm. ‘New’ viral could be categorised as ‘expected’ – it’s known and predictable when the content gets thousands of shares, fast.

Where does that leave original (‘old’) viral content?

Where do the cat videos and Charlie bit videos of the new era reside? Is it possible to have a breakout hit anymore, without the clout of a major social publisher? Well, ‘pre-viral’ might, ironically, prove the best label (lovingly borrowed from NewsWhip), for what we once termed ‘viral’.

That’s because the threshold for viral has changed from thousands to millions. Things which we once described as ‘viral’ are now merely ‘trending’ – too ephemeral and small in their reach – only blessed with the approbation of ‘viral’ once they hit the millions, propelled by content factories like Huffington Post or the Daily Mail to a greater stratosphere of reach and social shares.

Or perhaps the term ‘viral’ should be abandoned. An aphorism left to describe content which is advertised heavily within social networks, and/or shared heavily by the content factories.

Totally natural peer-to-peer content,with discovery originating in niche communities and isolated individuals might be rebadged. The evolution of language might be to name it ‘chained’ content, describing the daisy chain like effect of organic content distribution. It would need to meet certain technical characteristics: as a starter, a high number of connector nodes, and a low ratio of distributing nodes vs edges on the first 5% of shares, might ensure that anything originating by a major social publisher wasn’t mis-labelled.

Does the drift in the term ‘viral’ matter?

Not enormously – as long as everyone is clear in what they are requesting, and receiving, when they ‘order a viral’. The halo of the great virals of the noughties ought not be attached to the evolution of paid marketing tactics now represented by large scale ‘viral’ marketing campaigns, but having done too poor a job of explaining the difference, a new form of language is required to describe the ‘pre viral’ or ‘chained’ content strategies of modern social marketing.

Ultimately the ‘viral’ appeal of the very term ‘viral’ sealed its own confuscation.
Thanks and credit to Andrea Lopez, who had a recent Twitter storm on the evolving meaning of the term which sparked our thinking on this topic, and has also provided valued input on this. She also, rightly, points out that one can still have an ‘old’ style viral, when a teenager gets 400 RTs on a photo of them + celeb, seemingly from nowhere. Maybe we’re becoming nostalgic too soon…

Blog / Archive