How do things become ‘viral’ on the Internet? And what exactly do we mean by ‘influence’? Marketing and PR people want their messages to spread in the most effective and efficient way possible, and so these questions have received a great deal of attention, particularly in recent years, as we’ve seen the rise of ‘viral memes’ on the Web. This article aims to cover the basics and also provide some new ideas for discussion.
Dramatic Hamster: 19mn views for the version that went viral; maybe 40mn from the spin-off versions. But why?
There are four parts to the solution (the one to the question about how things become viral and influence spreads; I have no idea about the hamster):
- the nature of the spreader or influencer(s);
- the nature of the audience;
- the nature of the network or media through which the information is moving;
- and the nature of the information/content itself.
So how do you get stuff spread on the Internet?
1) Through very influential people
Some people are more influential than others, it seems. If I tell you to buy a Blackberry rather than an iPhone, I’m not – on the face of things - likely to achieve as much of a result as if Brian Lam (the editor of Gizmodo) said it. Why? Because he’s more of an influencer on this subject than I am. He has a rather larger audience. He runs a site that is acknowledged as an authority for this sort of information. And because he’s being reviewing mobiles and other gadgets for some time. If you were a marketer working for RIM, you’d probably be inclined to take Brian out for a nice lunch and send him some Blackberries.
You might call this the ‘old-fashioned’ model of PR and marketing. You want a story about your product/services/company in a big publication – so you take the editors out to lunch, give them access to information not available elsewhere, such as an interview with the CEO and send them free stuff.
This old model has been somewhat reinvigorated in recent years with the realisation that the most influential people might not be working for the FT or even Gizmodo, but might work for what seem smaller publications or websites that are read by or otherwise reach all the other editors. These ideas stem from those published by Katz and Lagerfeld in Personal Influence (1955). Malcolm Gladwell’s influential 2002 book The Tipping Point revived this, suggesting the existence of thought-leaders (Mavens and Salesmen in his parlance) who – if they recommended something – will lead to that item spreading like wildfire – the spread of hush-puppy shoes in 90’s New York is his most-cited example. It didn’t come from fashion columns or fashion magazines. It came, Gladwell says, from a small number of eminent dudes (to use the scientific term) sporting this footwear brand.
That’s when things get tricky. Finding these ‘quiet influencers’ is difficult since you (as a marketer) haven’t got access to other people’s reading lists. Where do the editors of the FT and Gizmodo go to for their information? Where do they find their stories, outside of their Inbox? Where do other people go for their footwear and mobile phone advice? On the Web, you might look at the blogrolls most people publish on their websites – if you find the people that lots of others read or cite, then they are most probably very influential, whatever their apparent profile or position. (Tim Hoang wrote some interesting stuff about this at the start of the year).
Even greater access comes through Twitter, where following/follower lists are visible to any user of the service. I went to a seminar last week about online influence where Andrew Walker from digital agency Thin Martian described some work they did around the UK release of the film Anvil last year. They wanted to find out who the most influential rock-music twitterers were. But rather than looking to the likes of Coldplay (2.5mn followers) or Lily Allen (2mn), they traced back from the people who talk and write about rock, publishing blogs and zines. It emerges that the most influential rock-twitterer is shock-jock Howard Stern, with a mere 67,000 followers. Why? Because the people who talk, write and otherwise publish about rock all follow him. (Mat Morrison has done some analysis of the UK social media twittersphere that you may find interesting).
2) By finding influence-able audiences
That last finding, about the extent of Howard Stern’s influence, despite relatively modest audience figures points towards the second part of the ‘finding influence’ equation – finding audiences that are willing to be influenced. Stern’s read by people who are looking for stuff about rock music to publish on their own sites and publications. They’ll be receptive to all sorts of tidbits that wouldn’t necessarily be seen as interesting by other audiences. Back to the Blackberry example, if gadget-review magazine Stuff (ABC circ. 95,000 pcm) publishes a 10/10 review of the latest model, then that could well be more influential than Lily Allen giving it the thumbs-up, despite 20-times the reach. Give people the recommendation in the context they want it and it will carry. Otherwise, it probably won’t.
Research into influence and virality suggest that finding these audiences is far more important when it comes to spreading information than the apparent influence quotient of the person passing it on. Yahoo chief researcher Duncan Watts has poured scorn on The Tipping Point’s idea of key influencers:
“It just doesn’t work,” Watts says, when I meet him at his gray cubicle at Yahoo Research in midtown Manhattan, which is unadorned except for a whiteboard crammed with equations. “A rare bunch of cool people just don’t have that power. And when you test the way marketers say the world works, it falls apart. There’s no there there.”
Academic research into modelling the spread of ideas through the simulation of epidemics has supported Watts’ beliefs. Instead, Watts’ approach is far more akin to traditional – albeit well-targeted – advertising. Get the thing out there to lots of people who may be receptive. Watts recreated the famous Milgram experiment (not the one about following orders, the one about six degrees) and found that, yes, anyone is separated from anyone else by six-degrees-of-separation or fewer (just three for the population of the US). But furthermore, there weren’t key links in the chain – the alleged especially-connected people or gatekeepers. Accidents and apparently unrelated environmental variables were just as important to the spread. See Watts’ 2007 paper Viral Marketing in the Real World for more on this. These ideas should give marketers heart, though: even if you can’t get the editor of the FT to look at your company/thing, getting a few bloggers on the case may well result in the same amount of real-world influence.
And that brings us on to the next part of the equation…
3) By using media and networks that spread quickly
This helps to explain the importance of websites like Twitter, YouTube and digg when it comes to spreading the news. People go there for distraction. These sites are all about finding cool, new stuff you can either re-publish or pass on. An article that reaches the front page of digg can expect to receive 100X its normal level of readership, not because it represents world-class journalism, nor because it comes from a terribly influential site, nor even because the person proposing it is especially influential (though there are cartels of diggers and influential diggers – something the site has fought to curtail since its establishment). Rather, because it caught people’s attention as something worth passing on – I’ll go into the reasons later.
Marketers certainly can and do attempt to get hold of some of this. Ask bloggers to write about your stuff. Make short-form videos instead of microsites. Write and give away handy guides and white papers. Submit your stuff to all the social bookmarking sites. Get people to join their Facebook groups.
This can work spectacularly well. But, as most agencies and clients have found to their cost, it’s rather hit and miss. Tens of thousands of videos, groups and bookmarks are created every day. People can only take notice of a limited amount of cool, new stuff each day. Your stuff is likely to be missed. This is particularly invidious when working under the constraints of a time-limited campaign or project – items on the web can and do accrete considerable value over months and years, but that’s not much use if you’re being judged on figures to be delivered in six weeks’ time.
One important lesson here. You can’t do any of this stuff in isolation. Making a great video is all very well, but you also need to give it velocity through straight PR and marketing; social media strategy and positioning.
It’s so annoying that most efforts fail that you might be sorely tempted to cheat – get your entire staff, your client’s staff, their friends and relations, plus a load of made-up accounts to vote for your stuff in the hope of it either (a) being noticed by a larger audience because of your apparent popularity or (b) that this pseudo-popularity is enough to assuage your client. There are downsides to this, though, because people will notice. Then you will look inept, your client’s brand will suffer and it’s bye-bye social media budget.
So, as well as all the above, you’ll want to make sure that your stuff is both good and spreadable.
4) Creating stuff that’s good and spreadable
The ‘spreadable’ part is the easiest to deal with and has already been mentioned above. Make sure that people can email it in some way. Attach social bookmarking links to your items. Put it into contexts, like YouTube and delicious that encourage sharing and spreading. Instead of regular press releases, use the SMNR model to make them blogger (and journalist) friendly. Make artefacts remixable to help generate spin-off variants.
That’s all quite common sense and mechanical. Good is where the difficult part lies.
There’s already plenty of advice on creating linkbait – make something that’s either useful, surprising or controversial. (‘Linkbait’ is web content that people are likely to link to, share or otherwise transmit elsewhere). Unfortunately, everyone on the Web has already read those articles and we’re swimming with ‘Top Ten Ways to/Resources for X’, ‘Why Y will be the Z Killer’, zippy flash games and mind-blowing visualisations. There’s an SEO arms race afoot and while there is a lot of success to be had with these formats, there are thousands of people going after the same top slots.
To be more positive. A recent academic paper – Social Transmission and Viral Culture by Jonah Berger and Katherine Milkman – analysed the virality of 7500 items from the New York Times, monitoring the most-emailed list from the site. They focused on the psychological characteristics of the items, rather than their actual content, which provides some useful prompts. Unsurprisingly, perhaps, awesome is good: “content’s ability to inspire awe is strongly associated with its virality”. But awesome doesn’t mean – here – some sort of Bill and Ted mind-blowing. In a more classical sense, it means something bigger and wider:
One emotion we focus on in particular is awe. Stimuli that open the mind to vast and often unconsidered possibilities can inspire awe, a unique human emotion that expands a reader’s frame of reference (Keltner and Haidt 2003). Awe is the emotion of self-transcendence, a feeling of admiration and elevation in the face of something greater than the self (Haidt 2006). It occurs when two conditions are met (Keltner and Haidt 2003). First, people experience something vast: either physically vast such as the grand canyon, conceptually vast such as a grand theory or finding, or socially vast such as fame or power. Second, the vast experience cannot be accommodated by existing mental structures. Intellectual epiphanies, natural wonders, and great works of art can all make people feel a sense of awe (Shiota, Keltner, and Mossman 2007). Similarly, news stories about a treatment that may cure AIDS or a hockey goalie who continues to play even with brain cancer may both inspire some level of awe.
The other values tested, all of which indicated likely virality, though to a lesser extent, were:
- practical usefulness (e.g. how to get a cheaper mortgage)
- surprisingness (e.g. dog drives owner to hospital)
- positive stories vs. negative stories (e.g. X is great vs. Y is terrible: positive stories get greater word-of-mouth, which you may find surprising)
- emotional content (e.g. anger against X; tragedy of Y)
Awe-inspiring ought to be the goal, then. Hmm you’re thinking that’s all very well. But my client makes sprockets for Acme. Where does that leave me? In all honesty, it probably means that you aren’t going to achieve virality for a story without considerable imagination. A great example would be liquidiser manufacturer Blendtec who converted drab kitchenware into a runaway viral success with its Will it Blend videocast. The Best Job in the World campaign by Nitro converted a run-of-the-mill ‘Come to Queensland’ message into an awe-inspiring adventure to attain an idyllic lifestyle. Diet Coke and Mentos turned two totally familiar supermarket staples into boy’s-own science fun combined with Bellagio spectacular. Yes, that sounds bloody hard to emulate, but that’s the nature of this territory. Sorry.
To sum up
Four ingredients to spreadiness. None of them are totally foolproof, but each reinforces all of the others, so it only makes sense to work on all four. In essence, it’s not very different to a traditional media plan.
- Involve influencers (if you still believe in that tack); but don’t mistake volume for influence.
- Work out where your audiences are (who and where are the people who want to talk about this stuff?) and go after them with the best & most sources you can reach.
- Do blogger outreach (but don’t be a dick); social media sites; making it spreadable; etc. Get broad exposure as well as targeted. It can produce the same results, if not better.
- Take a long time to think up the ideas. Be really imaginative and make it awesome.
picture credits: jiparis and wikipedia
Many thanks to Pearson Education for sending me two recent books about blogging for review. The first of these is Blogging to Drive Business by Eric Butow and Rebecca Bollwitt. It seems that Eric has written the more business and strategy-centric chapters, and Rebecca the more practical information about blogging.
Social tools, devices and web evolution are creating epochal change in media, society and business. The plan is to 
Age of Social Network Users
New data from Pingdom on the age of social network users confirms the rumours. They are mostly quite old, or they lie a lot about their age.
The smallest group of people using social networks is the 18-24 age group, which rather confounds the idea that these sites are for young people. Across the board, only 9% of 18-24 year-olds are social network users, according to the research. Even on bebo, the ‘youthiest’, of the networks in the group, the mean age is 28. On facebook, which was originally conceived for college students, the average age is 38.
I have to throw in three cautions here:
Notwithstanding, I don’t think this is enough to explain away the figures. Let’s face it: most social networks are more popular with older people than the young.
More surprises about what young people do and don’t do comes via the Guardian this morning. The following graph shows the percentage of people who have paid, or are willing to pay, for media products, by age:
Across every media type, teenagers claim that they are more willing to pay or have paid than people in their forties.
If you combine the two pieces of research together, you discover that most social networks appeal to a demographic that is significantly less likely to pay for anything than teenagers. More headaches for Mr Zuckerberg.