Influence and Virality: A Primer

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

Share this post:

Possibly related:

• Under the Influence
• Viral Spiral
• Blog influ­ence