Skip to content

Exploding The Influentials Myth

2011 November 16
mgladwell

In the ancient world, primitive people knelt at altars and prayed to imaginary Gods for good fortune, a bountiful harvest, a pox on an enemy’s house, victory in battle or whatever.  This was a matter of faith, not fact.  They simply believed.

In our modern world, marketers pray to their own imaginary Gods.  They call them “Influentials,” anonymous people with godlike influence, who are about as real as the ones on Olympus or Valhalla.

Nevertheless, the idea has caught on.  Marketers have become positively smitten with it. After all, why pay to reach millions if you can get even greater effect from just a few special ones?  As we shall see, it’s a misguided, quixotic notion with little or no evidence to support it.  In fact, serious inquiry into the matter finds the notion completely baseless.

How It All Began

The idea of Influentials is not a new one.  It has its roots in work done on the two-step theory of communication popularized by Katz and Lazarsfeld back in the 40’s and 50’s and summarized in this 1957 paper.

The idea was simple enough.  They found that people were influenced by other people much more than they were influenced by mass media.  They called these people opinion leaders and supplied three key attributes to identify them:

Competence: They know a particular field well.

Trustworthiness: They personified certain values.

Access: They were in a position to influence others.

The idea got a further boost from Malcom Gladwell’s enormously popular book, The Tipping Point, which glorified “The Power of the Few” and posited that there are “mavens” among us with amazing abilities to spread information.  Smart marketers, he suggested, could identify and exploit these wonderful people to their own advantage.

Sounds great, right?  Unfortunately, it’s mostly bullshit.

Why the Tipping Point is Toast

The notion of Influentials is an attractive one.  It’s simple, easy to understand and holds great promise.  The problem is that for it to work, such people would have to be identifiable by specific traits, economical and achieve measurable results.  No one, anywhere, has been able to prove any of these things are true.

On the other hand, very knowledgeable people who have done serious research into the issue have found problems with the notion of Influentials. None more so than Duncan Watts, one of the seminal figures of social network theory, who positively rails at the notion of anonymous mavens and influentials.

Watts points out that, while there are some people are more connected than others, a viral chain consists of more than just two steps, and his research shows that highly connected people are only slightly more likely to initiate long chains.  He also points to the work of Solomon Asch, which suggests that it is groups of people, not individuals, who influence.

It is odd, to say the least, that the high stakes, data driven, numbers crazy realm of marketing is so willing to take the Influentials myth on pure faith.  Maybe it is because the image of Influentials running around doing their work for them is so compelling.

The Curious Case of Revolutions

While I admit that indulging in fantasies about mythical people with superhuman powers of influence has its attractions, I can think of far more powerful images of the power of viral messages.  For instance, this one:

Or this one:

Or even this one…

One thing that the Color Revolutions, the Arab Spring, and the Occupy movement all have in common is the general impotence of so-called Influentials.  Although these grass roots movements spread quickly and had enormous impact, even to the point of transforming whole societies, it’s unclear, even after the fact, who drove events.

Having had first-hand experience in this area, I can attest that the picture doesn’t get any clearer at the ground level.  Moreover, as I’ve pointed out before, Malcolm Gladwell, the pied piper of the Influentials concept, who is normally an astute observer, turns into a babbling idiot when he attempts to explain these movements.

The Importance of Network Structure

What all the blather about Influentials fails to account for is that influence doesn’t apply to individuals with specific attributes, but is a function of how information travels within and between groups.  We’re learning how to track that, but it’s far more complicated than the facile Influentials concept.

For instance, the day after the 9-11 attacks, the US government released not only the names, but the leadership structure of the terrorists.  Unfortunately, as Valdis Krebs shows in this paper, the approach is incredibly computationally intensive even for a few dozen people and, at this point at least, impractical for the large groups that marketers are concerned with.

There are, however, some promising alternatives.  Duncan Watts himself has proposed something he calls Big Seed Marketing that is much more in line with research into how networks function.  Another idea, demonstrated in this Harvard flu study, is to exploit the friendship paradox.

One thing is clear, if you want to exploit network effects, forget about Influentials.

A Workable Model

While the Influentials concept is a dud, influence is very real and we’re learning more every day about how ideas spread.  While much of it involves complicated, often counter-intuitive network math, there are some basic principles that are easy to apply.

Passion:  Every movement is built on the passion of true believers and igniting that passion is the key to viral marketing.  However, inspiring devotion is not new nor is it mysterious.  Brands like Apple, Harley Davidson and Trader Joes, for instance, have been able to build an army of passionate followers without an “influencers” strategy.

Density:  It’s been a longstanding principle in biology that organisms grow out of a substrate.  In a similar way, social messages flow through dense networks much more efficiently than sparse ones.

There are social analytical techniques that can evaluate connectedness, but in the interst of brevity, let’s just say that there is a good reason that movies are launched in NY and LA, before the rest of the country.

Empowerment: Probably the most important principle of viral marketing is to give people a forum to share.  It’s no accident that hi-tech brands like Apple and e-Bay rely heavily on low-tech events where the faithful can meet face-to-face.  Harley Davidson, quite famously, has built an amazingly strong community from local clubs.

Social marketing, therefore, is not new, but we do have exciting new tools.  Twitter, Facebook and Google+ offer us new possibilities to promote, encourage and track word of mouth marketing.  However, that is no reason to abandon good sense.

Brands are not built by influential consumers, but by influential ideas.

- Greg

8 Responses leave one →
  1. November 17, 2011

    I do not disagree completely with your thesis. However, labelling Influentials as “blather” or cavalierly dismissing the concept as “a dud” doesn’t have the talismanic effect you seek to achieve.

    For it is the “spark” or law of unintended consequences that we find lacking in your article. Revolutionary turning points are often undertaken by unknowns, or slightly knowns and then those ideas are taken on by the masses as a whole. We can see this throughout history. When we examine the French Revolution for instance, many lesser knowns took on the task of re igniting the passion within each citizen, then they were killed off, or disappeared, and their place was taken by someone else. Eventually, over a decade, a leader did emerge, a true influential: Napoleon. But history still pays tribute to the lessor known influentials before him.

    So maybe that is the point and concept we should take away from this notion? That a small influencer, take Rosa Parks for instance, can have a large impact. Not because of the size of his her network, but because of the passion their small act has upon the body politic.

    Managing multiple accounts and identities on Twitter and Facebook allows me the unique opportunity to learn that some things simply take off more than others. A Colleague tweeted a funny tweet the other day. To Wit: @Samuel_Clemons Holiday Update, turkeys are pissed … that’s it. A Simple Tweet, it was retweeted 17 times and favorited about 20 times!! A simple message resonated. It is a timing thing too. A week before Thanksgiving, yes, many turkeys are probably really pissed.

    Other times, we tweet excellent content and links, and nobody even bothers to retweet or comment. Simply put, the impact of social media is and always will be audience participation, it is on the receiving side that our passion, our thoughts, our attitude will show through. When we relate a notion, or spirit that has commonality among the audience, when it relates well to a vast majority of the audience, they will respond. We have to have the ability to relate to the common soul and that is about all the influence we can ask for at any given time.

    THANK YOU very much for taking the time to pen this most remarkably researched article, and taking the time to share it. I appreciate your efforts,

    Lonny Dunn Tweets @ProNetworkBuild

    [Reply]

    Greg Reply:

    Thanks for sharing, Lonny.

    Greg

    [Reply]

  2. November 19, 2011

    Greg,

    What an interesting post. As one who teaches the Process & Effects of Mediated Communications, this post really resonated with me.

    I’m not so sure that because we can not prove the existence of “influentials” means they don’t exist. Someone is stirring the pot. Someone is more passionate about an issue than most. Someone is causing people to think in a particular way vs. another.

    Certainly, you are not discounting folks like the conservative talk show folks from being influential wrt to how people form their ideas.

    I also know people who are ardent FNC viewers that mimmick the views expressed by this cable channel. Are they (FNC) not being influential?

    I think there are ACTIVE people and PASSIVE people.

    Actives tend to set the agenda for the passives. And in each silo, there is an active setting the agenda for those passives.

    Tim Pool had a LIVE feed of OWS. Would you not consider Tim an “influential” when it comes to this movement? His feed was picked up by TIME, the NY Times and others. He, in my opinion, was influencing a lot of people with his vantage point of the activities going on in NYC.

    Let’s not throw out the baby with the bath water.

    The technology is most certainly changing how we influence, but certain people are using this new platform to influence more than others. And they are, again in my opinion, Opinion Leaders.

    Dick Taylor
    Assistant Professor, WKU
    School of Journalism & Broadcasting

    P.S. Thank YOU for your posts. I am learning more and thinking harder about mediated communications because of your writing. Happy Thanksgiving Greg to you and your family.

    [Reply]

    Greg Reply:

    Dick,

    You are surely right. I glossed over that distinction both in the interest of brevity and because I was specifically referring to the usage that marketers employ. There classes influential people. For instance:

    - Politicians and community leaders like clergy, etc.
    - Celebrities
    - Doctors

    However, that is clearly not what marketers mean when the speak of “Influentials.” We have terms like lobbying,
    celebrity endorsement and trade marketing for those activities, so we have no need to describe mysterious “Influentials.”

    There are also people who have influence in organizations (as, in the example I gave of a terrorist network.) However, that type of influence is due to their position in the network (i.e. smokers in an office transmit information very efficiently). We can identify these people as well, not by a particular trait, but through network analysis which I alluded to above and described in more detail in other posts (i.e. http://www.digitaltonto.com/2011/how-social-network-analysis-solves-real-world-problems/)

    Sorry for the confusion.

    - Greg

    [Reply]

  3. January 16, 2012

    I think what the study of influence shows us is that there can be many different factors: i.e., Milgrim’s studies of minimal group and his authority experiments, reciprocity/gifting experiments, Ash’s experiments with social proof, and yes, even Katz & Lazarsfeld’s 2-step theory. No one theory obviates the others, but simply shows that yes, these other elements can be strong factors in influence.

    You say, “such people would have to be identifiable by specific traits, economical and achieve measurable results,” – OR – they can be identified using research with data. It doesn’t require a lot of research to unearth the five most influential people online with a fringe subject, say periodontal therapy – and then work to include those individuals in your own communications. This doesn’t suggest that those five are influential to a broad mass of people, either – but instead, to your identified hyper-segment.
    Ric Dragon´s last blog post ..Google Logo Colors: What Would Come Next?

    [Reply]

    Greg Reply:

    I think there is some confusion. The question isn’t whether we can identify individuals (i.e. Malcolm Gladwell) who are influential, that’s pretty easy to do, but whether we can target influentials as a segment, which we can’t without doing SNA analysis.

    Greg

    [Reply]

    Ric Dragon Reply:

    We might be talking about different things… but:

    We have designed what we call influencer projects within social media marketing, wherein we identify a pool of influencers, and work to engage those people on behalf of a brand. The target metric might be that xx amount of those influencers mention or discuss the brand on their blog, Twitter, or other social platform. It could mean, for instance, that when a journalist discusses a certain top and mentions a couple of brands, that our brand is one of those discussed. There’s a lot more to it than that – but that is the beginning.
    Ric Dragon´s last blog post ..Google Logo Colors: What Would Come Next?

    [Reply]

    Greg Reply:

    Again, the problem is that so called “influencers” aren’t that much more influential than anybody else. See this paper by Duncan Watts: http://research.yahoo.com/pub/3386

    Greg

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS

CommentLuv badge