Skip to content

4 Reasons To Kill Influencer Marketing

2014 October 1

Marketer’s like to repeat the quote, “I know I waste half of my ad budget, I just don’t know which half.”  No one knows who first said it—it’s been attributed to a number of people—but the fact that it gets repeated so often is testament to how strongly it resonates.

So it shouldn’t be surprising that marketers like the idea of “Influentials,” seemingly ordinary people who determine what others think, do and buy.  A recent study of 1300 marketers found that 74% of them planned to invest in influencer marketing over the next 12 months.  That is truly an incredible amount.

However, there’s good reason to believe that it’s all a waste of time and effort.  While the idea of influentials may be intuitively convincing, there is very little, if any, evidence that they actually can improve performance—or even exist at all.  So before you embark on another influencer campaign, consider these 4 reasons why you shouldn’t.

1. It’s The Wrong Metaphor

Malcolm Gladwell is probably the person most responsible for the massive interest in influencer marketing.  It was he who, in his blockbuster book, The Tipping Point, laid out his now famous “Law of the Few,” which he stated as:


The success of any kind of social epidemic is heavily dependent on the involvement of people with a particular and rare set of social gifts.


The Influentials idea makes intuitive sense because we all know people like the ones Gladwell described in his book:  “Connectors” who seem to know everyone, “mavens” who possess deep domain knowledge and “salesmen” who have the gift of gab.  We’ve seen how they’ve influenced us, so it seems plausible that they play a role in spreading ideas.

Yet social epidemics aren’t local phenomena.  They are long viral chains.  Just because someone might be good at getting an idea across, doesn’t mean that others are more likely to share the idea.  And if an idea doesn’t get shared, it doesn’t travel far.

A more accurate metaphor would be a wave at a stadium.  What “special traits” would it take to affect thousands of people throwing their arms up in sequence?  Could a 400 pound man do it?  If Jack Nicholson refused to stand up at a Lakers game, would a wave stop in its tracks?  Not likely.  Collective behavior requires a collective.

2. Science Finds No Evidence To Support Influencer Marketing

While Gladwell’s book certainly did much to popularize the notion of influentials, the idea is not exactly new.  In fact, it goes back to research done by Katz and Lazarsfeld, two prominent sociologists, in the 1940’s and 50’s.  Yet even in their original study, they found that influence was highly contextual.

Recent research raises even more serious questions about the influentials hypothesis.  In one study of e-mails, it was found that highly connected people weren’t necessary to produce a viral cascade.  In another, based on Twitter, it was found that they aren’t even sufficient.  So called “Influentials” are only slightly more likely to produce viral chains.

Duncan Watts, a researcher at Microsoft who co-created one of the most important models of how social networks function, says, “The Influentials hypothesis, is a theory that can be made to fit the facts once they are known, but it has little predictive power.  It is at best a convenient fiction; at worst a misleading model.  The real world is much more complicated.”

The empirical evidence is clear:  It’s time to debunk the myths about influentials.  Unless someone, somewhere, can produce evidence that these “special” people can further our marketing campaigns more efficiently than other approaches, we shouldn’t waste money chasing them.

3. Common Sense And Experience

So far, we’ve seen that the idea of influentials isn’t as intuitively appealing as it first seems. We’ve also seen that scientific evidence contradicts the viability of influencer marketing.  Yet intuition is always fallible and scientific studies, even if rigorously and carefully undertaken, can be wrong.  Real life doesn’t always align with what happens in controlled experiments.

But there is another reason to doubt the idea of influentials: recent events and common sense.  We’ve seen powerful social epidemics erupt in the Arab Spring, the Euromaidan protests in Ukraine and the 2004 Orange Revolution that preceded it.  Small, loosely connected groups overthrew powerful regimes.

Now, it hardly makes sense that Hosni Mubarak and Viktor Yanukovych, who controlled the media and the major organs of power, lacked influence or access to influential people.  Yet they were powerless to stop the street protests that eventually brought about their downfall.

It is, of course, possible that the protestors were driven by people with “rare social gifts” that trumped the dictators’ more traditional influence, but then why did those gifts fail them in the aftermath?  In Egypt, the Muslim Brotherhood, not the largely liberal protesters, prevailed in the sunsequent elections.  In Ukraine, the Pora movement never evolved into a political force.

3. Buzzfeed’s Viral Success

The fundamental problem with influencer marketing is not that some people aren’t more influential than others, but that there is little, if any, evidence that influencer strategies—other than celebrity endorsement—are viable.  Yet all is not lost.  There is a way to consistently increase the likelihood of viral chains.

In 2001, Jonah Peretti had an e-mail exchange with Nike that went viral on the Web.  He was fascinated and, a year later, he met Duncan Watts at a conference.  The two struck up a friendship and then a collaboration.  They did a number of projects together that had promising results, which they published in Harvard Business Review.

Their approach, which they called big seed marketing, does not rely on identifying a small number of special people, but rather on harnessing the power of a large number of ordinary people.  By reaching a mass audience, and encouraging them to share, you increase the likelihood that a viral chain emerges and, even if it doesn’t, you still improve performance.

Peretti went on to co-found Huffington Post, which was sold to AOL for $315 million in 2011. His second company, Buzzfeed, is now valued at $850 million.  In his extended interview with Felix Salmon, he credits not influentials, but “a constellation of connected things” for making his articles go viral so consistently.

So if you want your idea to spread, forget about special people with “rare qualities.”  Be interesting, reach as many people as you can and encourage them to share.

– Greg


An earlier version of this post appeared in Harvard Business Review

12 Responses leave one →
  1. jean-louis permalink
    October 1, 2014

    Hello Greg, things may have changed since Edward Bernays, a nephew of S Freud, published “Propaganda” in 1928.

    Still like to reread it and see how a few can influence a lot (e.g. doubling the market for cigarettes, getting corduroy to get a second life etc etc and having women wear furs amongst others)


  2. October 2, 2014

    Hi Greg,

    I’ve loved reading your blog for some time now. Please keep up the great work.

    I’m really interested in influencer marketing at the moment. What your hypothesis doesn’t seem to take into account are the new wave of amateur opinion formers in special interest areas (bloggers much like yourself). We have quite a lot of evidence that investing in these guys can get your message across twice as efficiently as traditional paid media and (it would seem) often much more effectively.

    What is this, if it isn’t influencer marketing? Would love you thoughts. Thanks!

  3. October 2, 2014

    Hi Lawrence,

    What you seem to be talking about is alternative media placement and it can be effective, but it is no different than trade marketing. There’s nothing wrong with targeting your message or rightsizing your media placement.

    But notice, you’re talking about bloggers with an audience in special interest areas, not people with “rare social gifts” who can spark social epidemics, just media with access to a particular audience.

    – Greg

  4. October 2, 2014

    Fair point, Greg. I don’t think we’ve ever started one of our “influencer” campaigns with the expectation that we would start a viral epidemic. I guess (as always) the argument will come down to definitions. While I agree that there are no magical individuals that, once harnessed, can start epidemics (except possibly celebrities) I don’t think we should ignore the value of influential people in efficiently getting your message across. This is what I think most people are describing when they talk about “Influencer Marketing”.

  5. October 3, 2014

    Yes, I think that’s true. Clearly influence exists. For example, CEO’s and CMO’s have influence on marketing budgets. Journalists and pundits have influence over people’s opinions in their respective areas. Oprah can have a major influence on book sales.

    Yet there isn’t anything new or mysterious about this type of targeting. So I think the problem comes when people start believing that unicorns, like Gladwell’s “Law of the Few” or worse, when snake oil salesmen tout their ability to identify and exploit such people.

    – Greg

  6. Mary Alberts permalink
    October 7, 2014

    Hi Greg,

    Interesting article, and I’d agree with a lot of it – influence (and influencer marketing) is a highly contentious topic these days, especially when it comes to brands and social influence.

    However, I’d counter the belief that influence doesn’t exist, or influence marketing (when implemented properly) isn’t effective and a real business solution. I notice you linked to the influence marketing study by Sensei Marketing. I would highly recommend you read the book “Influence Marketing” by Danny Brown and Sam Fiorella, which goes beyond Gladwell’s questionable definition of influence, and into what real influence looks like, what it means for brands, and where it fits in with the purchase life cycle of consumers. As a brand marketer for a large retailer of soft drinks, I’ve used the framework in the book and it’s very much the “real deal”.

    Thank you for inspiring my thoughts.

  7. October 7, 2014

    Thanks for sharing Mary.

    – Greg

  8. November 3, 2014

    Hi Greg,

    Thanks for sharing this article! I think one of the most important sentence here is that influence is highly contextual. From my point of view there’s a context for everyone to become an influencer (which will eventually lead to harness the power of ordinary people), however some have more impact on their peers than others in certain topics.

    My startup, Brandvee uses network science to find influential visitors on websites. We found that there are a few people who possess extraordinary capabilities for driving referral traffic (to both media and ecommerce sites), which can be also incentivized with proper campaigns. We separated the visitors into three segments:

    – Silent mass
    – Low virality sharers
    – High virality sharers

    Around 10-30% of all people were able to drive referral traffic at all. However the top third of these “sharers”were responsible for 80% of referral traffic.

    So to maximize the impact of influentials, you’d need to find the “high virality” segment. These people are ordinary people, but something (mostly the context, the topic, etc.) makes them being able to drive more traffic.

    What do you think?

  9. November 3, 2014


    Yes, influence does exist and it can be identified, either through social network analysis (which is why the NSA wants metadata) or through the type of statistical analysis you describe. What doesn’t work is looking for some mysterious class of people who have “rare social gifts” that enable them to influence the rest of us.

    Incidentally, there is a strange quirk in the math called “the friendship paradox.” Our friends are more influential than we are (and richer too).

    – Greg

  10. November 3, 2014

    Thanks for your reply Greg! I agree. That “rare social gift” is most likely to be nonexistent, and it’s all about context.

    (This is the single most amazing topic I’ve worked with in the last few years, and I’m really excited about what else we might find on our journey.)

  11. November 3, 2014

    Good luck with it David!

    – Greg

  12. Loop Garoo permalink
    May 24, 2015

    Think you are missing the point, Jean-Louis. Obviously everything starts somewhere, and in retrospect, you can trace back to one, or a few, originators. This does not mean those people are a special kind of person called an “influencer”, just that they are the first people who happened on something people found useful or interesting — a way to make fire, knap flints, fletch an arrow, etc. For every one such person, there are thousands who happen on things that nobody else finds useful or interesting. Problem is that there is no way to predict who will be an “influencer” in the future, in any given context, and that is why influencer marketing is useless. Interesting that you mention Bernays and Freud, as both were pseudo-scientists oblivious to the silly, circular character of their arguments.

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