Brand Marketing: Powerful Networks And Weak Ties

Walker SmithSeptember 18, 20129 min

A recently published study in Nature has been making minor headlines with its findings that close Facebook friends can boost the likelihood to vote. This study has caught the eye of brand marketers because these findings seem to validate the investments being made nowadays in viral marketing, particularly those involving online social networks. But it shouldn’t come as a surprise that people are influenced by their close friends. Instead, what’s surprising about this new study is just how weak this influence tends to be.

On the day of the 2010 mid-term elections, Facebook featured a News Feed message encouraging people to vote, with a link to find polling places and an “I Voted” button to click on. Most of the 60+ million people who visited Facebook that day also saw profile pictures of friends who had already clicked the “I Voted” button. Researchers carved out two randomly selected comparison groups, one that got everything in the News Feed except the profile pictures of friends and another that got nothing at all in the News Feed. The analysis looked at the impact of the message, both with and without profile pictures of friends, on the likelihood that someone would subsequently click on the “I Voted” button.

By the researchers’ calculations, the message with profile pictures of friends generated an additional 340,000 votes, of which 60,000 were directly influenced and the rest through a chain of social contagion. Moreover, the friends that accounted for the social contagion were close friends, hence, the headlines that strong online social ties influence the likelihood to vote.

But the researchers were mindful that the results were not unambiguous. Clicking the “I Voted” button is a far from a perfect measure of actual voting. More problematic is causality. It’s unclear whether these 340,000 people would have voted anyway and simply saw the Facebook News Feed first by serendipity. Most problematic is figuring out whether Facebook added anything unique. Since the friends that mattered for social contagion were close friends, some portion of the measured Facebook effect may not be due to Facebook at all but to other forms of equally influential interaction, an issue the researchers noted.

While these seem like nitpicking criticisms, they go straight to the heart of what’s at stake for brand marketers. The 340,000 additional votes constituted a mere 0.14 percentage points of the 2010 voting age population – by any standard, miniscule. Discounting some portion of this fourteen one-hundredths of a percent for the reasons just mentioned yields an even tinier effect.

The researchers acknowledged the infinitesimal size of the effect observed, which they admitted they were able to find only because they were working with a very large dataset. But they papered over this with two rhetorical flourishes. First, they noted that voting participation rose 0.6 percentage points from 2006 to 2010, and 0.14 is a sizeable part of 0.6. But it is a speculative assumption to presume that the 0.14 percentage points consist only of people who didn’t vote in 2006.

Second, the researchers pointed out that the 2000 election was decided by an infinitesimally small number of votes in Florida. But an extreme outlier does not establish average value, not to mention that if Facebook-generated votes had been cast in the same proportion for Gore and Bush as all other votes in Florida – the most plausible assumption – that election would been unchanged. The infinitesimally small Facebook effect found in this study is irrelevant even in an election decided by an infinitesimally small margin

This digression is simply to make the point that there is a finding from this research more fundamental than the one making the headlines, indeed, one that runs counter to these headlines, to wit, social networks are best understood these days in terms of weak ties not strong ties. What this study really shows is that even our strong ties are weak, a finding consistent with other landmark studies of contemporary trends in social networks.

Perhaps the best-known study about social connections is an analysis published in 2006 of historical data from the General Social Survey (GSS). This study made major headlines with its finding that people reporting no close confidants with whom they discuss important things in their lives had grown from 10 percent in 1985 to 24.6 percent in 2004.

What didn’t make the headlines, though, was the explicit reluctance of these researchers to conclude from this one data point that the percentage of people with no close confidants was growing, a reserve born of various methodological issues plus their unwillingness to draw such a dramatic conclusion without additional confirmation. Instead, the research team pointed to another finding they felt was more important.

The GSS asks respondents to check off which roles are fulfilled by each relationship. The surprising finding was that so-called multiplex relationships are rare, even for very close contacts. For example, a minority of respondents reported any links with their spouse beyond marriage, such as friend, advisor, coworker or co-member of another group. Very few reported more than two links. In short, even our closest relationships are narrowly and thus weakly defined.

The view that emerges from this and other research is that, increasingly, we live in a world of weak ties. Weak ties not strong ties define contemporary life and shape the ways in which people engage in everything, commerce included.

Of course, to speak of weak ties makes it sound as if people are isolated and disconnected. Not true. The reality is that people are reinventing their lives in full, satisfying and highly connected ways in the context of weak ties.

Indeed, the heartening paradox is that weak ties are sewing the seeds for stronger connections. To a large extent, this is nothing but scarcity at work. Too little of something people want gives rise to a boom in demand for it.  Ties are weaker but the desire for social connection is as strong as ever. So in a world of weak ties, people are more focused than ever on social engagement and the social currency that goes with it. This is the predominant motive force behind the emergence of the Kinship Economy, which I have written about before.

Critics have been making much of the fact that many people report feeling lonely and that more and more people are living alone. The Internet is often blamed as a prime culprit in social isolation.

But there is no evidence that loneliness is on the rise or that the Internet is isolating people. Other results from the GSS show that people who live alone are more socially engaged than married people. They socialize more, interact more with strangers, and are move involved in civic organizations. Additionally, other research established clearly over a decade ago that greater online social engagement is associated with more not less offline social engagement. In fact, loneliness is a phenomenon that tends to cluster not spread.

Weak ties are not bad; they’re just weak. This is important for brand marketers because too little attention is given to weak ties. Marketers are overly focused on leveraging strong ties when it’s weak ties that typify life today. Malcolm Gladwell’s 2000 bestseller, The Tipping Point, is one of the biggest reasons behind this marketing orientation.

Gladwell popularized a view of social networks that emphasized the strong ties of so-called Mavens and Connectors in bridging the gap for trends to take off and spread like wildfire. This emphasizes nodes in a network, or the key people who get the word out on the next big thing. As a result, brand marketers have invested enormous amounts of money and time to locate, woo and incentivize the people who constitute these key nodes in social networks.

But what if the effects reported by Gladwell are due instead to characteristics of the network as a whole not certain nodes within it? If so, brand marketers are looking for clout in all the wrong places.

Gladwell’s book was a layman’s guide to decades of previous social science research, but new work in the decade following the publication of his book has enjoyed the advantage of improved interdisciplinary analytics and far better databases, particularly those of online networks. While much of this new research has confirmed previous work, just as much of it is upending prior knowledge about social influence.

At the forefront of this new research is Duncan Watts, formerly of Yahoo! Labs, now with Microsoft Research. In his recent book, Everything is Obvious: *Once You Know the Answer (chapter 4), he debunks both Gladwell’s tipping point notion and the broader idea of so-called influentials. Watts notes that studies show instead, “when influence is spread via some contagious process, the outcome depends far more on the overall structure of the network than on the properties of the individuals who trigger it.” Simply put, brand marketers need to study network properties not network nodes, and weak ties are the defining property of networks today.

Here’s a way to think about it. There are two overarching theories of word-of-mouth. One is the traditional ‘opinion leader’ theory. The other is the emerging theory that I refer to as the ‘general opinion’ theory.

An opinion leader approach looks for people with strong ties or followings. The task is to get them to echo a brand’s message as part of what they are saying to other people.

A general opinion approach diagnoses the network itself. The task is to get a brand’s message into the right kind of network so that it becomes part of the broader context to which people are exposed.

An opinion leader approach means getting consumers to pay attention to particular people.

A general opinion approach means getting consumers to pay attention to the broader context.

Watts offers a hint of what this means. He writes that research shows that trends break out when they intersect with “a critical mass of easily influenced people” [emphasis in original], in which case “even an average individual [is] capable of triggering a large cascade,” a phenomenon that “has nothing to do with a few highly influential individuals at all.”

In other words, strong ties are not necessary to tip off trends. Certainly, strong ties can do it, but so can weak ties. What’s key, though, to word-of-mouth and viral success has nothing to do with the strength of ties. Rather, it is all about the right moment for the right network.

The challenge for brand marketers is efficiently leveraging the ties that exist once the right network is found. That means knowing more about weak ties, and this means a significant shift in focus, orientation and brand marketing practice, one that is less about finding opinion leaders to carry the message and more about finding networks ready for what marketers have to offer. Even weak ties will suffice if a network is ready, which is encouraging since weak not strong ties are the ties that bind these days.

The Blake Project Can Help: The Brand Positioning Workshop

Branding Strategy Insider is a service of The Blake Project: A strategic brand consultancy specializing in Brand Research, Brand Strategy, Brand Licensing and Brand Education

FREE Publications And Resources For Marketers

Walker Smith

One comment

  • Aolsz

    September 19, 2012 at 9:33 am

    I read with great interest your discussion of Watts’s theory that “trends break out when they intersect with ‘a critical mass of easily influenced people.'”

    Some years back, I worked on a political project that crafted an avant-garde, a cadre. The results are discussed here:
    http://www.bretschundler.org/campaignsandelections0801.shtml
    The intent was to form a group that would receive, amplify and transmit a message in the same way that a radio signal is propagated through the ionosphere. This approach implemented and melded both the General Opinion and Opinion Leader concepts. The network that we put together already was amenable to our message. Likely prospects were invited to attend talks and then were invited to provide contact information for more information and future events. The subsequent emails (or phone calls) at the very least contained talking points. Often — and increasingly so as election day approached — there was a call to action — provide the names and email addresses of like-minded family and friends and email/phone individuals in your personal circle who share the same ideas. In this way, the network was self-replicating with each of the members soon becoming the center of their own set of followers who were not in direct communication with the campaign.

    From a distance, my impression is that Procter and Gamble does something very similar with their Tremor:
    http://www.tremor.com/creating-effective-advocacy/methodology/

    Anthony Olszewski

Comments are closed.

Connect With Us