Contact Us

Who you spend time with really does matter

by Nigel Hollis | June 13, 2011

Paul Adams, global brand experience manager at Facebook, states:

The people closest to us are often the ones who influence us most.

I am sure he is right, but maybe we should consider an alternative interpretation to the word “closest” when it comes to word of mouth.

Paul Adams proposes a rough classification of influence from strongest to weakest: family and friends, people with who they have strong affiliation, groups of similar people, very large groups of people, and finally, least influential, random strangers. The continuum seems to be based on appreciation of the other people, from strong emotional connection to weak.

But before we start twittering on about the influence of word of mouth, let’s consider what Professor Alex “Sandy” Pentland of the Massachusetts Institute of Technology (MIT) Human Dynamics Laboratory has to say about influence.

Sandy appears to support the “monkey see, monkey do” school. He recounts an experiment that tracked the voting intentions of people in a college dorm. One third of people changed their political views based on face-to-face encounters. Overall, who the students spent time with correlated with their voting behavior. In another case, Sandy suggests that if you hang out with people who routinely eat three slices of pizza, guess what? You too will start to eat three slices of pizza.

It should come as no surprise that we humans are incredibly susceptible to the behavior of others. In an effort to fit in, we sometimes consciously, and often non-consciously, model our behavior on those around us. We want people to like us (and I do not just mean on Facebook).

This fact starts to suggest that when it comes to influencing behavior change, demonstration may be far more important than communication. Perhaps “closeness” should be defined as much by physical distance as emotional connection. At a specific point in time, the people who have the most influence on your behavior might be your colleagues, fellow shoppers or the crowd around you at a football game.

When it comes to measuring word of mouth, it makes me wonder if we need to start adding some new dimensions to fully capture the effects of social interaction. Alongside reputation, maybe we need to consider duration of exposure? Perhaps by profiling who a respondent hangs out with we can predict what influence they will have in terms of brand choice. A change in social or work context could signal a future shift in behavior. Any differences in brand choice between the individual and the people they associate with, should tend to diminish over time as they adopt the behavior of those around them.

So what do you think? Daft idea or worth exploring further?  


Leave a comment
  1. Evgenia Grinblo, June 27, 2011
    I can't help but wonder whether the intention matters here. If a person chooses to interact or be surrounded by certain people, wouldn't it make sense that their influence is stronger? This is especially relevant in the social network scene where seeking out information (following others on Twitter) versus receiving unsolicited updates (from Facebook friends who initiated the connection themselves) could have different results. Who knows, I'm just thinking out loud here.
  2. Nigel, June 15, 2011
    So if we agree on this one I wonder how best to integrate that into our research?
    Dom, I definitely believe that the power of digital is in speed and reach not actual impact. See this post.
    Cheers, Nigel
  3. Dominic, June 14, 2011
    I completely agree, and the research by Nicholas Christakis in his fascinating book Connected finds the same thing.
    It raises interesting questions about the power of digital social networks v 'real'ones.
  4. Phil Herr, June 13, 2011
    Agree. Goodness, we have been taught that non-verbal communications constitute the biggest factor in determining how people react to our messages. By applying this to the concept of "influence" this opens the discussion about what constitutes so-called word of mouth. So far we have been comfortable to accept definitions based on literal verbal communications -- oral or written. But as the MIT study points out, behavioral cues can be far more subtle and yet more pernicious at the same time. And so, I'd suggest the issue be opened to consider all factors that constitute interpersonal communication, including body language and actual behavior.

    Leave a comment