Understanding Influence
On Influence
On Influence

I don’t watch a ton of TV, but I do have a few favorite shows. I love AMC’s new series, Rubicon. I watch a lot of Top Chef with my five-year old, because he likes cooking and I like Padma. I generally catch Bri-Wi when I’m home.

So here’s a test: I’m going to name five shows, and I want you to rank them in order of popularity. Ready?

NBC Nightly News
Top Chef
American Pickers
The Closer
Mad Men

Where would you go to find this information? Why, the folks at Nielsen, of course, who measure this sort of thing. Here’s what they had to say (ranked by estimated total audience for last week):

The Closer: 7,209,000
NBC Nightly News: 7,170,000
American Pickers: 4,866,000
Top Chef: 2,898,000
Mad Men 2,312,000

Any surprises to you? I was actually most surprised by The Closer – I knew it was a hit, but those are pretty good numbers for a cable show. Now, all of these numbers are estimates, but they are reasonably good ones – the folks being measured worked with Nielsen to come up with a solution that benefited everyone – not a “black box,” proprietary algorithm, but a good, solid sampling methodology. Why bother? Well, knowing how many persons watched a given show at least demarcates the boundary for how many of them saw a given commercial message, and that’s a pretty important thing for both the networks AND advertisers to know.

Here’s another question: if The Closer is three times more popular than Mad Men, why would anyone advertise on the latter when the former has a significantly greater audience? Again, a simple answer: both parties measured something that matters – the composition of their audiences – so that products and services that were more appealing to Mad Men audiences (like the Chase Sapphire card) were shown during Mad Men. Context is an important part of engagement, and it’s all in the service of optimizing a limited ad spend. Again, all pretty proven and basic stuff here.

So, a third question: Which of the five shows is the most influential? What do you think? Let’s say for a moment that you believe Mad Men is the most influential TV show on my list. Influential how? Influential to whom? And in what context?

What does influential even mean?

By itself – nothing, of course. Influence requires an actor, action(s) and a context. Mad Men is arguably not more influential than The Closer, but Mad Men might resonate more strongly with you, and that resonance might make you more receptive to other things by association. That receptivity, in turn, is measurable in both overt and subtle ways. Did Mad Men influence me to change my Twitter avatar for a while? Definitely. Did it influence me to start smoking? Definitely not. Did it influence my martini consumption? Maybe. OK, definitely. You get the point. (The same is true of people, by the way. Is Barack Obama more influential than my wife? If you are married, you know this is a trick question.)

Devoid of a specific context, actors and actions, how would you arbitrarily measure whether or not Mad Men is more influential than, say, American Pickers? You could hazard a guess (let’s say, “3X,” which I’ve rectally derived), recite a few anecdotes or perhaps provide some other qualitative measures. You could total the mentions of both shows on various social media platforms, and even index them, if you’re clever. But there’s no way to check your work. Despite decades of measuring reach, frequency, demographics, psychographics and now engagement, there isn’t a Nielsen “Influence” rating. There isn’t an influence rating for a very simple reason – it would be really hard to figure out correctly, and figuring it out would net you very little in return. It isn’t a salable metric. All that matters to the advertisers on Mad Men is this: how many people saw my ad, who were they, and what happened as a result.

In fact, almost any use case for an influence measure is really a proxy for something else you’d rather measure: a purchase, an increase in loyalty, a shift in awareness or attitude, or the movement along a continuum from awareness to consideration to information-seeking to conversion. So we measure those things instead. Where influence may matter is with a specific life group – are we “moving the needle” with empty nesters? Young wealth acquirers? Recent retirees? In every case, however, a measure of influence is really a measure of something else (“the needle”), and that measure is meaningless unless it is accompanied by a context.

Influence is a derivative – an abstraction that’s exponentially more difficult to ascertain than the more useful measures it’s meant to amalgamate. To make meaning of “influence,” I have to know more than a simple number – I have to be able to know something about you – quite a bit, actually.

Here’s something about me: my Klout is 39 out of 100. My Tweetlevel is 60 out of 100. My Twitter Grade is 100 out of 100.

What do you know about me, now?

What are these numbers proxies for?

Enhanced by Zemanta

About the Author

Tom Webster
Tom Webster is Vice President of Strategy for Edison Research, sole provider of U.S. National Election exit polling data for all major news networks. Webster has 20 years of experience in market and opinion research, with a particular emphasis on consumer behavior and the adoption of new media and technology. He is the principal author of a number of widely-cited research studies, including Twitter Usage In America, The Social Habit, and The Podcast Consumer Revealed, and is co-author of the Edison Research/Arbitron Internet and Multimedia Research Series, now in its 18th iteration. Reach him on Twitter at Webby2001, or on his blog at BrandSavant.
  • Pingback: e-mail software()

  • Pingback: cloud web service()

  • Pingback: On Influence | Social Media Explorer | Influencing the Influencer()

  • Love it. I suggest that influence will change fluidly with both the message and the prevailing world circumstances. Therefore, it ought to be re-assessed frequently.

  • Pingback: Friday Social Media Reading List (1/21/11) | Fitzternet()

  • Pingback: HeadRadical.com()

  • Pingback: 26 Twitter Strategy Tips And Tricks()

  • Pingback: On Influence | BrandSavant()

  • I have found that watching TV shows always makes me feel like I want to be like certain characters. Not that I want to actually BE them, but characters that embody characteristics and goals similar to mine always attract and inspire me. I want to be like hank moody from californication. Lol

  • “Influence is a derivative – an abstraction that’s exponentially more difficult to ascertain than the more useful measures it’s meant to amalgamate.”

    Boom goes the dynamite.

    I like the “in person” version of this post you shared with me over my Antelope Tartar (with the Quayle egg), but…

    DJ Waldow
    Director of Community, Blue Sky Factory

  • A very interesting discussion on influence, indeed.

    What you are describing is something I talk about quite frequently, which I've dubbed the “Peripheral Influence Payout” … Peripheral Influence Payout is simply the unintended ways your decisions will affect others.

    There really is no good human measure for influence, because the peripheral influence that your decisions have is dynamic and constantly evolving.

    For example, if a young girl sees Lady Gaga and latches on to her as a role model, she is influenced by her. Her decisions to follow Lady Gaga may inspire other young girls (perhaps in her classroom, even) to do the same. Thus, influence has the potential to be infinite, and it's hard to tell what influenced what (as you've said) without an actor, action, and a context.

    Very nice post. Thanks for sharing.

  • Love this post. I do a ton of work in the entertainment vertical, and we sometimes find that lower-rated, niche shows develop the most passionate fans. Our higher-rated show may have a lot of volume in social channels, but these smaller shows have fans who will spread information, create their own show-related content, and passionately defend the show on branded Facebook pages. Regardless, the impact of volume of fans on larger shows is incomparable, even if the fans are overall less passionate.

    This realization drives the never-ending debate on ad spend: What kind of viewer will be most impacted by advertising around a show? That is, of course, the end-game for networks. Wish I had an answer.

  • Oh wow, the topic du jour! :) Influence is very hard to measure. We do it online / in social media because it's easy (or easier, rather). It's not really easy, rather — it's just quantifiable. True influence is much more complex and truly multi-channel. Unless it's a low involvement purchase (like buying a pen — NOT a Mont Blanc pen :), our decisions are probably driven by several channels. You learn something from people around you, whether or not in response to a particular question you have, or just by absorbing the world around you. You do it offline and online. The relationship between a social media action or gesture of an “influencer” and an action on your part isn't linear. Paul Greenberg, a man I admire tremendously posted the following about influence – http://the56group.typepad.com/pgreenblog/2010/08/the-naked-tweeter-traditional-v-contemporary-channels-of-influence-or-is-it-and-contemporary.html – worth a read.

    I think, at least in my mind, that influence is a people-to-people action, not something that a show can do. Characters in a show can be influential on you, because they may represent something aspirational or attractive in some other way. Here, just like someone else said, it's highly contextual and situational. If I'm a certain way, and this person is this other way and I somehow am able to be influenced by a person like them, then they are influential (if that made any sense). Influence has to be temporally relevant too — grunge rockstars were influential for me in 1993, but not today :)

  • Tom, I love this post. I think ultimately it cuts to a core basic surrounding social media and its influence…just because some things cannot be measured, quantified or analyzed in a traditional sense should not take away from their value. Who knows the amount of “influence” external factors have on behavior or trends? But just because I can't accurately measure, doesn't mean I haven't been drinking a lot more martinis lately.

    • +1 on the martinis. Influence metrics and number of martinis are both best left uncounted.

  • Nice post Tom and, naturally, coming from you it was spot on.

    Couple of comments. Influence is context sensitive, therefore can only be measured within a very narrow range of context. What's important, in my humble opinion, is the analysis that allows you to derive that context…only then does a 'influence measure' carry any weight.

    I actually think Klout for example is a very useful tool. But that's primarily because I understand the context in which its scores are derived, and more importantly, pay attention to the individual numbers that make up the final 'score' more than that score itself. A score of 39, or 99, is pretty meaningless in and of itself. Just like knowing the reach of a Nielson rated show is pretty meaningless without that context. If I'm selling an outdoor survival tool I feel relatively sure my dollars are better spent on “Survivor Man” than they are on “60 Minutes” regardless of what those Nielson numbers say.

    The problem, is that Influence is akin to ROI in the fact that both are really *post* activity measurements. Things like Nielson ratings and Klout scores can only give us some insight into influence *potential*, it's up to us to take it from there if we want to create a qualitative program vs. quantitative….which is really why in the absence of extremely granular data it's as much an art as a science.

    • Matt – I'd love to see a post from you on how you dive into the numbers-within-the-numbers on services like Klout – I have no doubts that you can, and I think getting everyone to think beyond the initial, meaningless “score” would do all parties a great service.

      I'll have another martini while I'm waiting.

      • That's a great idea for a post actually.

        The best piece of free advice I could give with Klout is that you never start there. I use Klout as a filtering tool vs. using it as a selection tool which I would then filter secondarily. Don't know if that made any sense?

  • Really interesting post Tom and absolutely true. If I don't ever watch a show -however popular, then I'm far less likely to be influenced by it , but it's interesting to think about the way popular culture can influence the world around us as well. I'm actually a fan of Mad Men, and while it hasn't increased the number of Martini's I drink, they do seem to be more readily available around London recently – could be a coincidence of course…
    One of my colleagues recently wrote about social network design and the way this affects behaviour, you may find it interesting: http://bit.ly/aft41M

  • Dawson Fercho

    Interesting post. Most likely value is not being determined due to a lack of understanding a true measure of influence given the current marketing paradigm on mass marketing driven by popularity metrics. You brush against this with your sentence of: “All that matters to the advertisers on Mad Men is this: how many people saw my ad, who were they, and what happened as a result”. Let’s focus on “… and what happened as a result”. That is inquisitive of influence. Not to be confused with awareness (which, by social science reference, would be defined as a neutral state of positive comprehension — awareness minus sentiment) or mentions or trackbacks or any of the broad array of popularity-driven awareness-focused measures. And maybe this is a case of misinterpreted semantics, but why would one not be able to check their work in this realm? A Proper analysis and study of language meaning underpinning Natural Language Processing enables a corpus of communication to be examined to a very high degree of significance (and relevance!). Silent lurkers (those who don’t discuss the latest episode of Mad Men in the “Mad Men Fanatics” forum) can be measured by more secondary measures of conversation thematic (relating to a particular subject of concept) identification and radiation tracking. Also, if Influence is really just a proxy for more granular measures (which isn’t being argued here), what then is popularity (buzz, chatter, etc.)? I’d offer that the predominant standard measures (based on popularity) that are currently leveraged in the attempt at understanding socialization (defined as awareness leading to a preference leading to a behavior — measured in digital communication and in “meatspace”) are mostly meaningless for tangible strategic use. The reality exists that advertisers only have a measure of popularity in which to hang their hats – and for the Networks to know how to charge them for the hanger. Popularity is a subjective abstraction for guessing at the equation of “an increase in eyeballs (automatically) equates to a lift in sales”. “Bob, our Neilsen score was off the charts with the number of impressions from last nights ad spot!”. Ok, great. Now what? Without delving into the dynamics of influence, one cannot advance much further without undue speculation.
    One could easily imagine this coming down to a more concentrated debate about constructing context for a valued measure of true influence.

  • Good post,

    I do believe actors have an impact on your actions and thoughts, its good acting. Obviously how influential a programme is on you personally is dependant on the amount of interest you show it.

  • Pingback: Derivative Measures in Social Media | BrandSavant()


Social Media Jobs

VIP Explorer’s Club