How Much Do We Really Know About the Influencers?

Social media is a land of many holy grails - if it's not completely antithetical to have more than one - but probably none is more prized by marketers than the Influencer, that person who can sway opinion, get people to buy products and otherwise, well, influence the hearts and minds of dozens, hundreds and thousands of other people.

So far, in social media, this has been a relatively rudimentary exercise. Lots of Facebook friends? Thousands of Twitter followers? Scads of traffic to your blog? Great! Let's sign you up in the name of selling lots of product!

But two bits of intelligence I've read over the last few days show me that we still have a long way to go in understanding influence. Before I go on, here are some of the counterintuitive data points I've come across about the people that everyone wants to friend:

From ReadWriteWeb: That some of the most influential people on Twitter don't do a lot of retweeting. According to January and February data from Tableau Software and TwitterStats, @scobleizer (Robert Scoble) with more than 118,000 followers, retweets one percent of the time (though a quick look at his Twitter account indicates it may be somewhat higher); @stevecase however, retweets 28 percent of the time. He has more than 340,000 followers.

From ICOM (as reported by Mediapost): The idea of the universal influencer is a myth. Individuals can be influential in certain, siloed categories, but don't tend to have influence across all categories. Further, having influence has less to do with demographics or connections than it does with behavior. As the white paper about the study says: "No demographic similarities emerged in the data; there was no skewing toward age, gender or income. Influencers may be grandfathers or twenty-somethings, working mothers or stay-at-home dads. They could be the well to do or the up-and-coming."

What's at issue when you look at these data points is not whether the role of influencers is overplayed though that could be a knee-jerk takeaway. This data underscores it's crucial to understand the nature of influence, and only then will it be possible, as a marketer, to really influence the influencers.

Take the retweeting data above. What this should mean to a marketer is pretty straightforward: that people with true Twinfluence (that's Twitter influence, for the uninitiated) don't spend much time taking other content from the Twittersphere and sharing it. However, that data point shouldn't be extrapolated to the larger thought that they are lacking in influence and aren't worth a marketer's time. The data shows that 71 percent of @scobleizer's tweets in the first two months of the year contained an @sign. That says - and following him on Twitter bears this out -he is constantly conversing with other Twitter users. That certainly is influence; it just takes a different form than retweeting does. @stevecase's tweets contained @ signs 51 percent. Thus, his twinfluence expresses itself differently.

That loops back neatly into the ICOM data, which emphasizes how important behavior is in determining the nature of influencers, not for instance, the number of Facebook friends. Data-mining types will not like to hear the following: most influencers like to spread their influence via non-keyboard initiated word-of-mouth that can't be tracked using an algorithm. Compared to the average user, they don't even spend appreciably more time on Facebook than the average user does -- five hours per week as compared to 4.5 hours, though it, along with texting, appear to be their primary social channels.

I could throw out more data points all day long, but they would point in the same basic direction: toward the knowledge that although we seem to have gotten a pretty good bead on who influencers are, we need to know more about how they operate, in both online and offline channels if we are to truly harness their power.

(Editor's note: We're in the planning stages of June's OMMA Social NY. Send panel and speaker ideas to Put "OMMA Social" in the subject header if you can.)
12 comments about "How Much Do We Really Know About the Influencers?".
Check to receive email when comments are posted.
  1. Jonathan Graber from Three Pillars Recruiting, March 25, 2010 at 10:24 a.m.

    A company like Buzzlogic has figured out this problem - at least from a technology standpoint. They're still a start-up....the jury is still out on their viability in the marketplace. They have a secret sauce - a technology - that allows them to see who the influential blogs are. Advertisers can then cater towards them. Right now, they're acting as a niche ad-network in the blog space. They are a very cool company and something advertisers should look more into!

  2. Lindsey Alexander, March 25, 2010 at 10:30 a.m.

    This post truly highlights the main goal of social media, interactions. It makes sense that people with more @’s in their tweets are strong influencers. They are engaging. They are using Twitter to create conversations that lead to relationships. Bravo.

    I agree 100% that marketers need to be pretty straightforward on Twitter. Be real, your customers can spot a fake and they don’t want to be sold. Provide information that can help them reach their goals in order to build these relationships. Their retweets about your brand, products and services are far more valuable to potential new clients than anything you can Tweet from the corporate account.

    I’m very interesting in learning more from further research on Twitter influencers. There is not yet a success formula for this form of marketing and the strategies are drastically different from company to company. It can be difficult to explain these facts to professionals who are not avid tweeters and help them to understand the significance of these crucial interactions.

  3. George Eberstadt from TurnTo, March 25, 2010 at 11:05 a.m.

    Seems like most of the thinking about influence focuses on the influencer rather than the person being influenced. Try starting with the question: "who is most influential in the purchase decision of a particular prospect?" Then "how do I encourage that prospect to connect with that influencer?"

  4. Ben Straley from Meteor Solutions, March 25, 2010 at 1:27 p.m.

    Great topic and post, Catharine. Thank you!

    When we track and quantify influence, we do so with a specific definition in mind. That is, when a specific individual shares content/information, how many additional online visitors go to the source and how many additional end actions (eg. signups, downloads, orders) occur as a result? It's also important to note that when we use the term "share" we specifically mean passing along a link via email, Tweet, Facebook Wall post, Blog post, Message Board post, etc.

    When you track and measure sharing behavior in this way and account for the multiple generations of pass-along, what you see clearly is that there is a small (roughly 1% on average) number of individuals that account for 10, 15, 20, even 30% of total unique visitors to a site and that this small percentage oftentimes motivates an even higher percentage of end actions (eg. signups, downloads, orders). In light of this data and the insights it reveals about the power of a small percentage of one's audience to activate a large number of highly valuable users out there, the value of influentials becomes clear. It's huge and something our customers are keen on tracking and orienting their campaigns around.

  5. Norman Rose from Travel Tech Consulting, Inc., March 25, 2010 at 1:58 p.m.

    I believe that the shift to the term Social Media from Social Networking has caused us to lose site of the social graph and the need to understand the community and influencers I recommend everyone consider NodeXL an open source plug-in for Excel that allows you to map the social graph of Twitter, Facebook, et al. Here is a short video we put together to demonstrate how it can be applied to the airline industry:

  6. David Libby, March 25, 2010 at 2:11 p.m.

    I read every one of your articles and love them. Thank you again for such valuable insight.


  7. Johanna Skilling from NYU-SCPS, March 25, 2010 at 2:37 p.m.

    This stands out for me: "most influencers like to spread their influence via non-keyboard initiated word-of-mouth that can't be tracked using an algorithm." Thanks for highlighting that "influencing" is a human action -- helped by technology, but not defined by it.

  8. Brian Allman from Sweet Spot Golf LLC, March 25, 2010 at 4:58 p.m.

    Great information and thought provoking on how to best use this. I like George Eberstadt's comment about concentrating on the more desired end user and then figure who is best to speak to that very specific consumer....

  9. Dana Lancaster from iWin, March 25, 2010 at 7:09 p.m.

    I'm skeptical about trying to measure "influence" by retweets as a % of total tweets. The retweet feature is used so differently by so many people these days, what's to say is a true RT anymore? From a quick skim of Scoble's tweets, it seems he uses the native RT feature to actually quote a tweet, whereas a lot of people still use RT@ exclusively, or a mix of RT@ and native RT (depending on the client used). Some people use (via @name), as well. Are both types accounted for here? What about "via" tweets (common by users of Tweetie as that's how the "quote" feature works - though really via should be used just when referencing a link not a direct original tweet from someone else)? Are they considered RTs or @mentions in RWW's study?

    Here's where it gets interesting: "71 percent of @scobleizer's tweets in the first two months of the year contained an @sign. That says - and following him on Twitter bears this out -he is constantly conversing with other Twitter users."

    If you look at Scoble's tweets, he does use @mentions quite a bit. But it's not just to converse directly with users - it's mostly to reference people and brands when he is voicing an opinion or observation. Examples:

    Hmm, more video ads coming to blogs thanks to @embedster (they are enabling ads on links to YouTube videos). I bet someone uses those.

    Hmm @infoharmoni is able to display a map of where @gowalla and @twitter users are checking in, wish I had that at SXSW.

    There are a TON of VCs at @ycombinator demo days: <-fun to watch this "rich" list.

    So, in fact, his mentions actually are a form of influence. Maybe Scoble prefers to "RT" in a way where he references a tweet or idea by someone else but just phrases it in his own voice. Isn't that more influential than simply an RT @ or a native RT, the latter of which allows no commentary or opinion attached, which ultimately limits influence?

  10. Doug Pruden from Customer Experience Partners, March 26, 2010 at 10:36 a.m.

    As marketers it’s difficult not to be seduced by dreams of finding and winning over that handful of key bloggers and Tweeters who can propel our brand’s sales upward. But as has been noted here, "most influencers like to spread their influence via non-keyboard initiated word-of-mouth that can't be tracked using an algorithm."

    We might be better prepared to pursue all the opportunities if we think about what can be done through “Public Social Media” (Facebook, blogs, online reviews, YouTube postings, Tweets, etc.), but also how we can encourage advocacy through “Private Social Media” (text messages, emails, phone conversations, and yes even face-to-face conversations) - that thankfully can’t be tracked.

    The research I’ve read tells me that stronger consideration and greater buying influences are still driven by what we hear and read from people we know and trust. There’s nothing wrong with pursuing that handful of Public Social Media influencers who potentially will reach thousands of consumers, but we should also begin to cultivate the thousands of current customers who have experience with, and commitment to, the brand. There is a tremendous untapped opportunity to better prepare them to use their influence, perhaps through Public Social Media, but more often probably through Private Social Media.

  11. Gregory Yankelovich from Amplified Analytics Inc, March 27, 2010 at 4:52 p.m.

    @ Johathan - Just to keep the record accurate - the Buzzlogic is a subsidiary of Nielsen Research. There are many semantic analysis companies trying to mine Social Media content. Mine is included.
    To the point of the article - most of us are rather look for magic wand or silver bullet, being it "influencers" or "secret sauce" rather than try to learn as much as possible how our products are not meeting our customer expectations - and fix it. I suppose it is too simple.

  12. Walter Sabo from HitViews, April 7, 2010 at 9:38 a.m.

    In the equation of influencing the sale of a product, technology is cute at best. A person is influenced to buy by emotion and facts from a trust source. Not a trusted app. Online video webstars are powerful friends, sitting in their fans LAPS presenting ideas the recipient has requested to see. Wow.

Next story loading loading..