Is Online Influence What You’re Really Looking For?

August 17th, 2011 by derek

 

Who is influential?  THAT is the question.  At least that is the question that TRAACKR was built to answer.

Or was it?

While this may be the most common question asked by marketing & PR pros when thinking about important people on the social web, we realized early on that this was not really the question everyone wanted the answer to (nor was it the most interesting question to answer, from our perspective).

We quickly realized the most important question to answer was not Who is influential?, but instead,

Who is going to be influential TOMORROW.

Or more specifically:

Who has the greatest POTENTIAL of being influential tomorrow.

And while this may seem like just a subtle re-wording of the original question, it has significant ramifications for how we all think about and approach this concept of “influence.”  It’s also (one of) the things that make this whole business of measuring online influence so tricky — because if done right, it must have a built-in predictive element.

What’s really important is what action one has the POTENTIAL to drive in the FUTURE.

This is what marketers ultimately care about.  What if I could uncover for you the most influential person in your market?  A person who has driven traffic, leads and deals to every business he blogs about.  Would you be interested?  Sure you would.  You’d be salivating for that info.  Then, what if I told you that yesterday this person decided to shut down all his social media channels and become a Belgian monk, effective immediately.  Would you care about him then?  Of course not (unless you were into some crazy good beer).  While this person’s past online influence was enormous, his potential influence is now zero.  And, when it comes right down to it…potential influence is all you care about.

THE FACTORS OF POTENTIAL INFLUENCE

Of course, the challenge with trying to hone in on a person’s potential influence is that data is, by it’s nature, historic.  All the stats that you can use to ‘calculate’ influence are based on past behavior.  This requires you to get creative with how you use this data so to make it more predictive.  So, when we think about how to gauge someone’s POTENTIAL to be influential in the future, we have to look at a few major factors.  To us, for a person to be influential in the future he/she must:

  • Have an engaged audience (the size of the audience is not necessarily very important here – more on this later)
  • Be active online — recently.  Understanding the trend of one’s activity is helpful in predicting potential influence.  If someone hasn’t published anything online in a while, the likelihood that they will publish something in the future is lower…lowering his/her potential influence.
  • Be interested in and willing to engage in whatever it is that you are trying to promote
  • Have an audience that is responsive to the particular topic that matters to you

If these things are all true, then a person has a high potential, future influence for a particular campaign.  Which means they have a higher likelihood of driving “action” in the future.

THE CONVERSION RATE

From a mathematical perspective, one’s potential influence is simply the raw number of actions that he/she can drive.  How many people he/she can get to do something.  And there is a very simple formula for this:

# of people influenced = (size of one’s audience) X (% of that audience that can be driven to action)

The first factor in this formula (Audience Size) is fairly simple.  It’s not always as straight forward to accurately determine as it sounds, but at least it’s relatively quantifiable & stable (audience size doesn’t vary widely over short periods of time).

The second piece of the equation (% of audience) is the really important piece — call it a Conversion Rate.   It defines the % of an audience that can be driven to action. The % of an audience that can be “converted.”  This Conversion Rate is affected by the factors described above.  The more engaged one’s audience is, the higher the Conversion Rate.  The more interested and passionate the influencer is with the topic in question, the more likely he/she is to passionately promote the topic, increasing the Conversion Rate.  The more relevant and responsive an audience is to a particular topic, the higher the Conversion Rate.

The Conversion Rate is how a person with a smaller audience can drive as much action as one with a bigger audience.  A person with an audience of 50,000 people who can convert 1% of that audience will drive the same amount of raw action as someone with an audience of 5,000 people and a conversion rate of 10%.

  • 50,000 x 1% = 500 people
  • 5,000 x 10% = 500 people

This is why Audience Size does not equal influence.  And it certainly doesn’t equal potential influence.  Don’t get me wrong…it’s a factor, but all the elements leading to that conversion rate are what really determine one’s potential influence.

HOW WE HAVE APPLIED THIS THINKING

This goal of predicting potential influence has been “baked into” what we do in a few ways:

  • We obviously take one’s audience size (Reach) into account when we assess someone’s influence…but we weigh audience engagement (we call it Resonance) at the same rate as we do Reach.  Higher engagement leads to higher Conversion Rates.
  • We have a scale which weighs more recent content more heavily in our Relevance algorithm (more on this in the next point).  The attempt here is to get a sense for a person’s publishing “velocity” and use that as a predictor of their likelihood of publishing in the near future.
  • Most importantly, we have always focused VERY heavily on a person’s Relevance to a topic in order to determine his/her potential influence.  To us, this is the greatest predictor of one’s ability to “convert” an audience.  The more Relevant a person is to a particular topic (a) the more likely they are to write about that topic; and (b) the more likely their audience is to respond with action.  Relevance is the biggest factor in one’s Conversion Rate.  In our influencer scoring engine, Relevance is the single most important factor in determining one’s rank on any search.

These things are all things we consider strong predictors of future influence, or Potential Influence.  And these are the steps we take to be more predictive in our analysis/algorithm.  Our goal is not to produce lists of “influencers,” but instead lists of people who have the highest Potential Influence for your particular search.

And it’s still very early in the game.  We are continuously improving on our system and scoring engine.  And as we do, we are constantly seeking that mystical beast which is a highly predictive model of a person’s POTENTIAL influence.

Thoughts?

DS

ps — and not for nothing…one element of one’s Potential Influence that we will never be able to measure is the quality of the engagement. The stronger and smarter your engagement with influencers is, the higher the Conversion Rate will be.  Your work definitely has an impact on that equation.  Don’t overlook this point and take “influence” for granted.  It’s not guaranteed.

Share

Tags: , , , ,

  • http://fredmcclimans.com Fred McClimans

    Derek – Excellent piece about the value of “potential” influence. You are correct in that understanding what influence a person is likely to have in the future is very important – perhaps more important than the influence that person has in the present or has had in the past. And while past influence is a good thing to understand, it doesn’t necessarily translate into long-term “actionable” influence. Musicians are a great example here, where at their peak they have true “actionable” influence. Over the course of their career (and even after) they may influence a new generation of musicians. But how many people hop online to download the latest release of a band that peaked 40 years ago?

    Influence is definitely time-sensitive.  Perhaps an interesting exercise would be to identify individuals with relatively low influence (they don’t measure up YET to your “Factors of Potential Influence”), but due to their present online (or offline) trajectory have a strong likelihood to meet those factors in the future and thus the potential to become influential as they grow/mature.

    Good post. – Fred

  • http://www.traackr.com Derek Skaletsky

    Thanks Fred (and good to ‘see’ you again).  You bring up a great point about uncovering the “rising stars.”  We are looking at ways to do a better job at this.  Part of it is theoretically simple — increasing the “weight” of “velocity” in a overall influencer score/ranking in order to emphasize the “potential.”  Obviously, in practice it’s a bit more challenging — especially with the rapid movements of the social web.  But we’re definitely working on it…

    Finding tomorrow’s rock stars.  That’s the next big challenge…

  • http://blog.intelligistgroup.com Alan Berkson

    Derek,
    Do you factor in “momentum?” There’s a tendency for influence “scores” to drop during weekends or vacations. How well they sustain could be an indication of future rock star. There’s a fade curve on influence. What’s the slope of that fade curve? 

    -Alan