In a recent article entitled, “The Future of E-Mail: Four New Marketing Segments You Need to Know About“, Jeanniey Mullen identified one of the new marketing segments as. “The Social Influencer”.  A follow-up article, “Social Influencers and BurntToast“, gives an example of one way that Wendy Ackland of BurntToast Marketing in Queensland, Australia was able to identify Social Influencers within an email campaigns.

As cool and thoughtful and well-reasoned the methods that Wendy used to identify the influencers of the campaign used as an example, that might not be efficient or effective on a larger scale.  Wendy’s example does support the concept of the Email Social Influencer, so let’s take a look on what it might take to identify and measure influencers across multiple email programs and campaigns….

If your email tool records the time/date stamp with each IP address that renders a tracking pixel along with any links clicked within a given message and associates that information to an email address – and you have access to it – then you’re just a few business rules and some queries away from identifying email influencers.  You’re not out of the game if your email tools aren’t this granular, but it may change how you effectively and effeciently  strategically manage influencer identification on a large scale at a measurable level.

Man, I really hope the FTC doesn’t come knocking on my door for this; hell, I already have a black car….

Okay, so now that we’ve established that we need some serious system logs to play hard and get granular, let’s look at how we might go about identifying social influencers within your subscriber base.

First we need to make some assumptions.  The first assumption is that anyone opening an email during business hours is probably opening their email at work.  Our next assumption is that anyone opening their email at work will probably be doing so from a static IP address. 

About here it’s good to know that there are two basic types of IP addresses, static and dynamic.  Without getting into all of the technical jargon of IP address structure, octets, and all that, an IP address is made up of four sets of numbers separated by dots.  With a static IP address, those numbers will always be the same.  With a dynamic IP address, the last set of numbers in the address will at some point change for each user, and in some instances both the third and fourth sets could change.

It may be kind of boring, but it’s kind of important to know for something like this….  It’s also a good idea to know if your email tool uses real record data when sent to you or your QA staff during a test.  Some tools will use data from and attribute actions to the first record on your distribution list when testing a message, and all activity you (or your QA staff or business owner/group or partner/client) make may inflate the behavioral data associated with that record. 

Since we’ll be looking at tracking pixels for this example we’ll want to identify records that have rendered a tracking pixel from the same static IP address during business hours.  We’ll also want to check that same email address for the rendering of any tracking pixels from any other IP address both within and outside of business hours, but we already have something to look at.

Let’s say that you send Bob an email and he opens the message at work.  Now let’s say that Bob’s email address shows 7 tracking pixels rendered to Bob’s email address all from the same IP address within 15 to 20 minutes of when the tracking pixel first rendered.  Do you really think that Bob “opened” your message 7 times within 20 minutes?  The assumption is that Bob forwarded the message to others within his company.  Even if he opened the message twice, probably not 5 more times within such a short amount of time….

The initial hypothesis may be that Bob is a Social Influencer at work and forwards your messages to others within his organization. 

So what about outside of work?  After all, we didn’t mention whether or not Bob received the email at his work or personal address….

Back to Bob.  Now we see that a tracking pixel rendered for Bob’s address at his work IP address.  (This would be even cooler if there were a way to tell how long Bob viewed the message or left it open in his preview pane while he went to a meeting.)   Anyway, we see a pixel render for Bob’s email address later in the same day – maybe after business hours or even a day or two later – from a different IP address. 

While not a static address we notice that the first three sets of numbers in Bob’s IP address are always the same (part of our query?).  Even though we haven’t identified Bob’s specific IP address at the exact time a pixel was rendered (you’d need a subpoena for that) we will assume in this instance that it is Bob and have a good idea what network he is on.  Bob’s a Verizon user.

So let’s put a stake in the ground and say that within 12 hours 15 more pixels render for Bob’s email address.  Because Bob is probably still on the same connection at the same time, we see that Bob rendered pixels to the image twice more after first rendering the pixel; that three other people on Verizon’s network rendered a pixel to Bob’s address shortly after Bob first rendered the pixel during his online session – one of them twice; and that nine other IP addresses rendered tracking pixels to Bob’s email address.  In the following 12 to 16 hours, we see spikes in pixels rendered to Bob’s email address that could also give us an idea of how viral the message may have become.

So what assumptions can we make.

  1. Bob didn’t get a chance to read the second email at work, or looked at it and saved it for home
  2. Bob shares information and may be a Social Influencer both in and out of work
  3. The people that Bob shares with share with others

Of course it helps when your emails have relevance to the recipient and are shareworthy.

In the end the value of Bob’s influence will eventually be determined based on any revenues that can be traced back to Bob. 

Some things like comparing invitations to registrations can be a Proof of Concept,  however it would be no small project to effectively and efficiently develop integrated lifecycle programs that target Social Influencers.  And we haven’t even discussed leveraging click-tracking and Web Analytics yet….

Email Social Influencers will come and go, and for this to be effective on a large or enterprise scale there must be the ability to dynamically segment recipients based on criteria set forth in business rules, filters and queries.  Don’t underestimate the Level of Effort that this type of project may take to fully and effectively implement.

Do you think that I’m being too granular, or do you think that looking at the number of pixels rendered or links clicked per email address is enough?  How many pixel renders or clicks attributed to an email address defines a Social Influencer to you?  Where do you draw the lines in defining your Social Influencers?