Influencer Measurement: What does it mean?

Over the past couple of weeks, I have noted an increasing fever pitch of discussion on Klout, Kred and PeerIndex - and how the concepts of measuring influence is both good and bad, the potential downfall of society, an enviable outcome of a more open, transparent community (great one Nick!) and so on.

Wired's article goes into the history of Klout and how it came about (Joe and his three-month speaking diaengagement), Kred's own story is borne out of their tool for social media analytics and monitoring and PeerIndex was borne out of an initial desire to find the best topic curator using crowdsourcing to provide the most interesting content being published at the time.

But the real question is: what are these companies doing?  What is the big deal about "influence marketing"?  Why do I get a score of 47 at Peerindex, 54 at Klout and 721(!!!) at Kred?

What We Measure is One Thing, What Our Clients Want Can Be Another

One of my favorite discussions in the past 15 months often consists of either an agency marketing person, or a brand director or other publication entrepreneurs working to understand what can "influence marketing" do for them?  In a previous post, I describe what I believe to be the rebranding of word-of-mouth marketing to influence marketing.  The difference is that in the new era of social networks, APIs for "exhaust data" and Big Data systems to process all of this data into actionable information brings about a new belief that we can "measure" word-of-mouth.  I would argue, like Nathan Gilliat did, that there is no such thing as a "unit of influence".  All of us - Klout, PeerIndex and Kred - are building models that become our own "truth" for determining actions.

And yes, these are models - not specifically measurement on influence if that was possible.  We use models to extrapolate what is happening - and all models have biases that are built around hypotheses.  As Nathan says:

Models reflect the opinion of the modeler and the objectives they support. Because apparently simple concepts might be used for different purposes by different specialists, we end up with diverse models using the same labels. In essence, we talk about the labels, because they represent familiar ideas (influence, et al), but the models represent what we really care about (such as positive word of mouth, leads, and sales).   

If you understand that the label is just a convenient shorthand for a model that takes too many words to describe in conversation, it's not a problem. If the model generates useful information, it's doing its job. Just don't assume that any one usage of the label is the correct usage. Modeling requires judgment, interpretation, and prioritization in context, which are incompatible with standardization.

So, if you understand that the models are our shorthand for creating meaning out of measurement, the real question is, what are our clients looking for?  And to be clear, by clients usually mean the ones who pay for our services.

Agencies and Brands are suffering.  They are awash in data and dashboards and monitoring and listening and so on - trying to get a grip on this world we call social media marketing.  Years ago, when I spoke on politics and social media , the ideas of "social media" was scoffed at.  Today, it will be the "biggest" area of growth in all businesses.  Shades of the Cluetrain Manifesto, anyone? 

So, when the clients are asking for help, what are they doing?  The same thing they have been asking for years on end - how do I get the message out to the right people to move the needle on my goal?  The problem is - what do you need to measure to be able to determine return-on-investment (ROI)?  And that is where the problem and the puzzle exists.

Politics in a World of Social Data - who knows what about whom?

This past month, I have been reading Eli Parser's "The Filter Bubble" on my iPad Kindle app - and I must say, Eli has done a bang up job of discussing the impact the Filter Bubble can and will have on our political and intellectual discourse.

As I continue to work on the problem of "influence marketing", it brings back concepts of control theory to mind, and how tuning the inputs to a system can probabilistically ensure a bounds on response.  And, as you begin to learn the "modes" of the "system", you are able to generate more desirable performance from the "system" overall.

Eli uses the phrase "persuasion styles" (from Dean Eckles at Stanford and Facebook) which discuss the modes on how a person can be influenced to perform an act/action based on how the request is made, not simply by the content.  And with so much Social Action data around on the web (and purchasable in large batches from companies like GNIP and DataSift), the ability to sift through the voting records, demographic records and Social Action records will be a powerful mix.

Political Influencer Marketing

So what is "influencer marketing"?  As I described in an earlier post, some circles consider the measure of influence corresponds to your perceived expertise in a particular topics or arena.  It harkens back to the old E.F. Hutton commercial, "When E.F. Hutton talks, people listen."  But in today's world of social media focused on driving action - and advertising dollars being spent to drive purchases or brand impressions, influence marketing becomes more of a word-of-mouth advocacy concept - who is able to drive actions from whom by creating reciprocal or responsive actions to the original action.

In Control Theory, we would call the original Social Action an "impulse" and then see what "modes" get "excited" (e.g., vibrate, resonate).  This processes is known as System Identification.  And while most "systems" are considered from a single-input/single-output (SISO) model, we know that we are bombarded by "impulses" from all sorts of sources (e.g., TV, adverts, Facebook shares, retweets, LinkedIn updates).

Influence, Authority or Word-of-Mouth?

For the past week, I have been watching a bunch of folks discuss the changes to Klout's algorithm and have been amused at the concerns people have had.  I should tell you that my Klout score has dropped from an average of 57 to an average of 49 - with variations in between.

And whilst I work as the CTO of PeerIndex (the competitive offering to Klout), I am bemused by the hub-bub.

I hand it to Joe and the Klout team for building an effective business model for a Perks-based Groupon.  In my estimation, the business model goes something like this:

• Create a ranking mechanism based on actions within the social web - primarily in twitter, but include other signals from LinkedIn, Facebook, etc.  Define influence as "causing an action in others in your network" and link it primarily to tweets, RTs, and shares.
• Create a Perks program that requires others to have to compare their Klout rank to an arbitrary threshold (one chosen by Klout with a partner) to suggest that if you have a rank higher, you are an "Influencer" and if lower, you are not influential enough - and target the social media engagers.
• People who get the Perk are thrilled and may be happy enough to tweet, blog or post about the Perk - with Klout branding and using the influencer's network.
• People who don't get the Perk are upset and comment on how it is not an accurate measure of their "influence" and either complain on the blogs, feeds or whatnot or track their score to see how to improve it - or both.
• More Perks means more chances to get freebies - and more chances for social media people to react - either with glee or jealousy.
• Klout ensures it has the channel to the individual social media influencers (by requiring the email address initially) and then triggers when a Perk is possibly available to them.

Resulting in:

1. A private Klout-engaged email mailing list for potential "influencers"
2. Brands paying for Klout's marketing expenses when Klout takes care of the logistics and fulfillment
3. A growing desire and ownership of the Klout score (e.g., "If I get my score higher, I will get Perks!")

Now, with their increasing growth, Klout has now removed the requirement to have their branding all over the packaging (thanks Steve Garfield) and even I got a Perk from one of the campaigns.

But I liked Robert Scoble's discussion from last week over what is Klout selling?  As in, is it being poorly pitched (click the link for the audio)?

Expertise versus Causes Action

When I first joined PeerIndex, I spent an evening with a gentleman who wanted to discuss with me the concerns he had about ranking individuals by their social media actions.  "How could you estimate my influence on any topic?  You know nothing about my education, my work, my writings, nothing."

And he is right.  And neither Klout nor PeerIndex suggests that.  Heck, see Klout's newest popup when you join - it clearly says "a higher score means you are driving more online action."

At PeerIndex, our PI Score you see in the Yellow Box has a similar metric - designed to suggest people who are great at driving online action - not that they are more famous or influential than anyone else.  The difference in in the Topic PeerIndex scores.

As a proxy for expertise, we use your engagement in a particular topic to identify your areas of expertise - even if you do not have a degree in that area.  If you are engaged on a topic like "Peer Influence", and you are effective at spreading a message in that community, then we rank you higher than others.  If you are engaging within that topic community - have others chat with you, share your link, comment on your blog posts, then we rank you higher.  We have been focusing on this segmentation because, in our opinion, helping brands and agencies connect with effective word-of-mouth advocates is not about popularity, but about engagement within their peers. 

Word of Mouth Marketing (circa 2005) is Influence Marketing (circa 2011)

What I love about this market today is that we are discussing the concepts that have been requested for time immemorial.  But unlike 2005, where we would spend all sorts of time trying to figure out blogger outreach and viral videos, we now have twitter's API, bit.ly URL shorteners with analytics. Facebook social graph and the Big Data tools to engage with the torrent of data that comes from this.

Google has had the corner on the Big Data market for some time for Google AdWords and AdSense, and Facebook is ramping up fast to use its data sources in a similar fashion.  And with Hadoop, other companies like ourselves and Klout are crunching through data sets that would have seem impossible or prohibitively costly to manage.

Today, Klout, Kred and PeerIndex are providing metrics for scoring yourself on how effective you are on word-of-mouth marketing - as measured on the social web.  Kred is going one step further in allowing you to provide offline influence credentials to add to your Kred score (yes, send your degree to them and you get +1000 Kred points!).  But the question that I ask - what are YOU looking for?

You - as in the owner of this exhaust data that is out there in the world?  Or you as the brand manager who is looking to improve your engagement with your advocates or simply your customers.  What is it you would like to see in this world of Influence / Engagement Marketing?

Kred Presents at PIVOT

So, last week, I spent a short amount of time at the Crown Plaza while Brian Solis and gang put on the PIVOT Conference. As the CTO of PeerIndex, a provider in the Influencer Marketing space, I was quite interested in hearing about what others were talking about in the influencer space. To that end, I got to spend some quality time with the founder of Kred, a new player in the space that has been focusing on developing a product to compete with the likes of Klout and PeerIndex.

From Jodee's presentation, Kred is made up of two major metrics - Influence and Outreach. The definitions of these metrics are:

• Influence is the "ability to inspire action" and the measure of action is based on other people retweeting, responding (@ messaging), liking and following you.
• Outreach is the "generosity that you give in responding to others" in the same areas of measure (retweets, responses and other acts of social kindness).

Kred breaks down the score into two segmentations: Global and Community.  In the Global measurements, you are given a score by the number of responses to you during the past (which goes back to up to three years) and each social action is weighted by the what particular action occurred.  So, someone retweets you, that person gets Outreach points, while you get Influence points.

The Global Influence Score is then applied to the entire listing of the Kred database and everyone is ranked from 1 to 1000.  Thus, people's Influence Kred overall is measured against everyone else.

In the Kred Community space, the population of each Kred Community is defined primarily by what their bio says - in particular, what titles, what words are used and so on.  There are currently 200 Kred Communities and will grow.  But in the case of Influence, all members of the particular Kred Community will be measured against each other and then ranked with a normalized score (from 1 to 1000).  In addition, the Kred Community is also ranked as an entity unto itself - allowing for the KC to have its own Influence and Outreach Scores, in comparison to the other KCs.

Outreach is not a normalized score, rather it is an accumulation of actions that a person does to share in the contribution of social interactions.  Right now, the scores are in the naughty digits (03, 04, 08) but should grow with each act of social kindness a person shares (oh, I look forward to seeing how @jetblue fares here).

Future Steps
From Jodee's conversion, we can take away that:

• Right now, only providing twitter from their past data - in the coming month, will support Facebook and LinkedIn.
• Next will be how to add blogs to the mix and how to map the bloggers to their twitter account. 
• People can generate "offline Kred" by submitting documentation to actions or credibility that can be confirmed in some fashion (e.g. a Presidency gives a person +1000 Kred points)
• At present, Kred does not rely on the twitter favoriting feature for affecting the Kred score

As for privacy, Jodee and team are very proud of the four privacy settings for customers on Kred (see below). The only concern I see is there is no way for a person to absolutely remove their data from the Kred database which might be an issue in Europe.

I, for one, welcome their entry into the play.  They offer a different take on the model and seem to have a lot more visible data that they can show people such that others can understand why they have the Kred they have.

If you wish to see the raw footage from this event, feel free to catch up at http://bit.ly/kredatpivot.

[Published on behalf of CenterNetworks]