Influence Measurement Optimization™ 2 – Rise of the Mathematicians

This is a follow up to my recent post Influence Measurement Optimization™. There was a lot of discussion surrounding the initial post, so I thought I’d expand on it a bit (with a little nudge from Jonny Bentwood of Edelman. Thanks, Jonny).

You Can’t Measure Influence

There. I said it. At least not empirically. There are too many variables and permutations. But you can measure things.  And there is value in measuring things.

In business, we try to measure lots of things. Some things we try to measure are pretty clear – how many people does it take to make how many widgets per hour?  Others are not so clear – what is the impact of this marketing campaign in terms of customer awareness and traction?  It’s here, where there is the least certainty, that we need to apply the most creativity. One way is to attempt to simplify the complex to a point where it provides some value.

So why all the talk of measuring influence? There are certainly some very complex issues in the world and, as disruptive technologies, the Internet and social media in particular seem to be playing a large role. So efforts to quantify the “unquantifiable”, particularly as they relate to influence within social media communities, are worthy of further scrutiny.

I love it when a complex idea can be reduced to a simple, 5th grade math equation.

That’s what Fred McClimans of the McClimans Group said to me in a recent discussion on influence in social media. He joked about the efforts of companies like Klout, TweetLevel, PeerIndex, Twitter Grader, and Twitalyzer to quantify influence, something with which Fred is well acquainted. His recent post “5 Trends Shaping Business Influence Today” sheds some light on the depth and complexity of the issues we are dealing with today. Based on many discussions we’ve had, I know what Fred meant: it’s all about context and scope.

Simple 5th grade math? They don’t teach math the same way they did when I was a kid. I love the way they teach it now because it aligns with how I naturally look at problems. While helping my kids with their math homework, I was pleased to see that they didn’t just teach rote memorization. They also taught strategy.

Can’t figure out that complex multiplication problem? Break it down into smaller problems that you can solve.

So What CAN We Measure?

Marketing and public relations are basically about influence. We want to measure how we can change someone’s perception of a product, service, political candidate, policy – the list goes on. That’s an awfully tough nut to crack. So how about some lessons from grade-school math: take influence and break it down into data points that can be measured.

The data-driven nature of social media is ripe for harvesting metrics. You can’t really tell how many people I’ve verbally told to see a movie, for example. It’s too subjective. I could lie. Or, more likely, just give an incorrect answer. So it’s not valid data.

But what if I tweet about it? You can measure my tweets and you can measure the number of my followers that were actually online when I tweeted. You can track the replies and retweets. And not just one level deep (my direct “followers”), you can track subsequent re-retweets, mentions and conversations. Will that tell you how influential I am about movies? Simple answer: yes – within a very specific scope and context. These metrics may not be earthshattering, but they have value. They are a step towards quantifying what was heretofore “unquantifiable.”

Not Quite 5th Grade Math

I had a chance to chat with Jonny Bentwood, head of Analyst and Influential Engagement for Edelman, about the new release of their influence measurement tool,  TweetLevel. Does TweetLevel empirically measure influence? No. But it has an interesting take on measuring your impact within Twitter. And, in an act of transparency which I admire, they even tell you how they do it.

Here’s how they calculate Twitter influence:

Tweet Level_Algorithm_01

You can see a detailed explanation on the TweetLevel About Us page. A bit more than a “5th grade math equation,” but simple and understandable. They also go beyond just boiling individual twitter handles down to numbers. You can also search on a topic (ex. “influence”) or even a hashtag (ex #influence).

Are these earth shattering metrics and analytics? No. But, within the narrow context in which they are defined, they have value. For TweetLevel it’s Twitter. Klout, for example, has expanded to also include other social networks like LinkedIn, Facebook and FourSquare.

Games People Play

I coined the term Influence Measurement Optimization™ (IMO) in my original post to foreshadow what I see as an analog to today’s search engine optimization (SEO). The goal of SEO is to make your web content more “friendly and inviting” for search engines, in hopes you will be put at the top of search results. The goal of IMO is “gaming” the influence measurement systems to enhance you influence “score.”

According to Jonny Bentwood, Edelman has no plans to “monetize” their measurement tool. It is simply a nice calling card for their ingenuity. By publishing their formula, they are showing transparency, a trait highly valued on the Internet and particularly in Social Media. But they’ve also opened themselves up to IMO. If you know the formula you can attempt to manipulate it. In contrast, Klout has a proprietary algorithm and has attempted to directly monetize their influence measurement. Since the algorithm is unknown, gaming is more difficult. I foresee an “arms race” similar to search engines vs. SEO or virus creators vs. the anti-virus vendors. Klout will have to continually adapt their algorithm as people figure out how to game the system.

So what does this mean?

I wrote my initial post because I saw a trend. Understanding trends takes away some of their power to overwhelm us. Is Edelman’s approach right? Is Klout’s? They are almost two different animals. But one thing is certain. Anyone who goes into one of these influence measurement sites is happy if their “number” is high and disappointed if it is low. How much would you pay to get it raised? How long before there is a market for Influence Measurement Optimization specialists?  Time will tell, but consider this:

  1. Like it or not, online influence is important to many people;
  2. With the identification of any metric comes the desire (and ability) to manipulate it

This is just the beginning. Do you agree? Where do you see this going?

  • David G H Phillips

    I tend to join Alice in her rabbit hole when these discussions take place.

    Let us for an instant attempt to examine what the real science is trying to tell a largely deaf profession.

    The convergence of values based on psychology, neuro psychology, anthropology, among a wide range of human behavioural sciences tell us that humans have values and that influence (for good and ill) is about the convergence of values.

    Does you client really want influence.

    Like Muammar Muhammad al-Gaddafi?

    The whole idea is much deeper than Klout, TweetLevel, PeerIndex, Twitter Grader, and Twitalyzer or even the Edelman formula.

    These are measures of transient channels at best and do not reflect the ability to change nature that was achieved by Gaddafi (and a history book of equally nasty clients looking for’influence’).

    Influence is a nice, shallow, marketing number.

    It is not fit for PR purpose.

    PR is much more subtle that that. PR can help people warm to, adopt and evangelise ideas without shouting or shooting.

    • “Influence is a nice, shallow, marketing number” — that hits the nail right on the head. Thanks for adding to the discussion, David. 

  • Hi Alan,

    Why focus on measuring social media influence at all?

    Wouldn’t you prefer metrics that tell how well marketing is driving prospects thru the marketing funnel all the way through to sales …

    social media influence impacts a small part of that process, but by itself influence doesn’t have much value

    metrics that help to measure and optimize end to end marketing effectiveness — now that would really get my attention

    • I agree completely, Rhonda. I wrote the initial post (Influence Measurement Optimization™) as mostly tongue-in-cheek. Is the trend good? Time will tell. We’ve relied on Nielsen ratings and FICO scores. Why not influence measurement? 

      Metrics and analysis are the hallmark of the burgeoning information economy.  Accuracy and importance are not necessarily exclusive, so regardless of whether you believe in it or agree with it, influence measurement is not going away any time soon.