Neicole Crepeau is a partner with Coherent Interactive, a digital shop specializing in all things web, social media, and digital, of course. Aaron Sachs is working on his Masters degree at the University of Tennessee-Knoxville, where I’m certain his coursework includes many live discussions about measurement.
Neicole kicks off:
I believe there are several factors that contribute to defining influence:
- Topic—nobody has influence in every area, most people have influence in specific areas
- Community—most influencers are active in certain communities or have the greatest impact in certain communities
- Activity—some influencers are content creators, others are mainly speakers, etc.
- Type—There are different types of influence. For example, there are people who are very good at spreading the word/content, and might be valuable for content marketing purposes or to gain traffic to your website. There are people who are very influential in terms of shaping opinions—think Walt Mossberg in tech. A recommendation from him on a tech topic carries a lot of weight. There are people who are influential in their ability to start conversations and connect with individuals. I think you and Mark Schaefer fall into that category, Jayme. (Thank you, Neicole!) There are probably other types, as well, thought-leaders, etc.
From a business perspective, you’d want to connect with different influencers depending upon your goals and audience. You want influencers who are in the communities where your audience is and who are influential in the ways that will help with your goals. If I want to get someone to be a speaker at my conferences or do webinars for me, I probably want to connect with an influencer in the relationship-builder model. On the other hand, if I want my content shared widely to people interested in a certain topic, I probably want to connect with a content curator, as per my post on Mark Schaefer’s blog. And so on.
Is there an effective way to measure influence? I don’t think so, not yet. Currently, systems are using proxies of real measurement. We use RTs and reshares to measure how good an influencer is at spreading the word. But we can’t really accurately measure that. If I share a link, someone clicks on it and reads the blog post I linked to, and then clicks the Retweet button on the blogger’s site, we’ve lost the trail. Someone may click the link that the reader shared, and that click should actually be attributed to me, but we can’t track it.
Klout measure shares and @’s and DM’s and comments on Facebook, too. That measure’s engagement of a sort, but from the number of interactions, Klout makes assumptions about someone’s influence. If the influence type is Walt Mossberg style, @’s and DM’s may not matter at all. What matters is how well-regarded the influencer’s opinion is and how well-spread his/her opinions are to a key audience, not whether he/she actually talks a lot with fans.
To really measure, we’d have to have a good breakdown of the different types of influencers, by topic and community, use different measures for each community, and measure all the way through. We’d need to know things like what audiences/customers the influencer is reaching, how far down the sales funnel it’s happening, who is sharing the opinion/information with customers, and whether it actually resulted in conversions (however you measure conversions: sales, registrations, etc.). We aren’t anywhere near that, yet.
Influencer identification and measurement is still in its infancy. We’ve got a long ways to go.
Aaron Sachs has some great
I want to begin with the tools that are out there–Klout, Backtype, PeerIndex. They all measure influence, but there’s a key thing missing, before we get to what the tools measure, and that is: Who is defining influence for these tools and what are they basing it on?
It’s all well and good to have an indicator based on partial metrics–Tweets, retweets, followers/following ratios, etc. But where is influence in all of this?
Michael Wu at Lithium has an excellent post on influence. In an interview with him, Michael stated that simply, influence is the ability to cause a change in mindset or actions. Essentially, influence is the art of persuasion. It is causing someone to think or behave a certain way. Whether the motives of the person influencing another person are “pure” or not is irrelevant to defining influence. However, we all hope that people influencing any field are pure of heart and have the best interests of an industry or field at heart and won’t influence the direction of an industry or field in a negative fashion. But that’s beside the point.
The tools that are being used to measure influence are broken. They don’t measure true influence. RAAK did several tests looking at PeerIndex and Klout. The key test was Klout, as that seems to be the standard right now. They found that the bot that tweeted the most had a higher score than the other three bots. By conducting that test, RAAK essentially found out that Klout’s definition of influence (and somewhat the industry standard, by proxy of accepting Klout as the defining measure of influence in the social media world) was relegated to Tweeting…a lot. Is that really how we want to define influence for social media? For that matter, do we want to define influence as simply retweeting, interacting with others, or posting content? I would hope not.
For social media, there has to be a CONCLUSIVE measure of influence. If the definition of influence is to cause a change of mind or behavior, then THAT is what should be measured. Simply measuring retweets or interaction does NOT prove conclusively that a person has experienced a change of mind or behavior. Sorry 🙁 long rant about the tools.
The other factors that are missing in influence are things like credibility, relevance of material to audience (or relevance in general), physical appearance, communication skills, and persuasive ability. These items are the major items that contribute to influence.
Then, you take into account the things that have contributed toward a person being influential–professional opportunity (essentially status–we’re more likely to believe someone who is in a position that lends them more influence or credibility), economic opportunity, and access to tools/communities.
Retweets, tweets, and engagement are not a conclusive measure of influence, not if it is defined as causing a change of mind or behavior. Really, the approach that the social media world should be taking to Klout and other influence measuring tools should be a mindset that understands that these tools don’t really measure influence. If anything, they measure how engaged you are with your audience. Engagement (does not =) Influence.
(Photo: NextWeb…never thought I’d see the day when Old Spice Man was relevant for an image; very cool.)