Is measuring real influence in social media possible?
One of the key discussions I continue to hear in the social media world is “how do we measure influence in social media?. Public relations companies, marketers are constantly seeking to pitch top influencers, in the hopes that those influencers will in fact help them increase their sales, etc. Jeremiah Owyang recently mentioned that number of followers is a bad measure.
Owyang asks for folks to look at quality and quantity of list memberships when measuring influence. The more time I spend analyzing social media (with an academic lens, even if I don’t actually study social media in my research), the more I think that measuring influence is next-to-impossible. Let me explain why. The graph to the left of the above paragraph is my Klout score of a few months ago (right now my Klout score has gone down to approximately 59, down from 75 which used to be my score, consistently).
For a while, I used to be in the top 5 Twitter Grader for Vancouver, even reaching #1 (that score also oscillated so much that it was almost useless). So when Klout came along, it seemed as though I had found a better measure of my own influence. Truth be told, in my own view, now that my Klout score is substantially lower, I believe that I have MUCH more influence, measured not by numbers, but by impact. I think my tweets have more impact now than ever before.
For example, recently, I asked the people who follow me on Twitter to re-tweet a job at UBC (I clearly marked my tweet as “please share widely”). I get one, two re-tweets here and there, maybe 6 on a good day, when I am not running contests that involve retweeting. But with this particular job, I got about a dozen or more retweets. That puzzled me. Is it the nature of the job, is it the fact that I have more influence, what made more people decide to retweet.
And the more time I spend trying to understand social networks (particularly for my research on ENGOs, but also transferring lessons to social media), the more I think influence is not associated with numbers. It’s associated with degree of trust. It’s associated with the right mixture of strong and weak ties. And please note that I didn’t create a causal link. I said that it’s associated (or even, has some degree of correlation).
In the next few weeks, I’ll be giving “tech-ie” talks (e.g. on blogging, on freelancing, on WordPress site building). But as I clear my desk from all the marking I still have to do, I want to start thinking more about issues of privacy, influence, safety and network modeling in social media. This post is the first of a series of thought entries. Obviously, it’s still half-baked, but it’s a start…
So, my question to you, my dear readers? What measures of influence do you take as most valuable? What constitutes influence, particularly as we think of social media?
Related posts:
- Does retweeting build community online? Twitter as a real social media platform?
- Measuring influence in social media (the Klout score)
- My recent Social Media Club Victoria talk: “Towards an Action-Focused Agenda for Social Change Using Social Media” (#smcvictoria) @smcvictoria
- Measuring influence Part I: Social media
- TUI – Texting Under the Influence
The overall Klout score is not a reliable indicator of anything except ego. However the underlying stats are useful. They’d be more useful if one knew their methodology but then you could game the system.
True socmed influence is subtle and diffuse. It is impossible to measure as it spreads out beyond the actual bits into real life connections off-line.
Ah the influence debate.
The problem with trying to measure it is that it’s incredibly subjective and there are a number of factors at play. I personally consider @ryananderson to be really influential but that’s because we’re old friends and I have a great deal of respect for him and his work.
Does that mean Ryan is influential in social media?
Furthermore, does it mean I’m going to retweet something he asks to be retweeted just because i find him influential?
You can measure reach, to some degree, but reach doesn’t equal influence.
FYI, there are ways to measure crossover but its much harder – so not impossible anyway.
Note that Joe points out importance of IRL – so true. It’s All About Trust
@ Paul – I disagree. Klout measures a lot of data that is valuable, it does NOT measure ego. It helps the ego of the people who are highly ranked as Personas (me, for example, one of them – not that I need anything to boost my ego!)
Also, please do note that I clearly said that it’s all about trust. I said it in my post, so not only Joe said it. I agree with what Joe said, that some other person may be much more influential to him because of what he and Ryan share (because they are good friends and Joe has a lot of respect for Ryan and his work).
@ Joe – I never said that this wasn’t a multifaceted issue. I said that I don’t think influence can be really measured.
@ both – I have spent a decade trying to study the influence of non-governmental organizations in international environmental negotiations (I can send you a few papers for you to read). I don’t think any quantitative or qualitative measure is enough. We need to measure many things.
What I meant in my post is – measuring influence goes well beyond numbers.
What comes to mind for me on this is that we need to define “influence.” Generally, “influence” means to get people to do something. So if you are interested in if person X can get people to repeat what person X has said, then the number of retweets is a relevant measure. But if you are interested in if person X can get people to take some other (probably more meaningful) action like buy a product or write a letter to their MP, then number of retweets isn’t a good measure at all (and it will be much harder to come up with a good measure for the latter).
Just my two cents (off the top of my head).
Interesting, I had to Google what influence means.
in·flu·ence [in-floo-uhns] Show IPA noun, verb,-enced, -enc·ing. –noun
1. the capacity or power of persons or things to be a compelling force on or produce effects on the actions, behavior, opinions, etc., of others: He used family influence to get the contract.
2. the action or process of producing effects on the actions, behavior, opinions, etc., of another or others: Her mother’s influence made her stay.
There are a few things in that definition that assists in answering the question. In the definition there is a quantifiable factor – actions (RT, replies) but to measure effects (behavior, opinions) is a completely different game. Bots and algorithm cannot fully capture that kind of transformation within a person thus to conclude that person influenced another person by their tweet or status update. Both quantitative and qualitative measurements have to be consider otherwise you have an incomplete picture.
As someone in the marketing industry, I measure the qualitative aspect of influence by figuring out who OFFLINE is able to quiet a room when he/she talks because it is much harder to be quiet and listen than it is to quickly chime in or dismiss the person altogether. If I am unable to be in the same geography, I figure out who’s paying to hear them talk :p
Personally though, this is how I feel about the topic of influence.
“You don’t have to be a “person of influence” to be influential. In fact, the most influential people in my life are probably not even aware of the things they’ve taught me.” – Scott Adams
My 1.5 cent, it’s 4:12am and I have NO IDEA what I am talking about.