Friday, September 9, 2011

Klout analysis of PainPoint's 4,000 Twitter followers

So I'm currently on a quest to "better know my followers" so I recently turned some attention to

It appears that Klout's primary service goal is to provide a "score" to your social media accounts based on your Twitter and/or Facebook "Network, Amplification and True Reach."   Their algorithm is proprietary, but includes the usual counting of followers, retweets and of course other secret sauce relevancy engine stuff.  Klout, PeerIndex and EmpireAvenue, are a few examples of companies that are doing very cool things in the social network relevancy space.

So now as a self proclaimed "new media entrepreneur," I guess one of my goals should be to maximize my Klout score and eventually get some of those Google Ad dollars or sell virtual goods myself. Apparently I was already high enough influencer last month that their partners over at MetroMint sent me a free box of water!  Free water doesn't pay the rent, but one of my personal health "pain points" is that I generally don't drink enough water, so it was a welcome promotion. 

As of today my Klout score is a 58, up a few points over the last week.  Awesome!  Naturally I had ask myself "is that good or bad?"  How does my Klout score stack up against my Twitter friends and followers?  Will analysis of my follower Klout scores give me some more insight into my followers? A few simple CSV exports and a couple Excel manipulations produced the following distribution summary:

Distribution of @PainPoint Followers Klout scores (X-Axis: 1=1 to19, 2=20 to 29, 3=30 to 39)
In other words, about 80% of my Twitter followers have Klout scores between 20-59), with a mostly even distribution across the 20s,30s,40s. You are in a slightly more exclusive club with a score in the 50s. A score in the 60s puts you in the top 9% and 4% of my followers have scores in the 70s. Congratulations to @Kim and @TheNextWeb as you are part of the top 1% of my followers with Klout scores in the 80s. Virtual hugs!  Unfortunately users with scores in the 90s like @justinbieber @ladygaga, @mashable, or @barackobama are not yet following me.

Same graphic as above as a pie chart.  mmmm pie!

I would agree that users with higher Klout scores are at least generally more likely to be offering a unique personality, quality content and/or are well engineered Twitter content robots or "celebrity produced" accounts. In any case, Klout would likely argue that media content produced by these "nodes" with higher Klout is more widely distributed due to the person's "Network, Amplification and True Reach."

Distribution of @PainPoint Followers based on their Klout Scores. 
A quick look below at some of my "big name" followers with high Klout scores...Wow - I'm truly humbled.  Thank you all for following! I wonder if you or your others in your organization are influenced by any of their messages?  Perhaps some of the tweets I've written over the years have been influenced by them? Do you find yourself more likely to retweet the following users or consume the original content they offer?

Forrester -
Gartner_Inc -
CIOonline -
Zappos -
NetworkWorld -
CiscoLive -
Cisco_Channels -
MarketingProfs -
Kim Sherrell -
R "Ray Wang -
Laura Fitton -
KXAN News in Austin -
ComcastMelissa -!/comcastmelissa
VerizonSupport -!/VerizonSupport
dups - CEO of Empire Avenue (mentioned earlier)

I'm sure I missed many of you who have a high Klout score. If you are now a follower of @PainPoint on Twitter, have a high Klout score, and are not on the list, please comment below and I may update the list. Remember my goal of this exercise was to better understand my followers to help guide my future content. Now I just hope some of them retweet me. :)

Have a great weekend!
-Andy Fields
CEO Vendor Voice Media, LLC
*Call to Action 1: Please help me better understand my followers (Survey):
*Call to Action 2: Drink more water!
*Call to Action 3: ReTweet this Blog Post and be part of my top 5%

Finally, here's my Klout page including the topics they think I have expertise.


  1. Interesting spike at Klout score = 1. It appears k=1 is the default score given to relatively inactive or new accounts. Also interesting to note that Klout offers no scores between 2-10, so they must have some initialization threshold in their algorithm that chooses whether your account is a 1 or at the true beginning of their variable scale at k=10.

  2. I follow you and am 81/82 depending on the day. I seem to be fluctuating between the two. Actually the score is only a tiny part of what is about and can do for you. Try this

  3. That 1 to 10 jump is really interesting to know.

    For what it's worth, the Klout product manager recently said that 20 is the average Klout score.

  4. 10/26/2011 - UPDATE from Klout:

    This will affect all scores and may invalidate some of the findings in the article above. Once I find some time I will re-do the analysis.


    After the algorithm changes, I dropped from 60, and have now dropped just below 50 because of the Google+ scoring bugs (my top network), but otherwise it looks like I'd qualify for your list. I'm (e)MetaThis on Empire Avenue.

  6. Updated Distribution Graphic: