My Klout Experiment And The Disturbing Results

In my professional, humble social media opinion, it is about value, engagement and relationships that lead to ROI, not what someone or something says about how influential you are. Having said that, many of you that read my blogs on social media marketing and Klout know that I have been fairly supportive of the company, it’s business model and their perceived desire to provide a reliable, yet imperfect overall influence score.

Fast forward to just after the recent substantial changes Klout introduced a week ago yesterday and the firestorm of unrest that ensued, I wrote a post that supported the changes and tried to put everything in perspective.  In my article titled “Klout Changes Affect Millions” I state “It’s NOT the number increase or decrease that is important here. What’s important with the score is what that number can tell you and how you can make changes to your activity”. As I watched my score and others in the industry continue to decline for no apparent reason, I also began to watch and compare my Klout data with some major brands to try to gain some insight to share to all of you. I noticed only three significant differences between my @fondalo account and some of these brands, celebrities and industry experts:

1) Number of followers – most had thousands more than I do.

2) Following percentage – most don’t follow more than 10-20% of their followers. I follow folks that aren’t spammers and that are my target market. (as should all social media marketers)

3) Engagement – most do not respond and engage with their fans/followers at any significant rate.

At this point there has been so many posts and articles about the substantial changes Klout made and the anger toward the lack of explanation and transparency or any semblance of sense around what has resulted. So I don’t want to add to that. I did however begin to have a hunch regarding the algorithm when comparing my score to some brands.

I know I influence thousands on a daily basis, based on conversations, mentions and RT’s, let alone the significant number of comments, likes and shares I experience. We see significant ROI with our social media efforts as well, which in my opinion is a huge indication of our influence. One rule I have within social media is that I respond to every single mention, comment and RT. Whether it is furthering conversation or a polite thank you, I firmly believe this is important. Most of the larger brands out there do not. I also know that as a social media marketing software company, our target market and therefore friend, fan and follower base is mostly made up of peers with similar or lower scores than myself.

*Note – the time between Klout’s change and the start of my experiment, I wrote a blog post that received the highest traffic, views and comments, as well as being shared more on Twitter and Facebook than any other article I have written. Furthermore, that same period included a Friday, where my twitter account receives literally thousands of mentions in a single day. All the while my score continued a steady decline. Is that possible if influence is truly being measure?

All of these things got me thinking, then asking myself some questions…

1) Does Klout now determine you are more influential if you DO NOT engage/respond to everything?

2) Does Klout now determine you are LESS influential if the folks/brands that mention engage with you have less influence than you do?

I embarked on experiment that was very painful. My hunch regarding number one above was combined with a change in activity to see if anything interesting resulted. Over three days I only responded and engaged when someone directly asked me a question or their mention or comment really required it. All other likes, comments and RT’s were ignored. (I want to note this was so hard for me. I pride myself on not being like the other arrogant social mediaexpertsby always responding and helping others, etc.) It was incredibly difficult to stay firm with my activity pattern change to ensure these 3 days resulted in accurate findings.

Results:

My hunch related to number one above was completely wrong, but I noticed something that is even more disturbing. Based on my reduction in engagement and response to my friends and followers as well as blog comments, you would expect that the rate my score is declining would substantially increase. This did NOT occur. As you see in the image, it maintained the approximately the same level of decrease during the experiment.

Why is that important? It says that according to Klout’s new algorithm, responding and engaging with your friends, followers, fans and Blog “commenters” (which I might add builds relationships and therefore increases opportunity for ROI) has little to do with your influence in the social graph. Disturbing doesn’t even begin to describe my feelings about this discovery.

What’s more, the data in the image shows two additional frightening discoveries that should be pointed out:

1) Amplification Stabilized – Even though I stopped conversing at usual levels, my amplification stopped declining at the previous steady pace since the Klout changes. Seriously? How does that remotely make sense or be in anyway possible? So by not engaging and thanking people, how much I influence them increases? Disturbing to say the least.

2) Network Impact Increase – After a steep immediate decline at the beginning of my experiment, my Network Impact began to show a pattern of slow steady rise.

So while my overall Klout score continued its steady decline, my overall amplification influence and network influence became more stable or even started to rise.

As a previous proponent of Klout and someone that consistently saw their system as one with proper focus, leadership and focus on accurate depiction of the social graph, I am dismayed. As a heavy social media user, consultant and someone that has driven ROI and strategy for many businesses large and small, I am now almost speechless at what I have found.

Based on my experiment, it appears that Klout’s algorithm changes are not focused on improving their social measurement system, but a clueless attempt to prop up larger brands and celebrities anti-social behavior and stifle effective relationship building that leads to ROI for those that do it right. -OR- even worse, tech geeks and scientific formulas that have no real understanding of social media and it’s proper use in business.

Being a social media technology company, you would have thought Klout would have better managed their decision to release this new tech as well as get in front of this story with better answers that made some kind of sense. Instead, they decided to let social media do what it does and react to it, albeit poorly. This entire thing is an incredible example to other brands on how NOT to manage a crisis and to tech companies on how NOT to let your head get too big. Your customers and users should have a lot to say about the game, not just you.

I am not upset at all that my score lost 11+ Points. I am pissed that a company I trusted, upheld and cheered for has fumbled in such a horrific way. The only way Klout could save face with me at this point is to do what Bank of America is doing in the banking sector. Roll back to previous algorithm and make small incremental changes that are accurate and thoughtful.

*note – I did not add or remove any social media platforms during my experiment. I only altered my engagement pattern. Nothing more, nothing less.

112 Comments

Filed under Facebook, Klout, Social Media, Social Media Marketing, Social Media ROI, Strategy, Twitter

112 responses to “My Klout Experiment And The Disturbing Results

  1. Thanks for sharing your findings Rob insightful as always. just some initial thoughts.

    I like you would interpret influence as an action taken by a third party as a result of my communication. So I would class a share, like, retweet etc as an action.

    New Klout says yeah that’s nice and all but there are a whole host of other factors like the influence of the people doing the sharing, likes, retweets, @ mentions and so on. – This is interpreted as Klout with a small k.

    “3) Engagement – most do not respond and engage with their fans/followers at any significant rate.”

    I suspect but do not have the data to prove this; but if your content or actions are responded to by people very close to you this means less than
    If the share etc is done by an account you have no connection with .

    So if i am on the right track you could ignore your immediate followers and as long as they want your content / message and share it with other beyond your circle this means Klout with a large K.

    Gaining influence this way makes no account of sentiment by which I mean if 200 people share your link and they are saying ‘Look what this fool is saying’ Would you class this as influence? For Klout…. hell yeah… good or bad people distanced from you are reacting.

    We will try this out by having a brand account and then personal accounts.
    As we can see that our main account has gone from being classed as expert to conversationalist.

    We will then use the brand account to share on message information and links. Our personal accounts will be for conversations and engagement.

    Rob great post!

    • Related to Brand/Personal account comments at the end – You should ALWAYS do this, regardless of the Klout situation. Leveraging a personal and brand account is second best, following combining them into a single account.

      Thanx for the input!

      Robert

  2. Thanks for sharing this. I really feel as though Klout is one of those new neighbors that everyone is talking about, knows very little about and is somewhat skeptical of. I find that the ecosystem that makes up social media seems to be growing but is fine where it is now (to a point). Klout, at least to me, seems to be something else to worry about rather than a tool to use.

  3. Hey Rob – Darn you. I just a few days ago stopped thinking about my Klout score. Now I’m back at it again.

    I went from a 66 to a 49 on black (whatever day it was). And went from a “specialist” to an “observer”. Really? And, wow, what a staggering drop. I changed nothing, of course. And klout still won’t count my wp.org blog. And won’t allow me to combine my multiple personalities (multiple blogs/twitter accounts ,etc).

    Klout’s distribution graph said most will not change. Some will go up. Do you know anyone who moved up? Oh, and Justin Bieber stayed at 100. Go figure.

    The thing I most dislike about Klout (and others like PeerIndex, Empire Avenue) now is the incessant addition of more and more places to influence. Now there’s instagram and so many others – how can anyone be influential on them all. I think we all need to stop letting influence sites tell us where to go and what to do. And just go be ourselves. Influence be damned. :-)

    There. I said it. Thanks for sharing your experiment.

    • I for one have not (but for this experiment) changed my activity at all. Nor will I. It’s about relationships, value and ROI. We do all three exceptionally well and will continue to do so.

      There is a place and a need for a third party measurement tool in my opinion. Klout just ensured they wont be that tool…

      Robert

    • Apparently 40 is the new 50 and so on….. if I am reading what others are saying correctly then a score of 50+ puts you in the 95th percentile

    • Tim, I totally agree regarding the “incessant addition of more and more places to influence.” The only thing I like about that is it has introduced me to some I hadn’t heard of, like Instagram. As a digital native I think it’s my job to go and explore them and at least figure out how they work and what that might mean for my own personal online social scene.

      Long story short (and thank you Robert too for this great post), I think Klout screwed up. You know how in a spreadsheet if you apply a rule to a rule to a rule sometimes everything just blows up? I think they started applying algorithms to algorithms until things got so messed up that they can no longer even figure out how they got there.

      Best thing to do? Scrap the whole thing and start fresh. You can’t fix a broken system. The only possible way to resolve the issues now is to throw out all the algorithms and start anew. What that means to me is that right now there is a HUGE opportunity for anyone out there who wants to step into the arena with a new influence site that gets it right.

      • Thanx for the input Dana. I think the problem is they tried to go too far, too fast with their “improvements”. It’s science, not art. Therefore a stagger step approach of incremental improvements should have been the approach.

  4. Great article! Thanks for doing the dirty work by putting the SM scene on hold for 3 whole days to come up with these results.

    I truly hope that they simply are too tech-savvy to understand the real-time communications community that we have become. Engaging is the key and there is no substitution. If Klout thinks that success comes from interacting like a celeb they are mistaken; their celebrity came before SM not because of it.

  5. Excellent article. I had noticed odd things as well. At first I was excited that my Klout score jumped back up to over 50, but the more I engaged the lower it became. Then after a few busy, non-SM days, it grew again. It just doesn’t make sense.

  6. Great post Robert. My score (albeit a lot lower than your account) followed the same patterns. The disturbing part of this whole thing is that it started to alter who I was on twitter as I thought perhaps I was “off” in my approach. Now, I’ve decided perhaps they are off in their approach. – Holly

  7. Klout scores are like the SATs–they’re not entirely fair, but it’s most widely accepted standard of measurement. It should by no means be used as the only indicator of Klout, but it provides some parameters in a cluttered universe.

    However, even before the recent changes to Klout, I was often bewildered at the lack of movement of the Klout score for a brand I manage. RTs, replies, and followership all grew steadily over a period of months without any change in Klout (with high engagement both from us and our audience).

  8. Two things,

    First, i’m not sure on the legal ramifications, but Hootsuite should consider offering alternatives to Klout. (I’m not sure if they do this already and I’m unaware). In preferences, select the rating system you are interested in using, or get rid of the ratings in general. The reminder to me of Klout is viewing user accounts in Hootsuite.

    Second, there should be an interaction rating in Klout. Rewarding companies, brands, and individuals that reach out to others and spread messages. It is nonsense that the highest Klout is do to those that already have the advantage of being famous. It shouldn’t penalize celebrities, rather, offer opportunities for harnessing relationships, as you mentioned before. After all, social media marketers everywhere say it’s not the size of your following, it’s how you use it. Quality over quantity.

    Thanks for sharing

  9. Robert – A great experiment. Thank you for sharing.

    Like you…
    – I ALWAYS respond back. That must have been hard for you not to respond back as you always do! In terms of the Amplification Score and your experiment, I wonder if this is a case of “thee with the last word wins” meaning if you thank me for a RT, and I do not respond to your thanks, Klout registers that you made a post without receiving engagement back. This could explain high scores for low engagers like Bieber or large brands.
    – I started out giving Klout the benefit of the doubt. Their original platform was built on Twitter, and soon will consist of 17 or so connected platforms (many of which make no sense, but that’s a different subject!) I sort of understood why they needed to re-think their scoring. However, I observed that after the initial score “hit” (I lost 14.4), the scoring seemed to make zero sense. It seems that complete strangers (not following/followers) with very low scores who may have RT’d me one time months ago have a bigger day-to-day score impact than a RT or @ mention from the likes of Ann Tran, Mama Britt, Dabney Porte, Glen Gilmore, Michael Todd and you (!) all in one day. “Dismayed” was the perfect word choice.

    Despite Klout’s assertion that “adding a new platform will never hurt your score,” I’m considering my own experiment where I disable Facebook to see the score impact (personal account, high engagement with every silly post among fabulous, but very low scoring, friends.)

    Thanks again, Robert.

    • All great points. I think we may never know on some of this. Crazy!!!

      Good input as always Susan

    • Susan – I removed every one of my accounts (Facebook, Google+, FourSquare, YouTube, Flickr, Instagram, LInkedIn etc.) and there was little to NO impact on my score. This in itself proves to me the lack of credibility in their numbers.

      Also, check out the 3 peeps they had listed as top peeps I influenced last week. Tell me how can these 3 people have the exact same stats (number of followers, following, tweets) AND the only two tweets they have ever sent are the same? I get hundreds of retweets and mentions a day and they determined these are the people I influence most.

      Seems their amplification score is what has really wacked the scoring system out. It appears to now be a percentage/probability based algorithm. So if you have a large following then the % chance that you’ll be amplified goes down.

      Regardless I am no longer even going to discuss Klout with clients, training seminars or online training we conduct and will actually encourage them to ignore it for now. There are many other aspects of online marketing and social media any business can be focusing on than a number that means next to nothing!

      http://twitpic.com/7a9u45

  10. Very interesting, thank you for the research and post.

  11. Fantastic insight in this post and as a BSN/microbiology geek, I love that you conducted a “lil experiment”; albeit, not a scientific one. I am quite disappointed in Klout’s “change”– seems they gave us the change without the hope (kind of like some others I know — ahem — our government). I’ve always felt skeptical about Klout. Even before their algorithm change, I noticed that If I actually worked on direct income producing activities (rather than tweeting all day), that my score dropped drastically as did my access to prime perks.

    I’m done. Broken up, at least for now. A friend of mine recommended PeerIndex. I need to check it out.

    @Andrew Clinkman – you’re right on target when you say, “Second, there should be an interaction rating in Klout. Rewarding companies, brands, and individuals that reach out to others and spread messages. It is nonsense that the highest Klout is do to those that already have the advantage of being famous.”

    • Thank you for the comments Samantha! Truly a disturbing situation…

    • “If I actually worked on direct income producing activities (rather than tweeting all day), that my score dropped drastically as did my access to prime perks.”

      Even before all this happened, I think there was some Klout fatigue going on, or at least backlash to the “is my number up?” obsession. One comment to a blog that really stuck in my head was a guy who wrote that he viewed dips in his Klout score as a good thing because it meant he was doing something more productive — be it work, spending time with the kids, etc. That really struck a chord with me, and rang a bell when I read your comment here.

      • Ha! That’s a fantastic perspective your friend has, Tom. I love it! While I do think Twitter counts as productive activity for business, like everything else in life, its use can become out of balance quite easily. Thank you for giving me that glitter today. I’m sparkling on, now, because of you.

  12. One of the things that Klout doesn’t take into account, and should is a personal blog. It doesn’t take into account the comments I get and all of that on my selfhosted site. Unfortunately, it doesn’t. That irks me

  13. Funny true story: While this was before the algorithm change, I took a weekend off Social Media. No Facebook, Twitter or Foursquare. I expected my score to drop, but I came back on Monday and it stayed the same. I got a ton of engagement that day and the next day it jumped up ten points. It was almost like it re-calibrated and zeroed out at a higher level so when I returned it saw a huge improvement. Not sure if this would work with the new (cough, cough) “improved” Klout.

  14. Jabers

    You people all seem to expect that the measurement will reflect your expectations, but maybe your assumptions are wrong? Someone who participates more in social activities isn’t going to necessarily be more popular. They may just be thought of as annoying. A bit of mystery and unavailability is considered attractive. My point is that perhaps Klout’s algorithm is perfectly accurate, but maybe you are just unhappy with what this says about society as a whole. I tend to agree and wish the world were different than it is.

    • Interesting thought Jabers, unfortunately as someone that has a decent handle on Social Media activity, what drives results and how to actually influence, I disagree. I think my experiment clearly shows this and a quick check of Klouts fanpage and Twitter interaction reveals much about their understanding of the medium.

      Thanx for the interesting view on the subject though.

      Robert

  15. Hi,
    Great article. For me my Klout score went up and now my score is on par with others who I believe are at a higher level of influence.

    There have been users of Klout who were always trying to “game” the system, now Klout has literally become a game – giving away +K’s, garnering achievements and badges like “Klout addict”, etc.

    Curious though, was your Klout only connected to Twitter during this experiment? If not, maybe that explains the uptick if you were engaging on other platforms.

    • As my note indicates at the bottom of the article, I left all the many connections I had with Klout as they were in the experiment. I did not remove any or add any more. I simply stopped engaging and having conversations, driving traffic and doing business. Crazy!!!!

  16. i know this new algorithm is heavily linked to facebook – more than twitter , i like you tried something and spent a few hours posting and interacting with facebook not twitter and my score jumped , spent less yesterday and it went down again – something wrong somewhere – I have friends who do nothing on twitter – dont even have an account and only post on facebook and they have higher scores and klout is recommending me to follow , so its really a score and how much time you spend on facebook not how influential you are – why does it encourage you to add all those platforms when they have no effect on the outcome

    • I get hundreds of likes, shares and comments on my posts on FB every single day. I engage there as much as I do on Twitter. In my experiment, just like on Twitter, I only stopped engaging, liking, commenting and having conversations.

  17. It seems very much like high school all over again – cliques and peer pressure galore.

    Almost everyone is submitting to the peer pressure of an external source. It’s as if people were addicted to seeing their score go up, and when it went down they spiraled down with it, and when it went up it was like a high that they needed more of. They needed another fix. And when their dealer changed the terms some freaked out, some quit all together, and others were even more addicted because their scores went up.

    While it’s nice for companies to try to measure someone’s influence, in the grand scheme of things what does it actually achieve? Especially when people are rewarded for being anti-social and NOT interacting with their loyal fans.

    If you are the type of person who wants to talk to everyone who talks to you – do it. If you are the type of person that only responds to private messages, then stick with it. If you’re the type of person who doesn’t interact at all, so be it. People are your fan because of how you are, not how some website says you should be.

  18. The only problem I see with your test is that it doesn’t take into account how Klout actually delivers a score.

    The score isn’t calculated daily, it’s calculated as a monthly average based on the average scores of your last 30 days.

    As such, a test like this won’t prove anything unfortunately, because your score is fluctuation based on 30 days of data and not just a single daily set.

    • We discussed this on Twitter. I completely disagree with your assumption or mislead information provided you via Klout. For 9 months I have watched my score move in direct connection to new blog posts and follow Friday mentions. Though it may be based on a monthly run, it uses the DAILY in that algorithm to produce the new daily score. This is undeniable…

  19. Wow…….that is seriously disturbing. I used to pay attention to it, stopped for a while, and came back just prior to the changes and this very disturbing to hear.

  20. Your observation and sentiments on Klout are very similar to mine. I am a huge supporter of Klout when it started. I knew it was not perfect but I always say it’s a good start. Now I am not sure if people should even start with it. I am a blogger and I am concern that the influence I have as a blogger is not affecting my klout score, when I think it should. I am on a self-hosted blog and tehy only connect to the wordpress.com account. I continue to watch what’s going on at Klout and your blog post is helpful. Thanks.

  21. Excellent article. I’m among those who have been wondering just how accurate Klout is. From what I read of your experience – they’re way off.

  22. I’ve been watching a similar freefall across the board with my scoring components (including amplification and true reach). I have two forays/ideas to further test the algorithm.

    – I did my first vlog which introduced another channel to my mix (YouTube). I already had it connected, but it was pretty much dormant with private videos (of judo throws) only

    – I’ve kicked around the idea of changing my comment login, when it is allowed, to use my Facebook credentials versus Twitter (which is my preference). A common theme is that Facebook still seems to carry a heavier weighting than the other networks. I guess you could argue that on relative size of audience (800M FB to 200M Twitter users…120M for LinkedIn). However, I am one of those that keeps my FB to primarily personal. I can still influence there, but not as it relates to my job which is where a lot of us want to have the most influence as thought leaders so that it results in measurable ROI.

    BTW, I am still lobbying (voice in wilderness) for the addition of Disqus and Livefyre to the scoring immediately. I think they are great measures of engagement and implied influence.

    Hmm, wish this blog post would have come out before my vlog discussing Klout. It would have made a nice contribution to the other post I used from you!

  23. Robert, you’re the first person I’ve seen backlashing against Klout’s recent changes that has actually taken the time and energy to conduct an experiment with which to base your opinion on. Kudos for that. I have a contrary opinion, and here are some of my thoughts:

    1. I think your experiment uncovers quite a few truths, but I don’t think the truth is that Klout is an irrelevant gauge of influence. I think the truth is that influence means a lot to many different people, and Klout has shown that their algorithm is agnostic to a few key issues we previously thought mattered in overall influence. After speaking with Joe Fernandez, I got a pretty firm understanding that the influence scores and algorithms will show how much you can cause activity with the least amount of action. This is the Justin Bieber case. Joe hated that Justin Bieber had a 100, but he truly creates a lot of interaction from very little. Thus extremely influential.

    2. Engaging doesn’t make one influential. This realization stings. Or it’s just a different metric altogether. Klout doesn’t measure engagement, they measure influence. I’m not entirely influential if I start and continue many conversations, that makes me engaging. I am influential if I post one idea that thousands act upon. I’m even more influential if those thousands also have a high level of influence. That means that with little effort I can cause more influential people to react, which means my message is being reached by more people.

    3. Influence is touchy for folks, but the problem is somewhat embedded in our ego. Understand I’m not undermining your opinion by saying Klout hurt your ego. I am saying that Klout measure overall influence, and in a world of 7 billion, it’s a pretty tall order to be considered influential. As Klout gets better at their algorithm, it will be harder to maintain a high score. Most people dropped, which means to me that they were inaccurately influential before. Google does this all the time with their search algorithm. SEO folks pull their hair out because Google is out to get them by changing their algorithms so their previously relevant search results are no longer relevant. Klout did the same thing here.

    4. Rather than turn your back on Klout, who previously had your loyalty, maybe consider what they’re trying to achieve. For instance, you mentioned “it appears that Klout’s algorithm changes are … a clueless attempt to prop up larger brands and celebrities anti-social behavior and stifle effective relationship building that leads to ROI for those that do it right -OR- even worse, tech geeks and scientific formulas that have no real understanding of social media and it’s proper use in business.” On this, I’d argue that either of these positions don’t make sense for Klout to do. For one, by selling out to larger brands and celebrities Klout would render it’s service completely irrelevant and go out of business. This is not their intent. Secondly, they people behind their business do understand social very well, they just have a completely different level of view than you do. They aren’t looking at what brand is bringing more ROI by using it well. ROI isn’t part of the equation in any way.

    Basically, none of us are as influential as we’d like to be. Use these new algorithm changes as a challenge. It excites me to see that my score lowered because now I have to figure out new ways to be more influential. It’s all a game, and it’s shorthand for understanding a true influence, but it’s more useful than it was before in my opinion. Remember, it’s not about being engaging. Very sorry for the verbose response!

    • I appreciate the counter thoughts, but let’s be very clear here.

      1) I never stated nor do I believe that engagement equals influence. On the contrary I detail what influence is related to sharing of content, likes, RT’s comments and the like. Very big difference.

      2) My Klout score has zero to do with an ego. If you took the time to read my various other posts and support of Klout, you would realize that is just not the case. My attempt is not to complain about a score that is said (by Klout themselves) is in the 95th percentile, after all who in their right mind would complain about that? I am rather trying to point out failings with technology and business decisions that have made them irrelevant to many, including myself.

      3) Again, read my previous posts. They have continued to dodge questions or produce anything resembling transparency related to the changes they made.

      4) I am exactly as influential as I’d like to be and our return on investment and value provided to our friends, fans and followers can easily be seen.

      And again, you remember that I never said it was about being engaging. I think I have been in this game long enough not to make such a statement.

      *steps down off soap box…

      I do appreciate the counter views you clearly spent a long time writing my friend, but I wish you had taken that amount of time seeing what I have written extensively on related to Klout and Social Media in general before making the comments.

      • Hey man, I’ve only got so much time available… =) And I wasn’t attempting to be accusatory on most of those points. Like I said before, I wasn’t stating your ego was the reason. I was stating that the problem most have with the drop is “somewhat embedded in our ego”, not that your ego is the reason you’re disenfranchised.

        I’m not saying you don’t understand the difference between influence and engagement, but I do believe you stated quite a bit of opinion around engagement. Specifically that you were disturbed that “responding and engaging with your friends, followers, fans and Blog ‘commenters’ … has little to do with your influence in the social graph”. I don’t believe it should have impact on your influence, as this is from the position of engagement.

        Also, Klout should never release their algorithm or provide transparency to a large degree. Should they do that, any competitor could hijack their algorithm. Imagine if Google did this. No one is up in arms that their algorithm is proprietary, and it affects the entire landscape of internet marketing. Klout should stick to this model in my opinion, or else they will be rendered completely irrelevant by competitors.

        Unfortunately, time runs short in my life. Should you be able to afford me some, I’d be happy to read all the content you’ve written on the subject! =) In the meantime, I wouldn’t have had a response to this blog for another 3 weeks…

      • Brad,

        Thanx for clarifying… I won’t break it down point by point. Love the banter and discussion!!!!

        More importantly I appreciate the feedback and thought you put to your comments. A rare thing!

        Robert

  24. Mary C Barkley

    Hello!

    Nice article! As many others have indicated, the new scoring system is not transparent as the corporate doublespeak promises. Thanks so much for offering up the experiment with your own account also.

    I suspected when the precursor announcement came out about upcoming changes, we were going to have some unhappy campers, me not being an exception. Then BOOM on Monday 10/24 there is was. New, lower scores to the many folks that have championed KLOUT and helped them gain traction.

    I think the long term end game is for KLOUT to make money selling to brands with their giveaways now, even FTC laws are being questioned.

    Coincidentally that same Monday, I coined #OccupyKlout after a phone call from OccupyWallStreet that same day. A high school buddy of mine is in NYC and wanted to know if I’d help write about their cause. Then it hit me.
    Ironically, several folks have challenged me about the name, but it is in my Twitter stream.

    More to come in the next few days….

    This is not about self-promotion but answering a cause-related marketing need.
    Many of us depend on integrity in the marketplace.

    I’m still curious as to who is handling the PR at Klout, who the investors behind KLOUT are, and who are on actually their board of directors. I’m starting to thing this is a little bit of a country club and we got pulled in based on a very basic need of acceptance and wanting to be liked.

    All for now, cheers,

    Mary Clarkson Barkley
    Founder and CEO
    Beatrix Group
    Brands Made Better
    980.355.1740 EST
    LinkedIn/@MaryClarksonBarkley
    Twitter/@Marycbarkley

  25. I haven’t used Klout much but I have been skeptical of what I have seen. In all honesty I don’t see how an algorithm is going to judge what I post and how people interact with it. Really I think we have to take a step back and think that social media is not a popularity contest (contrary to popular belief) but a means of communication. In other words social media should be about collaboration and team work rather than competing against your “friends.”

  26. Problem is, if you don’t have enough klout, then people won’t get the message about klout.

  27. Wonderful post Robert. As someone who I look up to and who engages constantly (seriously how do you sleep) your input on this matter is really important to me. For me this quote said a ton:

    ‘I also know that as a social media marketing software company, our target market and therefore friend, fan and follower base is mostly made up of peers with similar or lower scores than myself.’

    As someone who made a really big career change this year into social media marketing I value the power of engaging with everyone who wants to engage with me. Many of who, like you state, might have scores similar or lower than my own. Engaging with many people not only increases my knowledge about a ton of things, (it’s like learning without paying a huge tuition bill) but allows me to assist the companies I work for. While I do take issue with the fact that my score seems to be negatively effected if I choose to concentrate on engaging with people who have lower scores than myself, what really offends me is that this new klout seems to penalize people just joining who really have something to contribute. This new ‘system’ makes it harder for people to engage with new individuals if they want to keep their klout scores up. As someone who came back to twitter in jan of this year, and whose klout at the time was probably really low, I can imagine it would have taken me longer to get people to pay attention to me in the space if they were constantly thinking about klout. While I personally don’t care too much about the actual number anymore ( I won’t lie. For a while I did care) the reality is some people really do. Klout markets themselves as ‘the standard for influence’ and with that comes responsibility. I do not believe them when they say we care about what our consumers think. If they truly did someone should have stepped up and answered some questions immediately. This new system they set up creates too much inequality in my opinion. While I understand the value of metrics, the thing I love about social media seems to be cheapened by klout’s system. What is it that I love about social media: Engaging with new people and companies who have great things to contribute and who want to connect in some meaningful way.

  28. Fascinated to see how much of the re-scoring has to do with Klout working with Wahooly
    http://mashable.com/2011/11/03/klout-startups/#comment-17788763

  29. bauserdotcom

    I can barely figure out what you think your experiment was testing. Klout has ALWAYS said they were NOT measuring your activity — they’re measuring responses to it. Likewise, they’ve previously said that follower count was not a major factor. You (and many other Klout fans) appear to have been operating under a major misconception there.

    I think you ALMOST saw something important when you realized a lot of high influencers aren’t following many of their followers, but you’re looking at it too simply. It’s unlikely that simple reciprocity reduces the value of a response. More likely, high interconnectedness within groups does.

    I’d hypothesize the problem with you marketing guys who experienced K-score drops ISN’T that you always follow your followers; it’s that all of you follow all the other marketing guys. Think about it this way: Which spreads your message farther — A retweet from somebody who is followed by the same people who follow you, or a retweet from somebody whose followers don’t follow you? (Answer: The second one, because all the people in the first one ALREADY saw your tweet the first time it went out.)

    Celebrities with poorly-networked followings (like Bieber or Obama) get more bump from retweets because their followers aren’t as tightly interconnected. When their tweets get retweeted, they’re more likely to be seen by somebody who didn’t see the original tweet.

    Remember, the K-score isn’t a measure of influence within a specific community; it’s a measure of universe-wide (in the sociological sense) influence. If the majority of your feedback is coming from the echo chamber of a tightly-networked group, it’s probably not affecting your global influence.

    • Michael,

      It was not about the activity measurement. Activity drives action, which results in influence. That’s social media. So not engaging in conversations which lead to additional activity, links, clicks, comments and likes, the influence you have SHOULD decline at a higher rate than when you do nothing. It’s really pretty simple.

      We disagree but I appreciate the comments. Always spices it up!

      Robert

      • bauserdotcom

        Ah, but you’ve assumed two things not in evidence:

        1) That the actions measured by Klout (retweets and other mentions) are statistically correlated with the amount of directed interaction you have with others. Simply put, you haven’t demonstrated that thanking people for retweets increases the amount of retweets you get (or perhaps better put: that NOT thanking them decreases the amount). You needed to do that before your “experiment” could test what you claim it was testing.

        2) That solicited responses (replies from people you mentioned) have equal weight with unsolicited replies. It would actually be logical for Klout to weight unsolicited mentions higher, to minimize people gaming the system by begging high-influencers for attention. (On a related note, Joe Fernandez has gone on record saying Klout filters for people who “use twitter as a chat room,” which suggest they are thinking along those lines.)

      • And you are assuming you understand effective social media. Statistics and algorithms clearly can’t effectively measure influence, especially when scientists or number crunchers like yourself are determining those algorithms. Let me be clear. Social Media influence can only be determined by how well a brand/account drives action. What action? That’s the question. That question is not answered the same for every account. For us it is directly connected to shares, comments, likes and retweets, all of which matter none without user base growth and revenue, which we see dramatically.

        Regardless, I stand by my article and what it states. Now, I will let you continue to be the Internet Trouble Maker your website claims you to be and allow you to get back to the totally expert role in social media with your 90 followers. Troll…

        Good luck my friend!

    • Michael,

      Furthermore, your Hypothesis surrounding the problem with us “marketing guys” and the RT scenario is that Bieber and Obama idea you share later. They do not gel. You can’t have your cake and eat it too…

  30. Thanks Sunanda, your response was well received. Social media should be a tool for engagement plain and simple. Its similar to the phone. I pick up the phone to become “engaged” in a conversation. That conversation serves a purpose and then we part ways and go on to other tasks. Do we really need to measure these things? Even from a Marketing perspective if we become too focused on metrics I think we’ll all lose the point of the whole thing. We have to ask ourselves what we want to achieve with social media and if it’s purely a numbers game we are really taking the “social” part out of the equation. Social interactions are not built on equations and algorithms but respect, love and understanding. Sorry if I’m sounding a little too “soft” but that’s the truth plain and simple.

  31. Excuse me but who gave these Klout guys the “keys” to the city anyway….as Robert would say…”just saying”

  32. What an eye opener. (A polite thank you will suffice.)

  33. Nothing gets people worked up quite like Klout. How do you think your experiment would have been different if your score initially rose? Don’t sweat the trolls, keep up the good work!

  34. Pingback: Klout for the Little Guy « The Blog

  35. Great experiment, and great post about it!

    But beware of one thing. Klout is a piece of code sitting on some server. Like a piece of code that is in your iPhone. It does great things. As people who designed converted their great ideas into a great product.

    But developers are developers. Your new super trooper iPhone 4S is leaking battery juice like mad. Because someone made a simple mistake in the code somewhere. Great product otherwise. Same is Klout. The great idea, and the product almost follows it. Until it starts leaking the re-tweets. As unimportant measurement. Yeah right…

    Currently iPhone 4s battery lasts shorter than the iPhone 4 battery. A few years older technology (and we know what it means in IT) – and the new one simply does not deliver. Same is Klout. We accepted it because it brought some measurement in the social activity. The current version is worse than the last. That is normal in the IT world. The next one will (maybye) be better. Or someone else will do a better product before Klout gets its product back on track. And we will have a new and better social media measurement tool. And Klout simply bites the dust.

    • Ivan,

      You are correct and mirror my several posts on the subject. Some decisions, lack or, or failings in the marketplace render some pieces of code, business models and companies obsolete. BlackBerry, Palm, AOL, etc. Sometimes when you make a horrid business calculation it’s like jumping off a bridge without a parachute.

      I am a firm believer (and have written on this extensively) that a good third party measurement tool is required. Unfortunately, the leader that provides that will have trust from advertisers and more importantly users, that what they are providing and the changes they make are real, valuable and thought out. Klout has lost all Clout in that area.

      Thanx for the great comment!

      Robert

  36. Personally, I don’t understand why we have to “measure our influence”. Social media is more than a popularity contest!

    What annoys me most about Klout is the fact it signs you up without asking if you want to opt in, and from my research it’s not a simple task to get removed either (and removed properly, not just “hidden”).

    I can’t simply add people to my newsletter mailing list, yet a company can add me to their silly scoring system and use my details on their site? It doesn’t make sense to me.

    I’ve never given much regard to Klout scores and I won’t in the future either, particulary after your (rather worrying) findings.

  37. I pretty much did the same thing that you did. With the new algorithm, I lost 12 points. But a guy who is in my stream that just shovels out tweet after tweet after tweet with no engagement, fared worse than I did. That made me happy. I felt justified. Yay! Because, as you said, social media should be social and this guy is not social. After that, I too noticed that my score continued to decline for no apparent reason. Although I didn’t do such a detailed analysis as you did, I learned that if I spoke to “less influential” followers that my score would decrease. Well, I’m no one’s mean girl and I just deleted my entire klout profile today. I don’t need a number to tell me if I’m influential or not. I love all of my followers no matter how “influential’ they are and I will continue being social with any and all of them! :-)

  38. Robert, aloha. Horrifyingly amazing. Thanks for sharing the results of your experiment. Now i am off to share them with others. Aloha. Janet

  39. “Based on my experiment, it appears that Klout’s algorithm changes are not focused on improving their social measurement system, but a clueless attempt to prop up larger brands and celebrities anti-social behavior and stifle effective relationship building that leads to ROI for those that do it right. -OR- even worse, tech geeks and scientific formulas that have no real understanding of social media and it’s proper use in business.”

    — This… this is why so many people are getting disgusted by the Klout situation. It’s not honest, transparent or reflective of our musings online. How is being disingenuous and stiff online a model of influence? If that’s the case, my shoes have a higher Klout score than I do! No thanks!

  40. ReverendShaft

    Klout is complete rubbish. I (intentionally) have less than 200 Facebook friends and about 25 Twitter followers, yet apparently have a higher score (upper-50s) than a lot of the social media professionals who have thousands of engaged followers. All this despite the fact that I post mostly random crap.

  41. This is the second time I read this post. I admire you for sticking to your guns and getting through the rigors of the experiment – even though it was against your nature to not be your responsive self.

    Wish the results would have been different, but thankful to you for bringing awareness.

    • You are so kind Dawn. Sometimes I get a hunch on something and track it down until I have enough information. This was one that, I agree was a huge disappointment.

      Thank you for reading it and the kind thoughts. Appreciate it!

  42. Pingback: Klout is a Catch-22 « The Insatiable Solopreneur™

  43. Robert, I love this post and the comments on it. One thing that occurs to me as I look through the comments on this Klout event, it reminds me of the Shared Twitter Experience that I wrote about on ItsDifferent4Girls. It is happening to a large group of us at the same time. As far as I can see, it is mainly hitting marketers and social media pros. The vast majority of us strongly supported Klout before this happened. It is emotional. Kind of like going through a natural crisis together. I have been looking for those that are doing research on the issue. I am adding your post and this comment to my collection of thoughts on my casual Posterous blog where I am working through this bit by bit http://asklindasherman.com/klout-alternatives-and-trying-to-understand-t

    • thank you so much Linda. I appreciate you including this on your blog and all the comments and insight you shared.

      For me, I am not spending anymore of my time on this subject. Klout is now irrelevant in my mind and I will keep on driving ahead with my main measurement = results :-)

      Thanx for the comments!

      Robert

  44. Liz

    Jessica from @FourPlusAnAngel pointed me in your direction this morning. I wrote a post last week about my accidental Klout experiment: http://www.bellebeanchicagodog.com/2011/10/klout-algorithm-is-inaccurate.html

    In short, my completely INactive, zero engagement, barely any followers Twitter handle jumped up in score to a point that it exceeded my main Twitter handle that has tons of engagement, way more followers, etc.

    I loved your line about them fumbling in a major way. It makes me wonder how well they tested their new algorithm before rolling out. Also, I’ve heard another great point raised about how all historical data was changed, too. From a true analytics standpoint, that should never be done. A notation needs to be made in all data/graphs that marks the day the algorithm changed. People who were interviewing for jobs where Klout scores were involved ended up losing opportunities because not only did their current score change, but it looks – from the historical data – like they never actually had the higher score they said they did.

    • So true on many points here Liz. I am now at the point that Klout not only made a huge mistake in this change, they have handled the resulting firestorm so poorly in hiding and not taking steps to manage the story and social screaming that ensued, but they are now irrelevant. I am disengaging from the Klout conversation. They are supposed to be in social media measurement, yet they conduct their own social media management, marketing and PR so poorly, their measurement is clearly being skewed by an incorrect world view of what social media truly is.

      Sad…

      Robert

  45. I enjoyed your post and agree with the funky behavior. I was curious if you factored in your trend analysis the two day lag Klout has from when you had the activity to when it is reflected in your score. For example, I participate in a few twitter chats on Tuesday evening and that activity is not reflected until my Thursday morning Klout score update.

  46. Truly appreciate your taking the time to present the facts as you see them. I too was bewildered by the 23 point drop in my score. My mantra is to build relationships wherever I go and so I have a high interaction with my followings. You helped provide insight as to what happened seemingly overnight – thank you!

  47. Marie Haggberg

    Very fascinating read. Unfortunately for Klout, even if they roll back their algorithm, thieir self-inflicted damage may be long term. Consider that Bank of America and other large banks lost about 650K customers to credit unions and community banks, following the now-cancelled B of A debit card fee.

    BTW, from a software development perspective, your quality assurance testing approach and discipline were spot on. Well done!

    Thanks for a great post.

  48. Pingback: Mathew Lowry’s Tagsmanian Devil » Blog Archive » What is influence? or, Why I don’t care about my Klout score (updated)

  49. Like everyone else, my Klout score fell significantly and I have since been conducting my own experiments…

    I unlinked all of my accounts aside from Twitter, which i am most active on. I believe that you are “punished” in some way when you link accounts that you are less active on. Think of it as averages. If I am very influential on Twitter but not at all on Facebook, they will judge me somewhere in the middle. (All conjecture based on my accounts.)

    Second, Klout is no longer judging what WE do (or at least not nearly as much). Klout is more accurately judging what we get other people to do – and how important THEY are. It doesnt matter if I talk to everyone and tweet 100X a day, if no one responds, clicks, etc.

    I found that my Klout score will rise when I get other people with high scores to engage with me – either by ReTweeting or Replying.

    With that said, I think Klouts algorithm change sucks – because it knocked me down. But in reality I think it may be more accurate now as an “influencer” ranker. Klout isnt trying to judge who engages the most or whos better at social media, they are judging who can get a message across that people will take action on.

    Its a flawed system to be true and I hope they improve it, but I think they are trying to do so. Its a free service and WE are the product they are selling to companies who offer perks, so they will tweak their algorithm to give THEIR customers the most accurate reading – not to feed our egos.

  50. Pingback: Why I Deleted My Klout Profile | The Marketing Nut

  51. Pingback: Klout isn't for end users. It's for brand looking to target themMoss Media Labs

  52. I have just joined Klout, so I’m new to all this, but my understanding is that Klout measures influence rather than engagement. While the latter can be relatively reliably measured (volume of two-way traffic), measuring the latter is not so easy, hence the somewhat arbitrary measurement (number of shares, responses, likes to one’s posts, etc.
    Having said that, influence is generally considered a qualitative concept, and the quantification of qualitative information is inherently difficult (I know this as a market researcher, and having just started a business that will rely on just such a methodology).
    I am not a major social media commentator – I just post, share and respond to what interests me – yet I am hovering around the 44-45 mark, which is, apparently, relatively good.

    • Clive,

      My recommendation to you is not worry about your score and instead focus on building relationships that add value and in turn result in ROI for your business. That is the proper measurement of social media.

      Robert

  53. Interesting. Something you are not exploring is the increased effort users will make to raise their score by inviting more friends to engage. This is a mechanism to bring in other users just to stand still. I see that my Facebook likes and retweets increased and my score went down. What can I do but get more users on their system. Doesn’t that sound like a scam? What if our savings interest went down unless I got more friends to deposit in the bank. Clearly unfair, yet we don’t question when this is done with on line reputation.
    BTW, keep engaging. I like our attitude about social media. We meet real people and create real relationships. Strangers are worthless. I am as likely block a twitter follower as follow them. Twitter is a part of the company I keep.

    • Thanx for jumping in guys. Yes I believe it is nearly impossible for an algorithm to accurately measure social influence. It does a great job measuring and tracking topics though!

  54. Pingback: Klout: Promoting Unsocial Media | Albert Qian: The Social Media Dude

  55. Pingback: Klout Alternatives and Trying to Understand the New Klout

    • Thank you Linda. I don’t care what Klout or the others are doing and how it effects any score, nor should you. There is no algorithm that can effectively measure everyone’s influence. There are way too many variables, all of which are different for every single brand, individual and account that would specifically measure that persons influence. Klout is a data gathering tool and add delivery platform for big brands, pure and simple. Don’t get emotionally invested in something that matters none.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s