Why Klout and Kred Fail

Image by Continuum

Last Friday, I spoke at TEDx Peachtree about how any of us can become influential and affect change in our world (you can read my speech here).

The speech took to task social scoring technologies Klout and Kred for failing to predict influencers and movements as they arise.

Attention — and more specifically public recognition of influence — almost always occurs after the fact. Someone who achieves a high Klout score for something notable already created that act.

That does not mean they will repeat their successes and inspire widespread action over and over again.

This core weakness – the inability to forecast influential moments, and the people that create them – forces social scoring into a statistical record, which may or may not predict a future success.

Coupled with endemic failures of rewarding attention rather than real acts of influence, and you have quite a gamble for the business community.

Topical and Time Based Relevancy

Shonali Burke
Shonali Burke Will Tackle Social Scoring at xPotomac

Social scoring will always have a high quota of failure. It cannot predict how complex humans change nor the circumstances that make their ideas relevant (or irrelevant) to their larger communities.

The business world continues to use social scoring with alarming frequency, and that in turn makes understanding what these scores actually mean from an influence and marketing perspective all the more critical.

It’s for this very reason that PR Maven Shonali Burke was added to the xPotomac speaking roster this February 25 to bring a measured discussion of social scoring and how it impacts businesses looking to market online.

I asked Shonali what she thought:

“Here’s the thing with social scores: They’re just numbers. And numbers without context don’t mean anything.

“When people ask me how important social scores are, I tell the story of how, a while back, Klout deemed me a “bacon” and “Kim Kardashian” influencer, simply because some friends and I were playing around on Twitter and Facebook with those topics. We were having fun… it didn’t mean that we had legions of fans hanging on our every word about either of those topics.

“Kred is a little bit more interesting because of what they are trying to do with their Kred Story feature. Andrew Grill has been a guest on #measurePR and he is pretty vocal about “influence score myopia” (as is my friend Pierre-Loic Assayag of Traackr). Now, you could certainly argue that these are all competitive services, so one must take differing opinions with a handful of salt, and you’d be right.

“For what it’s worth, I think Kred is slightly better than Klout, but I still don’t like the idea of people having numbers on their heads, as it were. And no matter how smart tools get, it’s how they are used by people that is what counts.

“Are people trying to game them? Of course.

“Will that stop? Probably not.

“But the fact is that social scoring is here to stay, so we may as well figure out how to live with in a world that, more and more, is moving in that direction. Tools like Klout and Kred (and numerous others) may give you a starting point when it comes to determining relevant influencers, but it’s just that – a starting point. Your brain needs to do the rest of the job.”

The Great Game of Statistics

From my personal standpoint, the problem with statistical performance is that the way the human condition works, we change over time. Quickly. We age, we evolve and devolve, grow and recede. Performance gets better or worse.

In the case of winners, repeat performances are rare, and with each success the likelihood of another one decreases.

How many of Ernest Hemingway’s 20+ books (some published posthumously) were truly great books? Arguably three or four, and we are talking about one of the all time masters.

Further reality throws a huge wrench into the equation, facts, circumstance, changing societies, evolving environments, and more all impact the human performance. Relevancy shifts very quickly.

How do Klout and Kred handle topical real-time relevancy? Right, now they can’t other than to update scores in near real time.

However, Kred is evolving to become a more utilitarian tool for brands, managing conversations about them. At the same time, while a smart use, it’s not really the influence predictor businesses want.

In many ways, the great gamble — the one business make when they chase influencers online with high social scores — can be likened to free agent pick-ups in sports. Teams purchase contracts based on past performance.

So we have the prediction game, where evaluating talent comes into play.

What do you think about social scoring’s ability to denote influence?

Register for xPotomac Today!


  • Geoff,

    Social influence scoring is a real bug bear of mine and I maintain that an external agency is not capable of accurately measuring it due to only having public data. The data host would be in the best position (be it Google, Twitter, FB etc.) as they can monitor both private and public activity as well as accurately measure additional factors like click-throughs to help determine who is driving traffic rather than “soft influence” of just likes and +1s.

    I wrote a proposal/essay a while back called the 3 R’s of Influence which states that influence is a composite of reach, reputation and relevance. The current activity model is ridiculous and does not give any true indication.

    Relevance is a big issue, as you state, and I defined it as trust + interest + location + platform. Timing is obviously a major factor.

    Identity and context also plays a part. Who are we and why are we influential? Currently we are graded the same regardless of whether we are a content creator or curator – there needs to be a form of measurement based on what we do not just on the clicks we get.

    Influence is, however, an increasingly important area (rightly or wrongly) and we must be aware of how it works, whatever method is employed.

    • Wow, this is a very interesting comment. How familiar are you with the science of networks as a theory? It seems like you may be within that school of thought on influence.

      • Must admit it’s not something I’ve specifically looked into. Will have to do so. I just have an idea of how I envision things ;)

        Off to do some research.

    • Colin, it’s very interesting, because Traackr uses three criteria in its analysis – reach, resonance & relevance. I’d love to read your essay, btw, if you’d share?

      Geoff – thank you for the inclusion – this is going to be a SUPER event and I am so excited to participate!

  • Great speech, by the way.

  • Geoff,

    Andrew Grill, CEO of Kred here.

    I thought this was a great article, and I agree with pretty much everything you’ve said.

    Just addressing a few points:

    I don’t believe Kred “fails” because we haven’t gone to the market and said we’re THE standard for influence. We’re a part of the equation.

    Also with humans being the way they are (and changing their minds!) we can’t always use machines to predict behaviour exclusively.

    The same argument applies to sentiment – and no-one has really nailed online sentiment because again there are humans involved in the process and we are complex beasts.

    Neville Hobson and others have heard me speak at length (see the presentation on the front page of http://lc.t/a or my piece about influence score myopia at http://lc.tl/myopia

    The scores are a PART of the piece. You still need people to analyse the results, form relationships and speak to each other.

    I am probably unique as I am the CEO of one of the leading influencer platforms, and over the last week in the UK I have become a key influencer myself.

    Read the whole story at http://lc.tl/eefail and also see how my blog post hit mainstream media at http:/lc.tl/eetele

    Only by being a real world influencer can you hope to understand how to use the platforms and advise clients.

    I think all platforms are a way off from “forecasting influential moments” on their own – the need a human touch will always be there.

    Take the proliferation of badly worded direct email campaigns that use all the prediction engines we can throw at them.

    In the influencer world, it is tens or hundreds of influencers, not millions – so we can use the human touch, and influencers like that – we don’t like to have our behaviour predicted!

    Shonali gets what we’re doing at Kred, and I appreciated you including her view on our platform in the piece – I think we move away from a “fail” just on what she has to say.

    Yes any system (including life) can be gamed.

    With Kred, we can SEE Kred being gamed because for every one of the 130 million profiles we update them every HALF SECOND, and we also have a full activity statement with every tweet and every point awarded going back two years.

    I think a profile update every 1/2 second is pretty close to real time when you understand we are processing the twitter firehose (which we have direct access to – from Twitter not a 3rd party) which delivers us 10,000 tweets a second.

    So I totally agree that Kred gives you a starting point – we will always need humans to interpret the social signals.

    As the PR team at EE can tell you from my meeting with them today, sometimes meeting influencers in real life can be a really enjoyable experience.

    Andrew Grill
    CEO, Kred

    • Andrew:

      Thank you for your commentary, and taking the time to write this extensive comment. And we are agreed that KRED, a formula if you would for measuring online interactivity, is a statistic which someone could use as a starting point.

      I think the issue is that even when someone is identified within the system as an “influencer”, it fails to predict time based context, relevance, or even the likelihood of said person being able to repeat their performance.

      Outside of pundits, who repeatedly publish and are in essence modern media, individual success online is really something that happens after someone achieves a notable act.

      As to the meeting someone IRL who is influential, everyone is influential. I mean that. Everyone can be influential to those around them, and I truly believe that. And I have met more than my fair share of real influential people, including two heads of state.

      We need to be careful in our use of these words. Someone who is famous or who has a blog or a big media story is not better or more relevant than others, unless time and context agree with that person being able to deliver actionable results.

      • Again, we’re in wild agreement. In every talk I give about influence, I always say that platforms like Kred and Klout need to come with a “health warning”.

        Yes we also agree that everyone has influence, in our own special way. I like to think we can find those non-obvious influencers that can enhance and enrich our lives in some small way.

        Loved your TEDx talk also.


        • Thanks! LOL. Well, I wish I could have gotten more particular about my professional experiences in that, but TED rules are TED rules.

    • If you’re doing 260 million profile updates a second based on 10,000 pieces of new content then presumably 99% of those profile updates aren’t updates at all.

      • David, just to clarify, we update all Kred profiles in real time that have activity that require an update.

        Clearly from one second to the next if nothing has happened, the raw and normalised scores will not change.

        When we review what comes in from Twitter, LinkedIn and Facebook, for those profiles that have an influence transaction, then we calculate the new Kred score on the spot.

        You can see this on your activity statement.

        Head to http://kred.com/daveg and log in then click on the panel at the top right that says “total influence points” and you will see the live update of your raw and normalised Kred score.

        As I type, you have 55,289 points raw influence points, and 20,783 raw outreach points.

  • Another example, besides Hemingway, is Orson Welles. For every great film he made, he also had 3 or 4 horrendous projects. Then again, he was unabashed in his defense of this, saying he had to do the horrible projects because they made him money, and he could use that money to make the horrible films.

    Personally, I think things like Klout and Kred will come and go, but I’m not sure that we’ll ever have a measure that is a predictor. There are so many other elements of influence that these algorithms can’t measure.

    • Yes, it’s an unfortunate circumstance. Their continuing use, that is, by the business community, but I did like @andrewgrill:disqus’s response, and pragmatic view.

      I do liken them to a baseball statistic, they show a very small part of the game. We need to consider them in that context.

      • Funny, I just now checked Klout and got a perk. Apparently I’m an influencer in iPhones and I get a nice new case for my iPhone 5. The one I don’t have. Oh, and did I mention I generally talk smack on Apple products? Yeah.

  • Then you have the connect.me and then you have an about.me to wed through to see who has vouched for you. I like to keep connections going on one of the main tools,twitter. But when I really need to focus my attention away from the white noise, i simply make a phone call to get down to the specifics.

    • Great point: I agree and think that real influencers meet in person. That’s how business is done!

  • I sign up for virtually anything that offers rewards for “influencers” because I like getting free stuff but on a professional level I am very skeptical about these things.

    For example, I convinced a friend who does nothing online to purchase a particular cellphone and car. There is no way to measure that and I know there are infinite number of similar examples.

    There is just no way to take into account the face-to-face and telephone calls that are part of influence so we speculate, extrapolate and guess.

    • Good point! Or the private conversations that are occurring online behind firewalls and on email. Real dialogue often extends well beyond the public socnets.

  • This thread is really getting interesting. For my tuppence worth ..

    It is easy to get bogged down in trying to define “influence”. Whether it is “social capital” or “influence”, I think it is better just to try to recognise and integrate as many of the moving parts of the equation as possible, and in that regard I think Andrew is absolutely right.

    Like it or not, this type of analysis is definitely here to stay. Klout and Kred are the pioneers, so there will inevitably be flaws, but with greater sophistication, I would imagine they will improve relatively quickly.

    Ref Colin Walker’s essay, you should definitely research network science as you are getting into it, perhaps unintentionally it would appear. I am currently reading Mark Schaefer’s book, Return on Influence, and whilst it contains accepted, unarguable and well-founded reasoning, the network impact is missing, as it appears to be from most current analysis. I am not an expert, but I am pals with a brain-box who is, and I do believe it has the potential to take the existing activity/engagement model to a deeper level. It also enables visibility of “influence” across the network of a vertical, not just isolated people within it. Geoff – sounds like you know a bit about complex networks? Or Andrew, any thoughts in this area?

    • Yeah, the theory is science of networks, which details the larger academic community’s study of influence in off and online social networks. This school of thought is certainly less popular than online theories discussed by bloggers, but in my mind it holds a lot of weight.

      Schaeffer’s book is based on some research by Robert Cialdini, which is primarily sales based, and is very good. I think the book can help you become more popular online!

      • I am reading up on network science and initial impressions are ringing a number of bells. It is also easy to see how Facebook has made a lot of decisions about how the network functions based on the general principles.

        • Agreed, I actually think Facebook leverages network science to a fault. It’s why people get addicted to it.

  • I want to know which 3 or 4 Hemingway books were the good ones?

    I’ve only read “A Farewell to Arms” and it was dreadful. I’m willing to give E.H. another chance, but only one more. He needs to NOT suck in this next book. So, please tell me which ones are the good ones.

    • Hahahahah, I loved Farewell to Arms. Try the Sun Also Rises or For Whom the Bell Tolls. I enjoyed both immensely.

  • Excellent article as usual. Social scoring is a fascinating topic to me because of its implications and lack of precedent. Regular people have never before had the opportunity to be scored. Outside of school and possible work evaluations, we’ve never had scores, especially one or two numbers that were calculated using these services.

    The closest we’ve ever had to having a number score assigned to us would probably be income. Those with high incomes are generally considered to be more “successful”, just like those with higher Kred or Klout are more “influential.” Yet income was never open to the public like this.

    So Klout and Kred have absolutely no precedent in their mission, mostly because it was impossible to measure someone’s life in any meaningful way, especially before the Internet and offline.

    What this means, to me at least, is that it’s not only unfair to say that Klout and Kred are failing, and possibly inaccurate. To fail, there must be some form of success. This may seem like petty semantics, but semantics here play an important role.

    Yes Kred and Klout are failing at measuring influence because our definition of influence is tied to an offline world in which it was impossible to measure it (because we have no means of quantifying individual influences on every other individual). Going back to income, it’s like developing a system to determine who’s richer without having money.

    How Kred and Klout can succeed, I think, is by dropping their close tie to the very loaded word “influence.” How about we accept that Klout measures your Klout, and Kred measures your Kred. Yes they are artificially created metrics, but if we accept them for what they are in their own context then, and stop trying to pigeon-hole them into the definition of “influence” a lot more people would be happier.

    Sorry if the post was all over, it’s late :)

    • No, I don’t think you were all over the place. I agree with you, I think separation from the word influence is necessary for long-term viability, at least without a constant argument for viability.

      I do think there is a precedent, however, and that is in the world of sports, where every action, positive and negative seems to be weighed with a statistic. I liken Klout and Kred to Twitter or public socnet update statistics, and not much more.

      They are far from comprehensive, and lack so many intangibles that they cannot be considered accurate, but certainly the first precursors of statistical evaluation of online behavior. Yay. Just what most of us did not want.

    • Great post and great reply @PavelNovel:disqus – I think social seo & social stats will continue to be a growing topic for 2013. No one has it right yet but many folks really can’t live without some sort of score or reporting system.

  • I think the biggest problem with these platforms is, as you say, they are “rewarding attention rather than real acts of influence.”

    I find these platforms pretty ridiculous, myself. But, unfortunately, I can’t fully ignore them because it looks like they’re not going anywhere…at least not for a while.

    And while I may not take them seriously, a lot of people do.

    In fact, my friend’s girlfriend’s dad (follow that one?) is obsessed with Klout. And checks MY Klout score daily – which is way more often than I check it myself :).

    It definitely creates a sense of competition – a game. And people love games.

    • It’s funny, the conversations we have at one of my clients about this particular topic. We really laugh at so called influencers and self proclaimed experts, and of course I get hazed as one of them because of the blog. Deservedly so. We all know where these people really sit in life.

  • first, I need to watch you TEDx talk :) Second, this is a terrific post and the discussion below is outstanding. The exchange with @andrewgrill:disqus is excellent and I recommend everyone to read. Determining the influential people in the different fields is something that is usually best determined by those working in the same area. This is presumably the ‘human’ touch that Andrew mentioned in his comment. I don’t have an answer on how to do this? I know there are new companies such as LittleBird that are trying to do such things but they are still in the early stages. I work for a company, Engagio, that believes that conversation is a strong social gesture and is an interesting filter to discover interesting sites and people. Perhaps, the level of engagement in commenting/conversations could be used in the future. Finally, I am intrigued by Kred after reading this post and will check them out!!

    • So what we are really talking about are new media voices. I don’t know. I think the term influence is misappropriated. I think Engagio is right that people serve as information outlets, but that’s not actionable social networking and grassroots activity. It’s unfortunate that we’re seeing so much weight tied to being a loud mouth.

      • I agree with your point about the notion of influence and how everyone is influential in some way. The social web has created a lot of ‘loud mouth’ and Saturday Night Live had a funny sketch last week comparing how the different political viewpoints on the social web are weighted (i.e a respected reporter’s tweet versus a random person tweeting Obama = Poop). Engagio might not have the solution for determining an ‘influence’ metric but at a personal level it has allowed me to focus on the right conversations and in the process identify people that have become influential to me. The key word here is “me” as I think this is a personal tool that people may use to tailor their networks and determine their influencers based on their custom filters. Still, having a global metric that helps determine ‘influential’ people would be very useful but is still elusive!

  • Finally getting to read through this. Good intro/context for @Shonali:disqus … I still stand by my chop busting though. I’d also argue that In our Time was an excellent collection from Hemmingway and the Old Man and the Sea is my all-tie favorite read.

  • I agree with many of your points but think you are rolling old tapes in some cases and applying offline paradigms to an online world where they are not necessarily relevant. Also, you are not considering significant progress in the field. Have you checked out Appinions?

    What Klout and Kred do is show a relative ability to create content that moves and creates a reaction. It’s really not that complicated but to a large extent, that is my source of influence in the online world, and yours too. If you did not posses this ability, it is likely I never wold have heard of you. This ability is also demonstrated by the reaction you have created in me now and the reaction you create consistently over time. To the extent that a formula can measure your ability to do this (and ONLY this) it would be an indicator of this one small (but important) sliver of influence. The math is still in the silent movie stages. But it does indicate something increasingly useful.

    Michael Arrington, Kleiner-Perkins and Microsoft have all invested in Klout. Despite the company’s limited scope and PR blunders, they probably see some potential there, right?. Neither Klout nor Kred will ever measure all influence, but like a credit score, it provides a short-cut indicator of something — namely an ability to create buzz. And while that may not live up to the all-encompassing, traditional and academic views of influence you mention, it is certainly not a “fail.” In fact, it is an achievement.

    • I think this is the crux of our disagreement about influence. What you call influence is what I call a measure of someones ability to create attention, and in most cases, that is a content creator. So high scores go to bloggers, photographers like Thomas Hawke, etc.

      Reactions to new media shares are not influence, in my opinion. That’s someone sharing a story. What content creators represent are new trade press covering niche topics throughout the long tail, and in the best cases, the head of the tail.

      So these things represent a statistical PR measure, not a true measure of somebody’s ability to trigger societal action, sales, or any other meaningful community action. They literally reward the loudest mouths on the Internet, folks like you and me.

      In my extensive work fundraising, I have seen over and over again, new media outlets — personalities and influencers in particular — rarely trigger action. Only when they have a very engaged community, an authentic tie to the cause, and the willingness to spend that social capital do they succeed.

      Much more frequently, those that do fundraise well are not “influencers” in a Klout and Kred sense, but are so in the real world. They do possess those three elements — engaged peers, authenticity and willingness — and successfully work their not so public and win.

      This is grassroots influence, and if you need an example, look no further than the Obama Campaign’s grassroots strategy for key counties in battleground states. The same formula was at work.

      • I completely agree that it is exceedingly difficult to push these weak online connections to offline actions and I discuss that extensively in my book.
        However, weak reactions like clicking a link, commenting on a post or eliciting tweets are also actions that may indicate relative influence.
        Further, the algorithms are progressing quickly to link online conversations to offline behavior. For example, a trusted music authority’s review may result in a wave of record purchases recorded on a Facebook timeline or any number of sites.
        Not only is this influence, companies are assigning a dollar value to it. This has vast implications for both for-profit and non-profit organizations. Just the ability to find passionate advocates who are willing to love and share our content is an amazing advantage provided by these companies that we did not have just a few years ago.
        You do influence me Geoff through your content alone as much as you might resist that idea! You may not be able to get me to buy something, but you do impact my thoughts and actions, which is quite evident today. In fact I am taking time to respond to you instead of engaging in a business or family activity. That is legitimate influence.
        Thanks for the great discussion.


        Please Pardon iPhone-related typos

        • I think this whole conversation needs to be reframed beyond the influence argument for new media voices. This seems to me to be a great overvaluation of the impact of an online marketing conversation.

          How does this influence dialogue mean anything to society when it needs change? How does this create an effective profitability plan for a company? How will this change a family suffering from unemployment?
          Influence or attention, what matters? Why are we discussing this? What’s the value for our culture?

          While I appreciate your gestures in this comment, I care very little about what bloggers think of their own impact. I do care about these larger questions.

          • Well in that case, if you expect Klout to help families suffering from unemployment then you’re right. It’s a failure.

          • I expect experts to talk about dynamics beyond individual blogger power, and get into wider more important topics about how influence impacts communities and social networks. I don’t think that’s too much to ask, Mark.

  • Pingback:Epic names, Fear Tactics and the F Word | Exploring Conversational Media

    […] Why Klout and Kred Fail – are numbers too big to fail? Sometimes, especially when it comes to social scoring. […]

  • Hey Geoff, this comes from a guy that is not in the “influence” game, per se.

    I have read your piece and many of the comments here. I agree that in a grass roots way true influence is akin to a relationship that you develop over time and can affect by what you do but influence is also like the guy (like me) you don’t know well who reads what you do, listens to what you say and then makes up his own mind whether you influence the decisions he makes to move in certain directions. That, to me, is influence.

    Unfortunately there is no real way to measure that but the fact that i come here consistently and read, RT your tweets, share stuff on FB and on and on certainly means that the measurement of that activity is valuable to you. Sure, bacon and Kim K expose the flaws of the current system of measurement but there has to be some form of relevance to the activity associated with your contribution besides personal connection, doesn’t there (I got that conclusion from your video)?

    It’s interesting to see so much debate about the tools when the discussion could turn to how to change the tools to better measure activity. Just like anything in life, we need to put a box around it so we have benchmarks otherwise there is no relevance to the data (I think @Shonali:disqus said something like that. Sorry Shanoli if I misunderstood).

    Cheers Geoff. Thanks for writing this. It helps me understand the on-line world a bit better.

    • Sure, and thank you for the influence tip. But there’s a difference between engaging in action and intellectualism. I’m as influential as Consumer Reports to a very small group of people on a very small subject area.

      Even when I do give you an idea that you consider, you weigh it, you just don’t act, and if it is a major decision, I’ll wager that you check with your peers, your real influencers, before you take action. You check to see what they think, what their experience is, what they think of the product idea, before enacting.

      So, I think a big reality check is needed in this sense. I would say the same thing to @markwilliamschaefer:disqus in response to his rebuttal. Let’s not overvalue intellectuals who get credit for presenting or representing ideas, and look at how real decisions are made and what spawns action.

      I don’t want to dismiss RTs and the like as non action, because they are certainly precursors to substantive movements, and in aggregate can create movements on their own. But I am chiefly concerned with influence on how societies evolve, how ideas are adopted by markets en masse, not the power of a new media voice. We need to take a big picture view of this.

  • Hey, I got to be influential about bacon at some point, too! But wait for it… my favorite one… Klout once told me I was influential in “pear”. Social scoring is the new achievement test. You know how when you’re in school the teacher “teaches to the test”? And everyone complains that there’s no time for gym and art because everyone has to learn how to answer the questions with the little bubbles correctly. But classrooms and schools with good numbers make people happy. This is the test bubble for the internet generation. It’s short-sighted either way but people latch onto it because it’s measurable and everyone wants to be better than someone else. My number is bigger than yours. Maybe the best we can hope to do is play the game.

    • Truly spot on. We can’t escape these metrics so we are stuck in a short sighted game. While we may have to deal with them, in some form. Now it’s time to demand more meaningful statistics that can actually help us do our job.

  • wow – great thread. Here’s my take, which addresses points made in several comments, so I’ll post it as a new comment vs. a reply:

    1. What you measure is what you get: Tools like Klout and Kred measure the result of someone’s online activity – (both also include different forms of “offline” influence measurement, but this is even less scientific and just confuses the issues further, so let’s leave it aside for now) – specifically their ability to trigger digital activity in other online users.

    2. Ask the right question: These debates get bogged down in the definition of “influence,” and whether something like a Kred score really captures true influence. I think that’s the wrong discussion to be having. Instead, we should be focused on understanding what these tools really are (and are not) measuring, what the value of that is, and how to best incorporate these tools into the overall tool kit.

    3. What is it good for? The emerging science of influence measurement will play a critical role as the balance of power in media continues to shift to where earned/shared media is worth as much if not more than traditional broadcast/paid media.

    4. Networks matter. We should be paying far more attention to how networks of influence are formed and behave, than just to how individuals behave. Yes, this is the science of network stuff referenced in other comments.

    5. Context matters. There are more sophisticated tools, such as Appinions, which Mark W Schaefer mentions in his comment, that measure not only your own personal activity, but look at how conversations and opinions are shaped around specific topics, and who is shaping them — even if the who is someone who’s not talking online, but simply being referenced by others. Cool stuff, and it delivers far richer insights than a single score or number that is very hard to interpret.

    I could go on — but my point is this: these conversations need to start getting more sophisticated & move beyond debating whether the tools measure “influence” and into understanding what can (and should) be measured – and how that data can be used.

    • Thanks for a well thought out comment, Tonia. I guess my response is to have a more sophisticated conversation about influence, we have to agree on what it is, and we don’t.

      To my point, your third summation says, “The emerging science of influence measurement will play a critical role
      as the balance of power in media continues to shift to where
      earned/shared media is worth as much if not more than traditional broadcast/paid media.” Again, we are talking media attention and not real actions that people take.

      That’s the crux of the issue. I would debate against any such media conversation being about influence. It’s not. You’re talking reach with stakeholders.

      So while I appreciate your depth of thought, you have to handle the horse before you address the cart.

      • So what else should we call it? I make a point of always at least saying “measures of online influence” vs plain “influence.”

        Traditional media measurements focused on reach, or some kind of direct “action” (opens, click-throughs). Klout measures your (retro-active) ability to generate online actions in other users. What would you call that?

        • So I see what you are saying here, and in the civic actions pace, these are often sneered at a slacktivist actions. So I guess you can call it that, but it’s not the action that companies and nonprofits often seek en masse. What companies and businesses want when they seek to influence people are leads and transactions. That’s what I’m talking about.

          What media and bloggers talk about with influence and what we are debating is a retweet ratio, which shows you you how far your post goes. Some metrics weigh the retweeter’s own network in this sense. So that’s reach, in my book. You can measure strictly the number of retweets as action, but as a marketer who is tied to measurable results beyond retweets, I can’t accept it as a viable marketing outcome other than the success of a post in garnering attention.

          • and that’s why any marketer who measures only reach, only retweets, or only Klout scores is a slacktivist marketer!
            But it doesn’t make those metrics invalid. It’s how they’re being used that’s the big fail.

          • Yeah, I agree with you. In fact, the statistical and contextual use of data to inform decisions can only help them. I also don’t want to devalue the branding and WOM value that bloggers/influencers bring.

            So the influencer label aside, we’d have a fantastic conversation about which additional statistics we’d need to really effectively manage campaigns and movements well.

        • Oh an yes, I realize that the algorithms take much more in consideration beyond retweet ratios. My last comment oversimplifies the issue.

          And thank you for the strong comments!

  • I love when you write super smart posts like this because I usually just say “Klout is poop and if you like it, you like poop,” and then I go on and talk about something else that’s even less interesting.

    Now, I can do that, but also point them towards this post.

    I like Kred guy. I’m still not overly fond of the whole “influence” thing, but if we switched it to “digital reach, regardless of potential value of the payload of those words,” they’d have me nodding.

    Okay, back to poop and similar. : )

    • I agree, the influence term needs to be tossed. Then it becomes a valuable statistic. And great to see you. As someone who has repeated successes, I bet you could speak to the issue of replicating wins and how hard it is.

      I hear the G+ Book for Business 2nd Edition Hangout went well. Kudos!

  • Pingback:Why Klout and Kred Fail | Geoff Livingstons Blog | lizjukovsky

    […] just read Why Klout and Kred Fail on Geoff Livingstons Blog and  – I know I’m missing the point (that social scoring […]

  • Hi Geoff,

    I like the idea to come back to basis and to the definition of influence.

    One thing that we do is to differentiate: “maven” type of influence, versus “connector” type of influence.

    In another words, I got your article from Gini. Does she influence me or do you influence me ?

    With this view, we put of lot of stress on “content author” as the core influencer. By the same token we catagorize the content /and the influence by domain. So instead of a score what we provide is:

    – Mr Geoff is a “maven” type of influencer in the social media tribe with a lot of influence and a mild connector in the same tribe.

    Not as sexy as a “one data point score” but more useful.

    We’ve still a long way to go. However, using this approach it is less a prediction game: One does not become a maven influencer overnight. It takes -cf M Gladwell – 20 hours a week for 10 years to become an influencer.


  • My fear is people will have a tendency to game their Klout score. Gaming your page rank in Google can have, ostensibly, ROI. But gaming your Klout score does not actually increase your reach, increase click throughs, etc. It’s can be just a vanity score.

  • This tool as well as others (like dandyID, for instance), fail insofar as they assume that users only have one identity or profile per social networking interface. So, even though my “personal” Twitter account is new and small, I have been managing at least 10 other Twitter Accounts, 23 Facebook Pages, and now a growing number of G+, Linkedin, and Pinterest accounts and pages!

    There is no way to consider the complete picture, if we are not able to integrate every social media platform along with every page/profile we personally work on and manage. This should also provide a different metric. If one person puts their “all” into one platform, certainly their Klout and Kred show favorably, but when someone does a great job with multiple accounts, pages, and profiles, that should also be taken into consideration. A person like this might not have as high a number on any one of them, but they are high considering the number of managed accounts. Considering this, don’t you think that a person operating in this way is rather formidable as well, and ought to net a higher score? For instance, does the Kred site consider the number of pages managed including their metrics along with the personal profile newsfeed posts of synced Facebook accounts?

    In addition, if someone is managing an account with a large business which already has a high following, the numbers are increased arbitrarily as their respective posts and interactions are not necessarily tied to their personal reach/influence as much as it is the pre-existing fan base.

    In these ways, the Kred/Klout scores are skewed data which isn’t taking factors into play like when you start at zero and grow to a thousand within months, while you grow other pages to 500 or more at the same time within months. This in and of itself is an important metric that deserves inclusion.

    The metrics as they stand today should be taken with a grain of salt. It will take some serious tinkering with some mathematicians to devise a system that is inclusive of the aforementioned parameters, before these sites are taken more seriously, IMHO.

Comments are closed