9 Comments

Summary:

Klout, the reputation-ranking service that recently confirmed a new round of funding, may not win the race to create a “PeopleRank” for the social web, but someone is going to do it — because the need to measure online influence is only going to increase.

5255901742_ea6350fb96_z

Klout, the online reputation-ranking service that recently raised a Series C round of financing estimated at $30 million from a host of venture capital firms, has taken some pretty heavy fire from critics over the past year. Some people dislike the way the company creates profiles on people without their permission, and others don’t appreciate being reduced to an arbitrary number that can change at any moment. But while Klout’s new round of funding doesn’t guarantee it will actually be able to build a business, the need for some kind of reputation-ranking system on the social web isn’t going away. If anything, it’s becoming more important with every passing day.

There have been rumors for some time that Klout was working on new funding, and the company confirmed on Tuesday it has closed a “strong” round led by Kleiner Perkins Caulfield Byers, along with Venrock and Institutional Venture Partners. The San Francisco-based startup hasn’t confirmed exactly how much money it raised, but sources have said it’s more than $30 million, which would give the company a market value of about $200 million. That’s not bad for a company CEO Joe Fernandez started in 2007 after spending months with his jaw wired shut following surgery.

Does the social web need a way to keep score?

Obviously, Klout still has to prove that it can generate revenue from its attempt to build a reputation graph, or what Fernandez has called a “PeopleRank” system similar to Google’s PageRank  for web pages. The company’s financial backers clearly believe it is (or could become) the leader in such an effort, as Venrock partners David Pakman and Marissa Campise noted in separate blog posts about their new investment. Campise said she believes Klout will be the “global standard” for identifying and measuring influence on social networks, and Pakman compared what the company does to media measurement services like Nielsen and comScore:

In every other mass media, measurement provides a benefit to the advertisers who subsidize that media… Klout has the benefit of being able to measure actual data, not inferred data. They aim to score the entire social web.

Keeping score is one of the things that seems to set many critics of Klout off. As Fernandez has noted in a number of interviews, people don’t like the feeling they are being marked on their social behavior — and in some cases getting what appears to be a failing grade. Others have complained that Klout creates “shadow profiles” for users based on their Twitter activity, regardless of whether they have actually signed up for the service or not. Author Charles Stross wrote a long post arguing this approach is fundamentally “evil,” and connecting to Klout is the “internet equivalent of herpes,” since anyone who uses it risks infecting their friends.

Other critics have focused on the arbitrariness of Klout’s measuring system, which the company recently tweaked in a way that caused most scores on the service (including Fernandez’s own score) to fall. Although Klout defended its changes by saying that it was trying to make them more reliable — and less susceptible to “gaming” attempts aimed at boosting scores through fake social activity — many complained since they didn’t know what Klout’s algorithm was based on, it wasn’t worth paying any attention to. The company’s defenders, meanwhile, argue that no one really knows exactly how Google’s algorithms work either, but that doesn’t stop them from using the service.

A reputation-based economy needs a measurement system

Sequoia Capital partner Roelof Botha recently noted, in response to a question from Hunch founder Chris Dixon, that services like Etsy and Kickstarter — which some have argued are the foundation of a “sharing economy” — are built on a web of trust, in which users are willing to engage in creating or buying products from others in part because they know (or can find out) enough about them to form such a relationship. This reputation-based economy requires some kind of measurement system, and possibly many competing systems, just as the traditional media market needs services like Nielsen and comScore.

We may not like the idea that we’re being measured, but it’s happening regardless — and Fernandez argues Klout is actually an improvement on some of the behind-the-scenes reputation ranking other services engage in, since it’s more open about the process. Klout also provides rewards or “perks” from its advertising partners to users based on their scores, such as discounts on airplane flights or the chance to test drive new cars, and this is a big part of how it generates revenue.

Whether Klout can create a “PageRank” for the people-based social web that accomplishes what Google’s PageRank did for the early web isn’t clear. There are dozens of competitors aimed at the same goal, including PeerIndex and Kred, and both Twitter and Facebook have their own internal reputation-ranking systems that could easily be externalized (some have argued eBay could do this as well). But there’s no question that someone is going to do it, because the social web requires it in order to function properly. As users, the most we can hope for is that the process is relatively obvious and that we get some benefit from allowing ourselves to be tracked.

Post and thumbnail photos courtesy of Flickr users Danny Cain and Andres Rodriguez

  1. I have a modestly good Klout score (55) but I’m very sad to say I even know it. As a twitter addict I know from experience that influence cannot be measured by RT’s and mentions. Its like measuring an ice-burg by its visible tip–a very poor and meaningless proxy.

    Share
    1. Thanks for the comment, James — I agree that RTs and mentions are the tip of the iceberg, but I think a reputation or influence-ranking system has to start somewhere, just as Google started with just links.

      Share
  2. Hello Mathew, great post and Happy New Year. You always seem to bring in social science into the discussion, of which I’m a big fan.

    There are a couple of things I think those charting the course for this space should consider, 1) Klout is not in the reputation graph game….that’s for the ProSkore’s of the world. And 2) PeopleRank is the term I gave the coming “social consumer hierarchy” in January 2008 before Klout’s launch: http://www.briansolis.com/2008/01/value-of-online-conversations/

    I have a report coming out on this very subject shortly…I believe Klout attempts to track both a digital form of social capital and through it’s Perks service, can also pinpoint a tiered interest graph, which only one other company can do well. What we can also use, is a service, or perhaps this is something that Klout can work on, is defining a true measure of influence…the cause, effect, or change that happens because of an individual in social networks. That’s what influence is…the ability to cause effect or change behavior. It would be interesting to also see to what extent someone contributes to social proof.

    It’s interesting, debatable, but more importantly, I hope all services providers here look beyond the score to find ways to deliver value beyond initial narcissism.

    #crosspost to Google+ post

    Share
    1. I agree, Brian. Thanks for the comment — and the compliments :-)

      Share
  3. Influence != Reputation !=Trust

    Share
    1. I agree they are not equivalent, Liad — but clearly they are related to each other, no? Any algorithm that tries to measure one of them will by necessity have to consider its relationship to the others.

      Share
      1. I don’t think that the “Social” web needs a score. Introducing a score leads to a sort of Observer effect, where the attempt at quantifying the values is directly influenced by the end score. It inevitably becomes a self fulfilling prophecy.

        As an interesting thought experiment, pretend for a moment that a person with a high level of “influence” were to receive a low “score”. His influence may then be undermined by those who base their opinion on score, or his score would be ignored as he already has sufficient influence over his listeners. Next, take a nobody in the social world and give them a high “score”. Suddenly, throngs of people flock to this “influential” person and hang on his every tweet, check in, and like. Ultimately, people would leave this impostor influencer, as he doesn’t actually have anything of substance to share.

        In the end, if this “score” were to disappear, influential people would still be paid attention to and non-influential people would be still be ignored. This system of quantifying a value can only lead to non-influential people gaming the system to achieve artificial influence. This system doesn’t benefit those who already have influence, which are the people the system actually needs to become a trusted reference for influence.

        Share
  4. Great post Matthew and echoing Brian, nice to begin the year with a little white hot social scientific debate.

    I think part of the issue is that currently users see the score as only of benefit to marketeers, it offers little value to they themselves (bragging rights aside!). I believe the reputation graph is here to stay, but that in the social enterprise realm at least, we’re measuring the wrong thing.

    It’s about relevance rather than reputation. That’s the issue we’re focused on. How is an opportunity relevant to me? Whenever a score is generated, be it Klout, PeerIndex, Kred or others, I need a little instant gratification – the score needs to work for me, give me something, it needs context.

    The reputation graph is mired in influence when all it needs is relevance.

    Share
  5. I’m not an icebergologist, but I’d imagine one could extrapolate rather accurately from the size of an iceberg’s visible tip to the size of the whole.

    Share

Comments have been disabled for this post