I’d like to influence you not to worry too much about online influence gamified metrics like Klout

Note from the legal department: This is a parody of the iconic Peter Steiner cartoon that appeared in the New Yorker in 1993. It can be purchased here: http://www.condenaststore.com/-sp/On-the-Internet-nobody-knows-you-re-a-dog-New-Yorker-Cartoon-Prints_i8562841_.htm

There’s a story by Anita Wadhwani appearing today on Tennessean.com regarding the efforts of a company called Klout to create a measurement of someone’s influence online about a particular topic. As the company is backed by the influential (in ways more important than may be indicated by its Klout score of 50) VC firm Kleiner Perkins, the company Klout has been gaining momentum (recognition? influence? — Klout has a Klout score of 86) in the past six months or so. With its success, the company has drawn criticism for a wide range of reasons, most of which are covered in the this October Gigaom piece by Mathew Ingram, so I’ll skipped all the “Klout is evil” stuff in this post.

Anita did a great job capturing the essence of what the company is all about, and added a local angle to the story by analyzing some people in Nashville who, for whatever reason, are liked by Klout’s algorithms. In the story, I got to play the role of the “humble bragger” (see: @humblebrag), wherein I say, in essence, “Klout is a joke — unless you’re talking about my score.”

There’s a reason for my ambivalence towards a service like Klout (beyond the humble bragging one). I completely understand why people — especially marketers — want a measurement, or even a currency, of influence to exist. The topic of such an online social currency has been around for as long as online communities have existed and was popularized in a very geekish and entertaining sci-fi way by Cory Doctorow in his novel Down and Out in the Magic Kingdom (free download), a book set in a future in which there is a “post-scarcity economy” based on a reputational currency called Whuffie. In the past, I’ve described Whuffie as similar to the role reputation management plays for those who buy and sell on eBay. On eBay, your reputation (or let’s call it a marketing term, “brand”) has a direct impact on your ability to sell something at all — or to sell it at the highest possible price. (For example, you’ll pay a premium three days before Christmas to purchase from a seller who has a 100% ranking and lots of comments about how they ship things when they promise to.)

However, unlike your reputation on eBay, a Klout score is, at best, some not-ready-for-prime-time Whuffie for many reasons. First off, Klout depends on algorithms that analyze expressions of influence, not measurements of actual influence. Expressions are, at best, a proxy of influence. No doubt, from a business-model standpoint, “expressions” are extremely valued by marketers. Google has become one of the most valuable companies on the planet by making search recommendations based on expressions (clicks and links) and not necessarily transactions. Nearly every measurement marketers depend on (TV ratings, for example) are measurements of proxies of influence — how many people may be watching a TV show may measure something, but it doesn’t measure how effective an specific ad is in making me want to purchase a product. Likewise, the circular nature of such proxies of influence — the more Google points to your content and generates clicks (expressions), the more expressions of influence you generate, influencing the algorithms of Google even more — is a challenge that will also impact Klout scores and those who will no-doubt want to game them.

(Note: There is an exception to that generalized observation of Google advertising: Some advertisers, online direct marketers for example, do know when a specific Google ad works and can measure precisely the return-on-investment of each ad they place — for them, the use of Google Adwords is a near-scientific practice akin to arbitrage. You know who you are, if this is the way you use Adwords.)

Let me try to simplify this: Buying something on eBay and then rating and reviewing the quality of the transaction and your opinion of the seller (or the buyer), is far more than a proxy measure of influence. It is currency that can add measurable value at the point of transaction — or, as we marketers like to call it, “conversion” (as in, conversion from freemium tire-kicker to subscriber). A click or Retweet or follow or friending are expressions of popularity and familiarity that, no doubt, represent something that might be termed “influence.” But a purchase that is then translated into a metric that captures both quantitative and qualitative data could be a source of influence that could serve both buyer and seller alike in ways only imagined today (or long-ago, in Cory Doctorow books).

Another challenge Klout faces is one that Google and its competitors must address daily. The algorithm that measures influence for you may not be the algorithm that measures influence for me. Consider the Google results page of 2011 (for a logged-in user) compared to the Google results page of ten years ago, the point about where Klout might be today, by comparison. Today, it is nearly impossible for you and me to search the same terms and receive the same Google results because the company’s algorithms take into consideration information I have provided it through my active request or passively permitted by my acceptance of their terms of usage.

Klout recently threw out an early generation algorithm that was terribly flawed. I know it was flawed because I was the beneficiary of a flaw in it. No matter how much flack they received by scrapping the old algorithm, I can tell you that it’s far closer to something half-way indicative of a certain type of influence today than it was before — but it’s still not close to what real influence is. But, like Google proves, algorithms are a long-distance journey.

Anita’s article points out another challenge a marketer might have in converting a Klout score into something of value. In Nashville, for example, there are a few high Klout scores that I would say are direct results of how actively the person plays good with social media (blogging for ten years and five years of tweeting goofy tweets that others re-tweet has been my strategy to influence Klout — except I was doing it long before Klout existed).

But other Nashville examples are more about the offline expertise and popularity of the individual, and how well they’ve integrated social media into their relationship with those who know them as, well, Taylor Swift or an expert in something related to church or parenting or a topic like cattle husbandry.

Pure online influence, or offline influence that has been converted to online influence — they’re both valid forms of influence, I guess.

So, bottomline, the Klout score is, at the end of the day, about as valuable as some of the metrics it tries to replace: sheer numbers of followers or “likes” or whatever. Until you can measure actual transactions (conversions) or the roles of offline vs. online sources of influence, when you measure a big basket of metrics all based on different expressions of the same sorts of likes and follows and retweets, you’re going to find out that a high Klout score is directly correlated to how long, how helpful and how actively a person uses the full set of tools social media provides.

In the end, Klout is another company that borrows all those social media things people do online and packages them up into something a venture company might fund and marketers might buy. But all it is at the end of the day, is us: users of social media. We are, as my friend Dave Winer sometimes uses as a metaphor — playing in Klout’s hamster cage.

No more, no less.

If you want to play with Klout, go ahead. Just don’t pretend it’s real influence. (Oh, unless it’s mine.)


When working on her story, Anita emailed to ask me who I would suggest as influential Nashville social media users, I didn’t have to think more than a second to respond the following:

Dave Delaney is the glue that holds together lots of those of us who have geekish tendencies. I have no idea what his “klout score” is, but I’d
give him a 100 for the influence he has earned and uses on behalf of
the company he works for, Griffin Technology, and for social media in
Nashville, in general.

This past week marked the fifth anniversary Geek Breakfast, just one of the many Nashville geek community activities Dave has created or fostered. Geek Breakfasts are now monthly traditions in cities around the country and internationally.

Dave’s a great evangelist for his employer — and for Nashville.

Thanks, Dave. You’re the real deal with it comes to influencing me (as you’ll see in a post I’ll add to this blog on Monday.)