Our Obsession with Scoring is Now Taking on Individuals on the Web

I had not been paying too much attention to Klout until I saw this article in the New York Times, Got Twitter? You’ve Been Scored. I skimmed it and tweeted about it to save the link for further reading. I often use my Twitter feed, @billlives, as a social bookmarking tool so I remember and have access to posts and articles I want to save. I then go through my tweets twice a month and it gives me an overview of what I thought was important during this time. Then I do a blog post that lists the tweets I found really important as a way archive these tweets. I am sure there are more efficient ways to archive Twitter but I find this practice a useful exercise. I also use the @darwineco Twitter feed for the same reason when the topic is more relavent to Darwin Ecosystems with the added goal of sharing it with my teammates at Darwin.  Then I follow the same steps ending up with a post on the Darwin blog.

Well, the tweet on Got Twitter? You’ve Been Scored got retweeted six times within a few hours so this action caused me to look more closely at the article. I then tweeted that I was inspired by the RTs to write a blog post on the article and I got two replies for the six who said they were looking forward to the post. I was pleased as this type of exchange is how Twitter is supposed to work. I also met a few new Twitter friends in the process.

The article itself exposed me more to the growing movement behind such tools as Klout and Peerindex. It is about all of us getting a number, similar to Robert Parker’s quantification of wine. While there is more objectivity here that with Parker, the underlying motive is the same, the quantification of the world. This is a topic I have covered several times.  I did a four part series on this blog that began with Rising Above the Over Quantification of Content: Part One: Parker vs. Piaget.

In this post I quoted, Adam Gopnik in the September 6, 2004 issue of the New Yorker, “(Parker) was uncannily successful because (he was an) apostle of a radical American empiricism – an insistence that facts and numbers could show you what was really going on, against everything tradition told you…The debate is not about whether the numbers are right but whether it is right to have numbers.”

In a similar perspective, Nick Carr quotes Google CEO, Eric Schmidt, in his Atlantic article,Is Google Making Us Stupid?, that it is “a company that’s founded around the science of measurement,” and it is striving to “systematize everything” it does. Carr adds that what Fred Taylor did for the work of the hand, Google is doing for the work of the mind. I would add: and what Robert Parker did for wine.  Nick goes on to write, “in Google’s world, the world we enter when we go online, there’s little place for the fuzziness of contemplation. Ambiguity is not an opening for insight but a bug to be fixed.”

Personally, I think a little ambiguity and complexity is good. So how does this relate to someone’s reputation on the Web? Personally, I follow people for many different reasons so a single score is not going to help too much. I also follow people related to three different twitter accounts I am connected with (@billives, @darwineco, and @outstart). The fact that Klout does offer the top ten topics a person is known for helps but there is still a gap between the complexity of why I follow people and a score, in the same way I might like a wine for many different reasons, occasions, and food parings.

The New York Times article quotes, Jeremiah Owyang, an analyst I respect with Altimeter Group. “Using a single metric to evaluate influence is dangerous. He noted that Klout “lacks sentiment analysis” — The Times adds, “so a user who generates a lot of digital chatter might receive a high score even though what’s being said about the user is negative. Also, a single metric can be misleading: someone with little Twitter experience can snag a high score if they happen to post a video that goes viral.”

I look at the Klout scores of people I know and there is a rough correlation with my views of what works for me but certainly not a precise one. Like wine, I think this is an area best left to the full range of human cognitive abilities and the multiple associations the human mind can make.  The Klout score of someone I do not known might help except for the fact that I know it is highly likely not completely accurate for my needs, only a clue about what to pursue in more depth. This also does not consider the opportunity to game this score, just as people do with Google. For the record, I did check my scores on two of these tools and my Klout score is 55 and my Peerindex score is 59 but I am not really sure what this means and I do not know how to game it. 

At Darwin, we take the opposite view of Google and Klout. We view the Web as too massive to understand and human reasoning too complex and individualized to fully progam. Rather that drawing inspiration for Fred Taylor’s time and motion studies and attempting to understand and control the universe of content on the Web, Darwin is inspired by Edward Norton Lorenz’s work on Chaos Theory in mathematics and weather prediction. Darwin does not try to impose an order on content, it lets content self-organize. During the 1950s, Lorenz became skeptical of the appropriateness of the linear statistical models in meteorology, as most atmospheric phenomena involved in weather forecasting are non-linear. Instead he proposed the concept of attractors that operate within a dynamic system that evolves over time in a complex, non-repeating pattern.  This is the complete opposite of Fred Taylor’s repeatable time and motion approach.

Now I am not saying the Darwin approach is better than the Google or Klout approach, just different. We actually think the two approaches complement each other. We call our tool an awareness engine™ to distinguish it from a search engine. The Darwin approach does not require human intervention to be organize the content; it is self-organizing based on the content itself and therefore more representative of what is happening as it occurs, and respectful of each unique context. It also eliminates the possibilities of search optimization and spam. It visualizes the content in a way that allows you to sort through the complexity a but better and make your own decisions on what is most important and relevant to you. 

 

 

 

Comments

[John Maloney]

Hi –

This is factor analysis. It is mature and valuable.

http://en.wikipedia.org/wiki/Factor_analysis

People deal with complex systems all the time, e.g., "...within a dynamic system that evolves over time in a complex, non-repeating pattern" using integers.

Financial markets are complex. We depend on DJIA and FICO to navigate them.

Human physiology is complex. I monitor my blood pressure. Don’t know/care much about diastolic, systolic, sphygmomanometers or biochemistry. I do know hypertension is an indicator for a ton of problems like stroke, kidney failure, diabetes, etc.

Also, ‘gaming’ the system must not be used pejoratively! If you could game you social index, it could mean you are in a way better or worse. It simply depends on the gaming algorithms.

In an increasing complex world, gaming 'factors' must be seen as positive!

Well, I’m off to ‘game’ my blood pressure at the gym. Then I’ll come home to ‘game’ my FICO by paying some bills on time. Then I’ll ‘game’ the airline reservation aggregators to get a low-cost flight to LA…

Thing is, I don’t know how any of these complex systems work, but I do know how to achieve a certain mastery vis-à-vis factor analysis and gaming.

Cheers,

-j

[Rob Peters]

This article has been invaluable to me. The march toward automation of on-line interactions will continue. I agree that we cannot rely on just one score to measure a person, organization, or product/service's reputation.

We do need smarter metrics on a relationship entity's trustworthiness. The quality of a relationship will be scored based on made & kept-commitments and perceptions. A kept-commitment and perception is accounted for as Relationship Capital (RC). This will provide the sentiment analysis you mentioned. From a bigger vision perspective, we need open industry relationship standards.

It will be difficult for various relationship entities (person, product/service, or organization) to consistently grow "RC" if they do not practice time-tested principles such as Honesty, Accountability, Responsibility, Support, Respect, Trustworthiness, and Boundaries.

The Standard of Trust ecosystem believes that the one's with the highest quality of relationships on-line and off, will be the most successful in this social world!

Regards,
-Rob