No Respect for Train Wrecks

The train wreck scene in Super 8 captivates you with its sheer level of destruction, power and unbridled fear. The scene is an awesome spectacle of sheer force and damage, one that you replay a couple of times to see which parts you missed. That doesn’t mean you want to hang out by the tracks for the next scheduled train wreck.

Yet isn’t that how some online personalities act online?

00695
Image by Grand Canyon NPS

A continuous train wreck of blog posts and social media updates detailing questionable acts and bad decisions definitely commands attention. Affairs, drunken debacles, bad business decisions, on and on. Kim Kardashian or Ozzy Osbourne imitations, the reality blogging is quite stunning. If the bumbling stumbling jalopy of voices keep it going for long enough, they may even command a significant online following. And why not? It’s entertaining (at least to some)!

However, garnering attention through a series of mishaps does not make a great marketer. On the contrary, it is simply a text version of reality TV.

Yet in the world of social media we like to anoint heroes based on follower counts and subscribership, one of the primary reasons why ROI is an elusive pursuit for many online practitioners.

In the end, cheap attention getting tactics don’t earn transactions. And that’s apparent when you look at data that examines conversion per follower with these folks. One chap boasts hundreds of thousands of followers, but can’t even raise $2000 in an online fundraiser.

The lesson: Discerning buyers don’t respect train wrecks. Neither should you.

How Social Semantic Search Defines People

BORKER-2-popup.jpeg
(Cartoon by David G. Klein from the New York Times)

Search is the underpinning of the Internet today, from the 1 billion traditional searches everyday on Google to providing references about a person on Twitter and delivering their stream feed on Facebook. Search has moved from simple page rank to an increasingly complex algorithm that weight’s social and semantic data points to deliver the outcomes most likely to please you. Personalization of search continues to evolve, but in turn it defines people and their choices.

Search — the technology itself — doesn’t bear responsibility for this. People do. People who use the Internet and its many free tools without understanding how the information is provided to them. They blindly accept search results or the search-based content feed without considering the source.

Consider the DecorMyEyes fiasco broken by the New York Times. Owner Vitaly Borker explained how he used intentionally created negative complaints about DecorMyEyes to game search results and place himself as a top ranked eyeglasses vendor. To Google’s credit, they promptly changed their algorithm to include more semantic weight (all negative or all positive disqualifying you), and the Department of Justice followed up with charges.

Social networks and applications also use search to source preferred content. Facebook’s activity feed is designed to source the most “interesting” content to people in your friends network are using the Open Graph API and likes. Search on Facebook is completely driven by the Open Graph (Like) protocol.

Of course, hashtags have demonstrated the power of search on Twitter. Twitter search was originally based on the acquired Summize search technology, and has been used to reference mentions and trends, too. Now Twitter (and other services) suggests people like you using semantic data.

The Danger of Homogeneous Definition

Google Organic.jpg

The danger in all of this personalized search — particularly when it’s largely based on peer interests — is creating a society of homogeneous sycophants that blindly accept the content sourced to them, either via search or feeds. Lest we think that people actually think through the click, consider organic click through rates on Google (as pictured above by SEO’s Neil Walker). Clicking through on the first few search terms is and has been the norm.

The addition of local semantic data to search only further complicates concepts of popularity. Algorithms tell people which burger joints, music venues, theaters, etc. are most likely to meet their interests.

When popularity is defined by an algorithm and served to people, homogeneos or mob thinking becomes the norm. This thinking feeds on the popular. Society is not currently trained to question the information presented to it. Thus algorithms — designed to create the output that will generate the most click throughs — become a critical determinant in defining people’s lives, and society as a whole.

That’s not necessarily a bad thing. Semantic information can weigh in when a system is gamed, and social search can provide the latest information based on people’s actual use and check-ins. However, idea markets are increasingly influenced by the popular, and not necessarily in a good way. Algorithms can keep bad ideas popular for longer periods of time.

It all points back to the need for society to teach better information skills. In an information economy, the ability to question and discern quality data presented via a plethora of media is an essential quality for democracy and individualism. It’s important to look deeper at online search, whether that’s because a search provided direct information or because an algorithm sourced a friend or influencer touting an idea or product. Quoting Doug Haslam, “Think for yourself. …you needn’t be part of some pack that can’t brook disagreement with your heroes.”

An educated Fifth Estate creates an evolutionary society, a mindless one creates results like Kim Kardashian as the number one search term on Bing for 2010. While many people find Kardashian attractive, should social semantic search tell every person — man and woman alike — what the icon of attractive is? Parents across America may object.

What do you think about how search and algorithms are defining our society?