Ground Control to Major Tom: Dealing with Rogue, Biased, and Unstructured Data

As marketers, must we be a humble servant to Big Data? Or can we be a Command Center observing and maneuvering through the chaos of distributed, biased, and unstructured content? 

Every day, marketers are faced with a dearth of jumbled, complicated data from multiple sources. I’ve suggested the role of the modern Chief Marketing Officer has evolved into more of a Chief Content Officer—analyzing, producing, and managing multiple forms of content across multiple media in order to capture, predict and find some semblance of order around a brand’s perception.

If content is Major Tom, marketer’s are Ground Control.

Major Tom was a fictional astronaut, one of the many characters created by David Bowie in the late 60’s and early 70’s. He appears in his songs “Space Oddity,” “Ashes to Ashes,” and “Hallo Spaceboy.” Major Tom’s fictional storyline sees his departure from Earth as a success, with everything going according to plan, until he cuts off contact with Ground Control.

Think of Ground Control as you, a brand marketer. Which makes Major Tom your brand—floating without tether, across the Internet and all major media sources.

Can we, will we, bring Major Tom back? Probably not. But can we continue to track him, support and encourage good moves, send him supplies, and observe what he does next? Yes we can.

As marketer’s we need to deal with the reality of a wayward brand, and we need to find ways to deal with rogue, unstructured, and biased data.

Rogue Data

Let’s face it, our data has gone rogue. Content is distributed across media, and users are generating a greater and greater percentage of the content, and thus defining brand perceptions in the marketplace with little-to-no control from marketers. And remember, we encouraged it—eliciting consumers to “spread the word” and creating promotions around “user generated content.” Major Tom’s cut off contact with our brands, and now it is time to use Big Data mining to track his progress. We can’t rein rogue data back in, but at least we can track it, analyze it, and use it to predict future behavior.

Unstructured Data

It is widely accepted that perhaps 80% of data exists as “unstructured”—radio, television, video, emails, “comments” on Internet sites, all exist as unstructured content. Harnessing linguistic, auditory, and visual content and analyzing this data is critical when making decisions around audience segmentation, go-to-market plans, and messaging. Using analyzed unstructured data to producing predictive modeling is an incredibly powerful tool for marketers. A 2010 Economist Intelligence Unit survey found that less than 33 percent of companies are using predictive analytics, but 70 percent said they will in the next three years.

Biased Analysis

Most big data analysis tools require human interaction—people “train” the data to deliver content analysis. The more human interaction, the more potential for bias-- pre-conceived ideas or an ideological disposition, the more careful you must be in taking data analysis with a grain of salt. Bias is anything that can “contaminate” the picture you are trying to get around consumer behavior, attitudes and beliefs.

While there are more and more technological advances helping to eliminate human bias by using machine learning (a variant of AI or Artificial Intelligence) that can lower human bias, but in the meantime, it is best to give a critical eye to who is performing the analysis, and what might be their pre-conceived notions and intent be around results. If your agency is analyzing your data, you better filter it through the lens of what services they are offering. If an internal employee is analyzing—consider their level of experience, and their industry expertise. All may bias your results.

So, by embracing your brand’s Major Tom—with his rogue and unstructured data, and his biases analysis, Ground Control can get back in the game, and participate more powerfully with your brand’s perception in the marketplace.

This article originally appeared on the Digital Marketing 2.0 blog 

blog comments powered by Disqus

The Featured Five