Big Data

Does Big Data Mean Big Ethics?

on July 25, 2014

A few weeks ago, some of you may have heard about Facebook’s big user experiment, wherein they manipulated user posts to find out if exposure to emotions led people to change their own posting behavior. Although the results of the study were positive (people respond to positive emotions more so than negative ones) the study quickly became controversial.

Why? Because Facebook did not gain permission of its users before performing the experiment. Which many consider unethical. In this article, the author Mary Branscomb, says it is downright unethical and creepy, and could lead to even bigger issues down the line, such as people not willing to share data, and the loss of benchmarking capabilities and opportunities for better, smarter growth that Big Data provides.

Responsibility and ethics are at the core of all businesses including those that involve Big Data.  I think the idea of a “Big Data code of conduct” is really one that is wrapped into a bigger idea of business code of conduct and ethics.  Credit card companies have long had insights into individual behaviors and tendencies, and they’ve managed to not spook people with their invasiveness. Within the article Branscomb cites several research studies that used existing data, but quietly.  Are these studies better because they were quiet?  Or was the lack of public outcry based on the lack of public knowledge? 

What they do with those insights directly impacts the perception their customers have of these companies and whether they continue to do business with them.  While the market doesn’t like “invasive” data analysis and subsequent actions, it “hates” data breaches even more. Consumers have proven in the case of Target that they react more negatively to data breaches than they do data “invasions.”  Facebook users have a choice with respect to who can see their posts and even whether or not they use Facebook.  I suspect that only very few are familiar with the terms of use of their Facebook data.

Either way, the bottom line is that Big Data businesses need to set data use expectations with their customers and then abide by them. They need to meet their commits to protect their customers from risks associated with data. Unwarranted data manipulation only serves to remind consumers that they are being watched.  Some organizations do this legally by burying their data use guidelines in legalese and data use statements with the expectation that their customers will not read them.  Others update user agreements nearly four months after the research took place.

Our philosophy at Oversight is to be as transparent as we can. We limit the data that we acquire from our customers so that we only meet our data analysis objectives.  This limits our risks, our customers’ risks, and establishes transparency.  The perception of Big Data analytics is directly related to public’s perceptions of the organizations who use the data. Since transparency and ethics will always play a key role in forming those perceptions, it is best to always engage in ethical best practices.

spend-analysis-vol3