How to Decipher Marketing’s Emotional Espionage

How to Decipher Marketing’s Emotional Espionage -

For eons, people have struggled to get in touch with their feelings … to understand what makes their precious selves tick. Marketers have been on the same quest for nearly as long. And now, in the highly-advanced 21st century, brand marketers finally might have cracked the code – might have the key to understanding the emotional import of consumer communications. Their tool for the touch is a process known quite aptly as sentiment analysis.

Sentiment analysis (aka opinion mining) is a method that enables researchers to gather subjective data from various forms of human communication. Essentially, it seeks to glean emotional meaning from strings of words. Are the communicators happy, sad, excited, bored, etc.? These, and all the other emotional states, are what the process seeks to uncover.


As with many other marketing trends, social media has done much to spur the interest in and pursuit of sentiment analysis. It’s a natural connection. Social sites such as Twitter, Facebook, and Instagram generate thousands of conversations daily. In the hands of brand marketers, sentiment analysis becomes a tool for gauging the emotional temperature of these conversations. Every message is a potential window to the true feelings that lie within. This transparency, it is assumed, will enable marketers to understand the real desires and attitudes floating around in consumer land.
Not surprisingly, marketers are combining sentiment analysis techniques with automated processes that collect consumer data and determine accurate responses. Researchers now are in command of sophisticated programs that evaluate consumer feelings toward a brand, identify any surge in negative messaging about the brand, and compare consumer feelings toward similar products or services.

Currently, several university research teams are using automated sentiment analysis to better understand social media communications. But the skies are anything but clear on that horizon. The problem is, people themselves often don’t comprehend the true feelings bundled into a given social media message. How then will an automated program pinpoint these elusive emotions?

Adding to the difficulties is the intrinsic complexity of a language such as English. A person might say “my supervising manager is excellent.” But if a message read “my supervising manager is excellent at making my life miserable”, sentiment analysis programs would be hard-pressed to judge the difference. That, apparently, is a mystery only humans will solve.

If you have any questions or comments about sentiment analysis in social media, or about any other brand-related topic, feel free to send them our way.