“In the last decade, we have learned more about how the brain thinks and how consumers make decisions than in the previous century.”

Emotions, when captured and analyzed, offer a much wider and deeper platform for insights into customers—their behaviors and their contexts. Emotional intelligence is one of the fastest growing fields in market research, marketing and advertising. Discovering a gut reaction—how customers truly feel toward your brand’s message, a new product you’ve developed or an experience—can ensure success and ultimately return on the investment.

While there are many ways to uncover subconscious truths, in this post, we explore the Emotion Mining™ tool against ZMET. Both provide substantive insights and offer a lens for accessing the emotions at work in each of us; however, there are unique aspects of each that marketers should discern.


The Zaltman metaphor elicitation technique (ZMET) is a technique that elicits both conscious and subconscious thoughts by exploring people’s non-literal or metaphoric expressions. Developed at the Harvard Business School in the early 1990s, it utilizes imagery to reveal ideas and customers’ hidden metaphors.

Broadly described, ZMET study participants are asked to collect a set of pictures that represent their thoughts and feelings about a topic of interest. The pictures collected are used as non-literal devices for uncovering deeply held, often subconscious, thoughts and feelings.

The goal of ZMET analysis is to reveal relevant metaphors, typically one or two, that reflect people’s feelings about a topic.

Emotion Mining™

Emotion Mining provides a linguistic technique, relying on the words people use, instead of images. It works by uncovering the emotions hidden in our language and the words we choose when answering questions about a brand, product, experience, etc. The online version of the tool elicits language from respondents through four primary exercises:

  • Give five words that describe how you feel about the topic
  • Give a one-sentence explanation for each word – “Why this word?”
  • Gamification – Conscious intensity measured
  • Gamification – Subconscious intensity measured

The patented linguistic algorithms can be used via several other techniques, including variations of text analytics, social listening and qualitative research techniques.

Once the language is captured, the Emotion Mining lexicon, which includes over four thousand words, whittles the collection of words down to 32 Emotional Channels that fall into four basic channel properties: Enjoyment, Interest, Commitment and Passion.

The goal of Emotion Mining is to identify the language that best describes the customer’s emotions regarding a topic.

Comparing the Two

Both market research techniques bring to light conscious and subconscious thoughts and feelings. Each is based in science and utilizes principles of psychology. What’s different is how each method extracts and delivers insights.

Dimensions Emotion Mining ZMET
Focus · Addresses customers’ emotions through the language ·  Addresses customers’ emotions through images and metaphors
General layers of findings · Emotion Mining results in a database of findings

· Insights are provided in typical/familiar analytical formats (quadrant charts, histograms, NPS/CSAT metrics, etc.)

·  ZMET provides a more informal set of outputs culminating in one or two overarching metaphors

·  Findings are presented in myriad formats

Data collection · Integrates with traditional market research methodologies.  Online/cloud based exercises, text analytics, focus groups and in-depth interviews, social listening and web-scraping, etc.

· Easily replicable – ideal for tracking

· Has an automated element that offers greater speed and consistency

·  Uses “collages” and other image sources to evoke aspects of metaphors

·  Not an easily replicable process. Replicability presents inherent difficulties—will different researchers (interviewers) churn out a similar set of data?

·  Requires intensive “interviewer” training

Segment and cells · Can be small or large samples

· Systematic analysis across segments/cells that can be connected into other analysis tasks

·  Smaller sample sizes typically

·  Difficult to execute segmentation and defined comparisons due to nature of analysis and interpretation – not data driven

Analysis · Formal algorithms built into the analysis process – reduces interpretation bottleneck/filtering

· Capacity to do broad sets of analyses (due to form of its data and breadth/depth of its application)

· Controls for biases connected with other data gathering modes—personal biases, interviewer biases, environmental biases

·  Not “automated”

·  Dependent of the inferential skills of the analyzer (and the interviewer in the earlier stages of data gathering)

·  Absence of formalized database inhibits structured and systematic analysis

·  Generates some standard findings as part of the analysis process but typically these are less consistent in form and structure

While the two take different approaches to discovering subconscious thoughts, both are grounded in neuroscience and can lead executives to a more refined understanding of the behavioral economics that drive bottom lines. To decide which is best for your company, consider the differences.

Tags: , , , , , , ,

More News