News at Smith

Cleaning the Clutter

About the University of Maryland's Robert H. Smith School of Business

The Robert H. Smith School of Business is an internationally recognized leader in management education and research. One of 12 colleges and schools at the University of Maryland, College Park, the Smith School offers undergraduate, full-time and part-time MBA, executive MBA, online MBA, specialty master's, PhD and executive education programs, as well as outreach services to the corporate community. The school offers its degree, custom and certification programs in learning locations in North America and Asia.

Oct 15, 2014
World Class Faculty & Research


Wendy Moe


By Wendy W. Moe

A social media forum is a terrible thing to waste. Yet market researchers rarely know what to do with the consumer commentary that abounds online. Some companies overreact, altering strategy based on the rants of a few unhappy customers. Other firms err on the opposite side, ignoring the feedback altogether because of its limitations.

A third alternative emerges in a new paper I co-authored, using data from social media strategy and analytics firm Converseon. Companies can navigate the noise that dominates the Internet and extract timely, inexpensive and accurate data when they control for basic human tendencies that shape online conversations.

The key is to apply the same rigor of traditional social science in the relatively young realm of Twitter, Amazon, Yelp and other virtual gathering spots. To clean the clutter effectively, online market analysts need to control for three main human effects.

Extremity effect

Traditional social scientists understand that people like to complain. When consumers take the time to fill out a comment card, or when they talk to their friends and co-workers about a product, usually they are venting about negative experiences. Market researchers understand this bias and control for its effect.

Something similar happens in social media, but with a twist. Extremely negative people go online to talk, but so do extremely positive people. The majority of consumers, the ones in the middle, usually keep quiet. Social media marketers can control for the extremity effect by using mathematical models to gauge public sentiment based on predictable patterns of human behavior.

Venue effect

Besides the need to express strong opinions, people also have a fundamental yearning to belong. Traditional social scientists know that when people converse offline, they tend to say things that they think their audience wants to hear.

They take care not to alienate themselves from the mainstream or to polarize the group. This means a person might express different opinions at work than at church or some other context. Over time, people gravitate to venues where their views are mostly embraced.

This effect translates online, where like-minded commenters cluster in different communities. Often the result is an echo chamber, where isolated groups reinforce each other’s views and grow increasingly confident in their extreme positions.

Savvy social media marketers understand that the choice of where to express an opinion is a function of the opinion itself. They control for the venue effect by monitoring a range of discussions in multiple locations simultaneously. They don’t just get stuck in Twitter or Facebook.

Topic effect

A third effect that traditional social scientists have learned to manage involves people’s ability to distinguish the parts from the whole when forming opinions. A sports fan can hate a coach’s play calling but still love the team. A voter can dislike a candidate’s position on one political issue but still support the broader campaign.

Likewise in social media, people can express negative views on a particular topic but still have positive feelings toward the overall brand. Market researchers miss this nuance when they go online, gather all the comments related to a brand, rate the comments as positive or negative, and then draw general conclusions.

Companies that ignore the effects of human behavior and take social media comments at face value get mostly noise. The correlation to well-developed offline market research is almost zero (0.008 to be more exact). When companies control for known biases, the correlation to traditional data sources improves one hundredfold to over 0.8 — with 1.0 representing perfect alignment.

Social media then becomes a rich source of continuous data that companies can use in tandem with traditional research tools.

Wendy W. Moe, PhD, is a professor of marketing and director of the MS in Marketing Analytics at the University of Maryland’s Smith School of Business. 


“Social Media Intelligence: Measuring Brand Sentiment from Online Conversations” by Wendy W. Moe appears in the August 2014 issue of Manufacturing & Service Operations Management. The paper is the 2014 winner of the Robert D. Buzzell Marketing Science Institute Best Paper Award.