Many brands have their ups and downs. What happens to a company’s brand reputation when its CEO makes a controversial political statement, for example? The best way to find out quickly is to see what people are saying about it on social media. New research from Maryland Smith’s Roland Rust uses an artificial intelligence-based text analysis tool to track how a brand’s reputation rises or falls, moment by moment, on brand-related events.
For the Journal of Marketing study, Rust, Distinguished University Professor and David Bruce Smith Chair in Marketing, worked with co-authors William Rand of North Carolina State University; Ming-Hui Huang of National Taiwan University; Andrew Stephen of the University of Oxford; Gillian Brooks of King’s College London; and Timur Chabuk, vice president of machine learning and advanced analytics at Perceptronics Solutions Inc.
All companies should be concerned with how consumers see them, say the researchers, and to get a read on what people think at any given time, social media provides a much better alternative to impractical and expensive consumer surveys.
The research team demonstrates how using AI-based text analysis of social media – in this case, Twitter – can monitor how much a brand’s reputation rises or falls over time. They merge the social media monitoring with Rust’s previous research (with co-authors Katherine Lemon and Valarie Zeithaml) on customer equity drivers to pinpoint exactly which dimensions of brand reputation are changing.
“It is one thing to know that brand reputation is improving or declining, but another thing entirely to figure out why,” write the researchers on the American Marketing Association’s website.
In the research, Rust and his co-authors develop the Brand Reputation Tracker, which implements Rust’s value-brand-relationship framework. The researchers used the tracker to mine social media text on Twitter.
The tracker uses 11 different criteria to measure things consumers say they care about, including a brand’s “coolness,” “goods quality,” “social responsibility,” and “trustworthiness.” Then it takes tweets and sorts them into three scores for a brand: one based on whether stakeholders think a brand is a good value; one that assesses how closely stakeholders identify with the brand; and a third, called the brand driver score, that accounts for everything else, such as style and popularity.
With their new tool, the researchers looked at 100 popular brands. They measured and compared brands’ reputations with each other. The tracker allowed them to identify changes in individual brand’s reputations in real time, as Tweets were updated constantly. They were able to track a particular brand’s trend data over time and compare its reputation assessment with how well that brand’s publicly traded stock fared. Changes in value, relationship and brand driver scores were reflected in each brand’s stock valuation.
Rust and his co-authors say companies could use the Brand Reputation Tracker to keep tabs on their reputation on social media or look at competitor brands within specific industries. Managers could use that information to drive ways to boost their standing with consumers, ultimately delivering more revenue and profits, say the researchers.
The research, “Real-Time Brand Reputation Tracking Using Social Media,” is currently available online, and will appear in the July issue of the Journal of Marketing.
Media Relations Manager