ISSN 2330-717X

How Social Media And AI Enable Companies To Track Brand Reputations In Real-Time


Researchers from University of Maryland, North Carolina State University, National Taiwan University, Oxford University, Kings College London, and Perceptronics Solutions, Inc. published a new paper in the Journal of Marketing that examines how artificial intelligence (AI)-based text analysis of social media can monitor the extent to which brand reputation rises and falls over time.

The study, forthcoming in the Journal of Marketing, is titled “Real-Time Brand Reputation Tracking using Social Media” and is authored by Roland Rust, William Rand, Ming-Hui Huang, Andrew Stephen, Gillian Brooks, and Timur Chabuk.

Organizations’ brand reputations can rise and fall based on brand-related events. For example, when Goya CEO Robert Unanue suggested that the 2020 U.S. presidential election was fraudulent, that controversial assertion likely offended a large segment of the population. How can we tell? This research team demonstrates that using artificial intelligence (AI)-based text analysis of social media can monitor the extent to which brand reputation rises and falls over time. What’s more, merging this social media monitoring with the Rust-Zeithaml-Lemon customer equity drivers can show exactly which dimensions of brand reputation are changing.

All marketers know that brands are important and that stakeholders’ views of the brand are reflective of many different factors. Also, brand reputation may rise or fall over time, due to events that affect the brand. The fact that brand reputation is not constant makes it essential for companies to monitor their brands continuously, to determine whether a brand’s reputation is changing, and to evaluate which aspects of the brand are causing these changes. Survey-based approaches exist, but surveying stakeholders on a daily basis is generally too expensive to be practical. Another approach is to infer what is happening to brand reputation by mining social media. Automatic, AI-based text analysis of social media posts is a realistic alternative.

Because Twitter is widely used by people to express opinions about brands and is often monitored by the public, the research team chose it as the platform to explore and baseline results. By analyzing millions of Twitter tweets, they demonstrate that their brand reputation tracker accurately reflected major brand events in real-time. For example, when it was revealed that Facebook had improperly shared personal information with an outside company (Cambridge Analytica), the brand tracker reflected that right away with a decline in brand reputation. On the positive side, when Google added new features, its brand reputation ratings went up.

Rust says that “It is one thing to know that brand reputation is improving or declining, but another thing entirely to figure out why. To ensure the actionability of our brand reputation tracker, we sorted the tweets according to the Rust-Zeithaml-Lemon customer equity drivers, which have been applied by many Fortune 500 companies. These three drivers, along with their sub-drivers, help managers know where to focus, making the brand tracker managerially relevant and actionable.” The three main drivers of customer equity according to this framework are value, brand, and relationship. The value driver considers the rational or objective aspects of the brand, such as price, quality, or convenience. The brand driver considers the emotional or subjective aspects of the brand, such as attitude toward the brand, or perceptions of the brand’s ethics. The relationship driver focuses on the aspects of the brand that create switching costs, such as loyalty programs or knowledge of the brand.

Rand explains that “Our brand reputation tracker is unique in that it can reflect the impact of brand events in real-time and connect them in a more granular way to managerially specific drivers of brand reputation. Managers can use it to drive programs that enhance their brands’ standing with customers, forging deeper relationships and ultimately delivering more revenues to their bottom line.”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.