Brand echo chambers

Arvin Moazemi, Netzer Oded, Verena Barbara Schoenmueller

Research output: Conference paperContribution


The social media echo-chamber has been widely discussed as a risk to the exploration of new ideas and political views because people are surrounded by like-minded people. Echo-chambers have been mostly discussed in the context of media and political ideology. In this research, we explore echo-chambers in the context of brands. How do consumers who are surrounded by consumers with similar brand preferences perceive and behave toward brands? To answer this question, we derive two complementary measures of brand echo-chambers either capturing echo chambers based on whether the focal consumer's social network also follows the brand, or whether the consumer's social network has similar brand preferences/brand followership to the focal consumer. To measure brand echo-chamber, we collected a large-scale Twitter dataset that comprises 604,182 followers of over 100 brands. We leverage our Twitter data to assess individuals' sentiment and frequency of speaking about brands and connect it to the echo-chamber measures. We find that people in more echo-chambered networks talk about brands more frequently and more favorably. To evaluate the relationship between brand echo-chambers and brand-based equity, we relate our brand echo-chamber scores with Young and Rubicam's Brand Asset Evaluator data. We find that brands with higher levels of echo-chambers, tend to be perceived by customers as more independent, intelligent, and straightforward kind, helpful, caring about its customers, socially responsible, down to earth, and gaining in popularity, but are perceived as less stylish and upper class.
Original languageEnglish
Publication statusPublished - 7 Jun 2023
Event45th ISMS Marketing Science Conference 2023 - Miami
Duration: 7 Jun 202310 Jun 2023


Conference45th ISMS Marketing Science Conference 2023


Dive into the research topics of 'Brand echo chambers'. Together they form a unique fingerprint.

Cite this