At a time when digital platforms influence nearly every aspect of consumer decision making, Isenberg Assistant Professor of Operations and Information Management Zhanfei Lei, PhD, is uncovering how consumers process and respond to the user-generated content that shapes their choices—especially online reviews.
Professor Lei’s research focuses on the hidden biases and automatic judgments that occur in digital environments. This interest began with a fascination with user-generated content (UGC); during her master’s program, she analyzed trending Twitter (now X) topics using topic modeling, a text-mining technique used to scan a collection of text and automatically group words into themes. That project laid the foundation for her later doctoral work exploring online word-of-mouth, specifically how product reviews influence purchasing decisions.
“I realized UGC’s powerful influence on shaping people’s opinions, behaviors, and decision making,” says Lei. “As I conducted a series of studies, I uncovered surprising biases and seemingly irrational behaviors in how consumers process and respond to reviews. These findings challenged conventional wisdom and inspired me to build a coherent research stream around uncovering the automatic and deliberative processes that shape user judgment and behavior in digital environments.”
Her 2025 article published in MIS Quarterly represents a turning point. Drawing on dual-process theories of cognition, the paper investigates the relative impacts of Type 1 (automatic) and Type 2 (deliberative) processing on consumers’ responses to online reviews.
“It not only integrates insights from my earlier studies but also systematically advances the argument that consumers’ use of online reviews in decision making might not be as deliberative and logical as commonly believed,” she says.
One of the most surprising insights, according to Lei, was how much review exposure—not just review quality—influences purchase decisions. Review exposure (or visibility) is shaped by factors such as consumers’ review-seeking strategies and the ranking algorithms used by review platforms.
“Instead, reviews or certain content in the reviews that get more exposure can substantially and automatically shape consumers’ purchase decisions,” she says. “In practice, high-quality reviews are often ranked as top reviews, so a rational business may disregard other reviews and focus its attention and resources on the top ones. However, such a strategy might be misguided because reviews with higher exposure are more persuasive, but they are not necessarily the top or high-quality ones.”
Lei found that frequently exposed reviews can steer consumers' decisions, especially when those reviews are negative. This insight has critical implications for how retailers and platforms structure their review content. Current strategies often emphasize the most helpful or top-rated reviews. But Lei suggests retailers rethink how reviews are ranked.
“Review platforms should take the exposure of reviews into account,” Lei says. “For example, when consumers choose to view more reviews, platforms may show additional reviews rather than repeating the top ones, thereby reducing undue and excessive exposure to the same content.”
The AI Effect
Lei is exploring how review summaries generated by artificial intelligence (AI), which are now common on platforms like Amazon, may further complicate or reinforce these processing biases.
“An important next step is to explore how the availability of such summaries may influence consumers’ Type 1- and Type 2-processing of reviews,” she says. “I’m particularly interested in how the design of AI interventions affects user judgment and the broader implications for digital influence in the age of AI.”
Beyond scholarly contribution, Lei integrates her research into her teaching at Isenberg.
“Bringing these findings into the classroom allows students to critically examine the user experience from a behavioral lens,” she says. “Rather than viewing digital platforms as neutral tools, they begin to see them as environments that can guide, nudge, or even distort decision making.”