Toward a Theoretical Model of Authentic and Fake User-Generated Online Reviews

Background

User-generated online reviews are heavily read by consumers to make purchase decisions. Thought to be unbiased accounts of actual post-purchase experiences, reviews are often seen as being highly authentic and trustworthy (Ott et al., 2011). However, the authenticity of reviews cannot be taken for granted. Their disembodied nature allows anyone with Internet access to post fake reviews easily without entering into any sales transactions (Singh et al., 2018). In fact, fake reviews have recently been shown to outnumber authentic ones. Specifically for electronics, as much as 61% of all reviews could be fake (Sterling, 2018).

Increasingly, websites such as Amazon and Expedia only allow individuals who have completed an actual transaction to post reviews. However, in the absence of user authentication on websites such as TripAdvisor, the authenticity of reviews remains a nagging question. Moreover, given that fake reviews are deliberately written to appear authentic, it is tough for consumers to distinguish between the two. When consumers overlook authentic reviews and follow those that are fake, they make sub-par purchase decisions that in turn unfairly affect the fate of businesses. This not only makes the relationship between consumers and businesses untenable but also diminishes the worth of user-generated content in general (Pal & Chua, 2016; Pal et al., 2017; Singh et al., 2018; Thomas & Elias, 2017).

The Missing Link in the Literature

The growing popularity of reviews—some authentic, others fake—coupled with consumers’ increasing proclivity to trust them have given rise to two disparate research strands. The first focuses on ways to distinguish between authentic and fake reviews (e.g., Harris, 2012). With a technical orientation, studies related to this strand are generally conducted by computer scientists using techniques such as machine learning (e.g., Ott et al., 2011). However, the consumer behavior perspective remains largely ignored.

The second research strand deals with consumers’ perceptions of reviews (e.g., Gupta & Harris, 2010). Usually undertaken by management scholars with a socio-psychological orientation, works related to this strand are characterized by user studies (e.g., Casaló et al., 2011). However, reviews shown to participants of the user studies were often implicitly assumed to be authentic.

Interestingly, little scholarly efforts have been invested hitherto in uniting the two research strands. As a result, the literature on authentic and fake reviews remains fragmented. The piecemeal scholarship begs the question: What theoretical lenses can be used to holistically study differences—encompassing both actual and perceived—between authentic and fake reviews? It appears that the question can be tackled effectively only by dovetailing the two research strands.