| Popis: |
Chatbots have been introduced as a new way to receive product reviews. However, little is known about how chatbot-mediated solicitation influences review characteristics such as valence and length, or its helpfulness, which, in turn, affects companies' sales and prospective customers' decision-making. We build upon social presence and audience-tuning theories and conduct four studies comprising 1,012 respondents to narrow this literature gap. Our results show that chatbot-mediated reviews are positively biased and shorter compared to reviews collected using a conventional review form. The effect persists even when controlling for the review elicitation source (seller vs. third-party). Further, we find that the positivity bias varies depending on the level of anthropomorphism in the chatbot (human vs. machine-like). Finally, chatbot-mediated solicitation negatively influences a range of review dimensions (e.g., helpfulness) mainly due to decreased review length. Our study is among the first to introduce chatbots as a new technology to drive the elicitation of online reviews, uncovering a direct link between chatbot use and review valence, length, and helpfulness. The results of our experiment inform platforms, sellers, consumers, and regulators of the advantages but also potential drawbacks of implementing robotic automation on review platforms. |
