AB testing for seller ratings of un-rated products in ecommerce platforms

Abstract

Product reviews are key to purchase decisions in ecommerce, and for unrated products there is little to no authentic information on ecommerce portals. Consumers increasingly prefer products with consumer ratings as consumers are making their buying decisions from cues of other consumers rather leading to sales of only few branded or rated products. Thus, Seller reviews can be seen as an additional attribute to improve the trust in consumers of the product. The theory is tested with randomized survey showing products with high, low and no seller ratings and gauging the trust and purchase intention. Buyers trust and purchase decision served as dependent variables with demographics as control variables. The results showed that high seller ratings led to high trust and lower seller ratings led to lower trust than no seller reviews. Also, online trust is proven to be positively correlated with purchase decision. Overall, the data suggests that good seller ratings can help e-commerce portals to push more unrated products for customer purchase.