Introduction

Following the stream of digitalization, and the rapid growth of e-commerce, consumer purchase decisions are increasingly made online. The large-scale adoption of the internet in daily life is the biggest event that has affected marketing over the past three decades. In the context of this internet era, information load, massive and varied options are provided to society. Having a large number of options might sometimes negatively affects customers buying intentions (Graeme, 2009), and the ensuing information overload and pressure on information processing can lead buyers to make irrational and simplified choices (Carmon et al., 2003). This not only affects customer satisfaction, but also hinders many opportunities for organizations to grow, especially under the e-commerce environment which brought larger perceived risk to customers (Pedro et al., 2014). It has formed a customer need of having a tool that could help them to make fast, convenient, and accurate decisions.

This phenomenon has offered a fundamental basis for the emergence of a automated online product recommendation systems, to help buyers and individual users reduce information overload by reflecting their specific needs, preferences, prior purchase histories, or demographic profiles (Izak & Weiquan, 2005). Moreover, numerous opportunities abound for businesses to better attach and serve customers with both the progress and prevalence of Webbased technologies (Izak & Weiquan, 2005).

Existing literature has already revealed that automatic recommendation systems are the most influential recommendation source on consumers’ online product choices (Bo & Izak, 2007), as it decreases customer’s search costs and effort especially on the online retailer platforms as it is an independent third-party website that provides more objective information. (Sylvain & Jacques, 2004) However, since the connection between the recommendation systems and their users is an agency relationship, users are uncertain about whether RAs are working for them specially or serving the other parties who have make them available, including merchants and manufacturers (Bo & Izak, 2007). The most pronounced issue involved in recommender adoption is the consumer’s trust in them. At the current condition, as recommendation systems develop and gradually become widely known, people start concerning more about their privacy and the intention behind the recommendations promoted by platforms, which causes dissatisfaction and decreases the credibility and user’s willingness of accepting recommendation systems.

Trustworthiness or credibility is a basic element that ensures the happening of trades between buyers and sellers, particularly in the rapidly evolving online environment. Due to consumers’ increased technical and commercial sensitivity, dispel the doubts of customers accepting recommendations gradually becomes one of the main task organizations should focus on. For instance, customers would doubt the online system due to the distance between suppliers and users, which emphasized by the absence of direct attachment of their recommendation service and one of the biggest concern of users is that it can be easier for the E-vendor to take advantage of online users (Izak & Weiquan, 2005).

In order to fulfill the empirical gap, the question this paper will mainly analyze is:

“How to enhance users’ willingness of accepting recommendations from automatic recommendation systems?”

Nowadays, with the rising and expansion of social media, there is a new concept called influencer marketing that has been described as the “next golden goose” of marketing (Newman, 2015). Organizations use or cultivate influencers who as third-party endorsers attract audiences and shapes their attitudes by posting pictures, videos, articles through social media (Karen et al., 2010). Sylvain and Jacques found that the recommendation source “other customers” was perceived as more trustworthy than recommendation systems (Sylvain & Jacques, 2004) but less professional than human experts. This research would like to take these two findings into consideration to develop a general recommendation system. The improved system will be named as Curator System in this paper according to the definition of the curator, which is another form of influencers, usually have a certain level of expertise or influence as an opinion leader(Jianling, Ziwei & James, 2020). Therefore, in this paper, two main types of the curator will be investigated, one is the curator who focuses on using their expertise, another one is influencer curator who focuses on using their fame and influence.

Despite some existing studies having investigated the importance of trust for RSs, effects, and efficiency of social media influencers on market, this paper instead combined influencer marketing as a strategy to see whether involving influencers into the recommendation system will have a positive impact on the trustworthiness of automatic recommendations, thereby enhancing customers’ willingness to look into the recommendations and take the recommendations into their consideration set.