Shopper Types and the Influence of Persuasive Strategies in E-Commerce Ifeoma Adaji, Kiemute Oyibo, Julita Vassileva University of Saskatchewan, Saskatoon, Saskatchewan, Canada ifeoma.adaji@usask.ca, kiemute.oyibo@usask.ca, jiv@cs.usask.ca Abstract. Identifying the shopper type of e-commerce users has been shown to help e-businesses better understand the attitude of their customers and what they look out for in their shopping decision-making process. However, applying the same influence strategy for all shopper types will not likely bring about the de- sired behavior change. Because influence strategies have higher efficacy when personalized, we propose a personalized approach to implementing influence strategies for the different shopper types by identifying what shopper type is sus- ceptible to which influence strategy. To advance research in this area, we explore the influence of the commonly used shopper types convenience shopper, store oriented shopper, balanced buyer and variety seeker on Cialdini's six influence strategies (reciprocation, commitment, consensus, liking, authority and scarcity) in order to determine what shopper type has the greatest influence on each per- suasive strategy. To achieve this, we conducted a study of 226 e-commerce shop- pers and developed a path model using Partial Least Squares - Structural Equa- tion Modelling (PLS-SEM). We further carried out multi-group analysis between female and male shoppers. Our results suggest significant differences of the var- ious shopper types’ susceptibility to the persuasive strategies. In particular, bal- anced buyer shopper type has the strongest influence on commitment persuasive strategy and insignificant effects on the other strategies. In addition, while the male variety seekers are susceptible to the influence strategy scarcity, the females are not. Similarly, while the female convenient shoppers are susceptible to the influence strategy scarcity, the males are not. 1 Introduction Identifying and understanding the shopping motivation of clients is essential to e-busi- nesses as this could help companies better understand the attitude of customers and what they look out for in their shopping decision-making process [1]. This information can help businesses create a better shopping experience for their clients [2]. There are various classifications of shoppers based on their shopping motivation. One popular classification is by Rohm et al. [1]. They classify e-commerce shoppers into four shop- per types: convenience shoppers, variety seekers, balanced buyers and store-oriented shoppers. The convenience shoppers are motivated by online shopping convenience. Shoppers in this category do not typically seek immediate possession of their products and they shop online for specific products and services. The variety seekers on the other Copyright © 2018 held by the paper’s authors. Copying permitted for private and academic purposes. In: R. Orji, M. Kaptein, J. Ham, K. Oyibo, J. Nwokeji (eds.): Proceedings of the Personalization in Persuasive Technology Workshop, Persuasive Technology 2018, Waterloo, Canada, 17-04-2018, published at http://ceur-ws.org Shopper Types and the Influence of Persuasive Strategies in E-Commerce 59 hand are more interested in seeking variety of products across various retailers and brands. The balanced buyers are motivated by the need to seek information online like the variety seeker. However, the balanced buyers do not plan their shopping ahead. The store-oriented shoppers want immediate possession of goods purchased and are more inclined to social interaction. Shoppers in this category prefer the feel of a physical store to an online marketplace. Despite these classifications of shoppers, there is little knowledge of their susceptibility to the various persuasive strategies. Because influence strategies are more effective when personalized [3], it is imperative to understand what strategies particular shopper types are influenced by in order to personalize these strat- egies to the shoppers. We hypothesize that identifying what influence strategies shop- pers in each category are susceptible to could lead to a more personalized shopping experience for the customer. For example, if the customers who belong to the shopper type variety seekers are susceptible to the influence strategy scarcity, the e-commerce system could personalize product selection, product display and product descriptions using various scarcity strategies. This could bring about the desired positive change in the customer. To advance research in this area, we carried out a study of 226 e-commerce shoppers to investigate how the various shopper types are influenced by persuasive strategies. We measured persuasive strategies using Cialdini’s six influence strategies [4] because they are commonly used in several domains including e-commerce [5]. We developed and tested a path model using partial least squares – structural equation modelling (PLS-SEM). We further explored the moderating effect of gender on the model. The result of our analysis suggests significant differences in the susceptibility of the various shopper types to the different influence strategies. In particular, balanced buyer shopper type had the highest influence on commitment persuasive strategy and has insignificant effects on the other strategies. This suggests that balanced buyers are likely susceptible to commitment strategy. In addition, convenience shopper had the highest influence on scarcity, while store oriented buyer had the highest influence on consensus. Variety shopper, on the other hand, had the highest effect on the influence strategy authority. Furthermore, while the male variety seekers are susceptible to the influence strategy scarcity, the females are not. Similarly, the female convenient shoppers are susceptible to the influence strategy scarcity while the males are not. These results suggest possible guidelines in the personalization of influence strate- gies in e-commerce. 2 Related Work 2.1 Shopper Types Classifying shoppers according to their shopping motivation and behavior has been suggested as a way to help businesses effectively tailor products and services to the various segments of customers [1]. It also helps businesses to better understand the attitude of their customers and what they look out for in their shopping decision-making process [2]. Several typologies of online shoppers have been developed in the past. Kau 60 Shopper Types and the Influence of Persuasive Strategies in E-Commerce et al. [2] classified online shoppers into six categories based on the motivation and con- cerns of online shoppers and their information seeking patterns. Rohm and Swamina- than [1] identified shopper types based on the motivations of shoppers and proposed four categories: convenience shoppers, variety seekers, balanced buyers and store-ori- ented shoppers. According to the authors, the convenience shoppers are motivated by online shopping convenience. Shoppers in this category do not typically seek immedi- ate possession of their products. The variety seekers are more interested in seeking va- riety of products across various retailers and brands. The balanced buyers are motivated by the need to seek information online just like the variety seeker. However, the bal- anced buyers plan their shopping ahead unlike the variety seekers. The store-oriented shopper wants immediate possession of goods purchased and are more inclined to so- cial interaction. Shoppers in this category prefer the feel of a physical store to an online marketplace. We used this typology in our study because it focuses on online shopping behavior and because the four categories identified by this typology are similar to that of other researchers such as [2] and [6]. 2.2 Persuasive Strategies Persuasive strategies change peoples’ attitude or behavior without coercion or decep- tion [4]. Several persuasive principles exist such as Cialdini’s six influence principles: reciprocation, commitment, consensus, liking, authority and scarcity [4]. The principle of reciprocation suggests that human societies subscribe to the rule of reciprocity, hence, humans feel obligated to return a favor they have received in the past. The prin- ciple of commitment suggests that humans tend to be consistent, therefore, it is likely that people will honor things they have committed to. Consensus principle proposes that people tend to manifest the same behavior and beliefs as others after observing several people behaving in a similar manner. Authority principle suggests that because humans are trained to believe in obedience of authority figures, hence in deciding what action to take in any situation, information from people in authority could help humans make decisions. Liking principle posits that people are more persuaded by some- thing/someone they like. Scarcity strategy suggests that humans seemingly have a de- sire for things that are scare, less readily available or limited in number. These strategies have been used extensively in consumer studies and other domains, thus, we adopted them in this study to explore how the various shopper types are influ- enced by them. 3 Research Design and Methodology In this study, we developed a path model using PLS-SEM to measure the susceptibility of the four shopper types (convenience shoppers, variety seekers, balanced buyers and store-oriented shoppers) to Cialdini’s six influence strategies reciprocation, commit- ment, consensus, liking, authority and scarcity [4]. Figure 1 describes our model which is made of four constructs that measure shopper type and six constructs that measure Shopper Types and the Influence of Persuasive Strategies in E-Commerce 61 the six persuasive strategies. Shopper type was measured using Rohm’s scale [1] while the persuasive strategies was measured using the scale of kaptein [7]. Fig. 1. Research model. All paths assumed positive. CONV = Convenience shopper, STOR = Store oriented buyer, BALA = Balanced buyer, VARS = Variety shopper Table 1. Demographics of participants Demographics Value Frequency (%) Age Less than 30 years 55 Between 30 and 49 40 Over 50 5 Gender Female 44 Male 56 Household size 1 to 3 people 63 4 to 5 people 34 6 or more people 4 Household income Less than US$30,000 40 Between US$ 30,000 and 75,000 42 Above US$ 75,000 18 Continent of origin Europe 8 Asia 35 North America 48 Others 9 To carry out this study, we recruited 226 e-commerce shoppers through Amazon’s mechanical Turk (AMT), online social media and news boards. We used AMT because we wanted a diverse set of participants and because AMT is an accepted method of recruiting participants [3]. We have successfully used online social media and news 62 Shopper Types and the Influence of Persuasive Strategies in E-Commerce boards in the past with success [8], thus, we used them in this study. This study was approved by the ethics board of the University of Saskatchewan. Table 1 describes the demographics of our participants. 4 Data Analysis and Results The aim of this paper is to examine the susceptibility of the four shopper types (con- venience shoppers, variety seekers, balanced buyers and store-oriented shoppers) to Cialdini’s six influence strategies: reciprocation, commitment, consensus, liking, au- thority and scarcity. To achieve this, we carried out structural equation modelling using the SmartPLS tool. We determined the reliability and validity of our constructs and the relationships between the indicators and constructs as recommended in structural equa- tion modelling [9]. Indicator reliability, composite reliability, convergent validity (us- ing Average Variance Extracted - AVE) and discriminant validity were all met as re- quired for structural equation modelling [9]. After establishing the reliability and validity of the constructs in our model, we ex- amined the structural model. We computed the coefficients of determination (R2 values) and the degree and significance of the path coefficients. Table 2 shows the path coeffi- cients between constructs. The number of asterisks represents the degree of significance of each direct effect. The number of asteriks ranges from 1 to 4 which corresponds with the p-value of <0.05, <0.01, <0.001 and <0.0001 respectively. Table 2. Path coefficient between constructs and their significance. AUTH = Authority, COMM = Commitment, CONS = Consensus, LIKE = Liking, RECI = Reciprocity, SCAR = Scarcity. n.s. = Not significant, *=P <0.05, **=p <0.01, ***=p<0.001 and ****=p<0.0001 AUTH COMM CONS LIKE RECI SCAR Balanced 0.116 n.s. 0.327**** -0.078 n.s. -0.062n.s. 0.126 n.s. -0.054 n.s. Buyer Conven- 0.186** 0.203** 0.138 n.s. 0.240** 0.259**** 0.295* ience Shopper Store 0.142* 0.084 n.s. 0.276**** 0.200* 0.105 n.s. 0.209* Oriented shopper Variety 0.260** 0.153 n.s. 0.211*** 0.240** 0.170* 0.173 n.s. seeker Our result suggests significant differences of the shopper types to the susceptibility of the persuasive strategies. Balanced buyer had the highest influence on the influence strategy commitment and had insignificant effects on the other strategies. This suggests that balanced buyers are likely susceptible to commitment strategy. According to the commitment strategy [4], because people typically try to be consistent in nature, getting people to commit to a behavior will likely mean that they will carry out that behavior Shopper Types and the Influence of Persuasive Strategies in E-Commerce 63 in order to remain consistent. Therefore, getting balanced buyers to commit to a behav- ior could likely lead to them carrying out the behavior because they are susceptible to commitment strategy. Convenience shopper had the highest influence on scarcity, while store-oriented buyer had the highest influence on consensus. Variety shopper on the other hand had the highest effect on authority. 4.1 Gender Based Multi-Group Analysis In order to understand the moderating role of gender in the susceptibility of shopper types to Cialdini’s six influence strategies, we carried out multi-group analysis between male and female participants. Our result is shown in table 3; it describes only the sig- nificant differences between females and males. The result suggests some significant differences between females and males. Of worthy mention is the influence of shopper type variety seeker on the influence strategy scarcity which is significant for males (p=0.361**) and insignificant for females (p=0.026 n.s.). This suggests that male variety seekers are susceptible to the influence strategy scarcity. Thus, when presenting prod- uct descriptions to male variety seekers, using strategies that describe scarcity could lead to the desired behavior change. One such strategy is Amazon 1’s use of “items left in stock” to show that an item is about to be scarce. On the other hand, female conven- ient shoppers are significantly susceptible to the influence strategy scarcity (p=0.384**) while the male convenient shoppers are not (p=0.075 n.s.). Thus, when presenting prod- uct descriptions to female convenient shoppers, using strategies that describe scarcity could lead to the desired behavior change. Table 3. Path coefficient between constructs and their significance for paths with significant differences between females and males. COMM = Commitment, CONS = Consensus, LIKE = Liking, RECI = Reciprocity, SCAR = Scarcity. n.s. = Not significant, *=P <0.05, **=p <0.01, ***=p<0.001 and ****=p<0.0001 COMM CONS LIKE RECI SCAR Balanced Female 0.257** -0.169n.s. -0.148n.s. 0.057n.s. Buyer Male 0.538**** 0.185* 0.132* 0.341*** Convenience Female 0.384** Shopper Male 0.075n.s. Store Oriented Female shopper Male Variety Female 0.026n.s. seeker Male 0.361** This study is still ongoing; in the future, we will explore the moderating role of age and culture on susceptibility of the shopper types to the persuasive strategies. In addi- tion, we will explore and list specific recommendations that can be applied in e-com- merce for each of the significant findings. 1 amazon.com 64 Shopper Types and the Influence of Persuasive Strategies in E-Commerce 5 Conclusion In this work in progress paper, we explored the susceptibility of different shopper types to Cialdini’s six influence strategies. Shopper types are used to categorize shoppers based on their shopping motivation and has been shown to help companies better un- derstand the attitude of customers and what they look out for in their shopping decision- making process. In order to tailor influence strategies to the various shopper types and increase their efficacy, we explore the influence of shopper types on the six persuasive strategies reciprocation, commitment, consensus, liking, authority and scarcity. The re- sults of the structural equation modeling carried out on 226 e-commerce shoppers sug- gest significant differences of the various shopper types’ susceptibility to the persuasive strategies. In particular, balanced buyer shopper type had the highest influence on the strategy commitment and had insignificant effects on the other strategies. This suggests that balanced buyers are likely susceptible to commitment strategy. In addition, con- venience shopper had the highest influence on scarcity, while store oriented buyer had the highest influence on consensus. Variety shopper on the other hand had the highest effect on the influence strategy authority. We further explored the moderating effect of gender on our model. The result of the multi-group analysis between females and males suggests that the male variety seekers are susceptible to the influence strategy scarcity, while the female convenient shoppers are significantly susceptible to scarcity. This study is still ongoing; in the future, we plan to explore the moderating effect of other demographic variables of the participants. 6 References [1] A. J. Rohm and V. Swaminathan, “A typology of online shoppers based on shopping motivations,” J. Bus. Res., vol. 57, no. 7, pp. 748–757, 2004. [2] A. Keng Kau, Y. E. Tang, and S. Ghose, “Typology of online shoppers,” J. Consum. Mark., vol. 20, no. 2, pp. 139–156, Apr. 2003. [3] R. Orji, L. E. Nacke, and C. Di Marco, “Towards Personality-driven Persuasive Health Games and Gamified Systems,” in Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems - CHI ’17, 2017, pp. 1015–1027. [4] R. B. Cialdini, Influence: Science and practice, vol. 4. Pearson Education Boston, 2009. [5] M. Kaptein and P. Parvinen, “DYNAMICALLY ADAPTING SALES INFLUENCE TACTICS IN E-COMMERCE,” in Marketing Dynamism & Sustainability: Things Change, Things Stay the Same…, Springer, 2015, pp. 445–454. [6] W. Moe, “Buying, searching, or browsing: Differentiating between online shoppers using in-store navigational clickstream,” J. Consum. Psychol., 2003. [7] M. Kaptein, “Adaptive persuasive messages in an e-commerce setting: the use of persuasion profiles.,” ECIS, 2011. [8] I. Adaji and J. Vassileva, “Perceived Effectiveness, Credibility and Shopper Types and the Influence of Persuasive Strategies in E-Commerce 65 Continuance Intention in E-commerce. A Study of Amazon,” in Proceedings of 12th International Conference on Persuasive Technology, 2017. [9] J. Hair Jr, T. Hult, C. Ringle, and M. Sarstedt, A primer on partial least squares structural equation modeling (PLS-SEM). Sage Publications, 2016.