“Customer switching intention from home delivery to smart locker delivery: Evidence from Vietnam”

The purpose of this study is to uncover evidence on the factors influencing switching intention from home delivery to smart lockers in the last-mile delivery service. The research model is constructed based on the Pull-Push-Mooring Theory and the Customer Perceived Value Theory using structural equation modelling to analyze data collected from 557 smart locker users in Vietnam. The results indicated a significantly positive influence of pull factors (convenience, environmental friendliness, and security) and push factors (delivery failure experience and risk), and confirmed that mooring factors (habit and switching cost) negatively impacted customers’ intention to switch from home delivery to smart lockers. The study also revealed that mooring factors moderate the relationship between pull factors and the intention to switch. Moreover, gender, age and frequency of shopping online are significant to switching intention, and usefulness mediates between them and switching intention. Several managerial implications were suggested for stakeholders in order to enhance customers’ switching intentions to use smart lockers, thereby improving the quality, efficiency, and sustainability of the last-mile delivery service in the future.


INTRODUCTION
The trend of global economic integration, economic recovery after the COVID-19 pandemic, and the explosion of e-commerce have contributed to the rapid growth of last-mile delivery services.According to the Vietnamese Ministry of Information and Communications, total postal output in 2022 will reach 1,550 million postal items with revenue reaching 2.37 billion USD, an increase of 16.2% compared to 2021, and predicted to reach 4.88 billion USD by 2030 (Vietnam Ministry of Information and Communications, 2023).E-commerce has been booming with a rapid increase in the frequency of online transactions, placing significant pressure on Vietnam's last-mile logistics businesses on home delivery time and cost efficiency (Huong & Thiet, 2020).Consequently, this has led to a high rate of home delivery failures (Buldeo Rai et al., 2021), as well as urban traffic congestion and environmental pollution (Zhang et al., 2019).As a result, a smart locker is believed as one of the technology solutions to optimize last-mile delivery to help improve service quality, reduce the risk of lost goods, and flexibly return goods (Wang et al., 2014).Smart lockers incorporate locking technology and personal identification codes, allowing users to de-liver, receive, and return goods without employee supervision (McKinnon & Tallam, 2003).Smart lockers contribute to improving the efficiency of last-mile delivery services thanks to the following features: reducing time and failure rate of delivery due to the recipient's absence, flexibility in delivery time for users when they can choose the time and location, contributing to reducing pressure on the traffic system in large cities, effectively integrating reverse logistics services (Quynh et al., 2023).In Vietnam, smart lockers are considered a new technology solution, attracting many consumers to use alternative services for home delivery (Iwan et al., 2016).Some studies on the topic of last-mile delivery mainly look at the technical characteristics and benefits and disadvantages of smart locker technology individually (Huong & Thiet, 2020;Quan et al., 2022).Nevertheless, the discussions about technology solutions for last-mile delivery services, especially about smart lockers, are still fragmented and need to be clearly oriented in the future (Lim et al., 2018).Hence, there is a significant gap in research on the comprehensive perspective on the factors that pull, push, and moor consumers' switching behavior from home delivery to the smart locker (including factors related to users' personal characteristics).

LITERATURE REVIEW AND HYPOTHESES
The Pull-Push-Mooring theory was first developed by Lee (1966) focusing on the explanation of the positive factors that promote the change of the subject's current condition, the factors that attract the change of the subject based on new characteristics, benefits, and values, and mooring factors that impede the transformation (Moon, 1995 The first mooring factor, habits, is learned behaviors that become reactive and automatic in order to achieve specific goals (Verplanken, 2006).Several researches have addressed the impact of the habit factor in lowering intention (Tuu et al., 2008).A strong habit prevents the intention to switch.In particular, home delivery is a common click-and-choice routine in online purchasing because traveling to pick up goods is considered more difficult than staying at home waiting for deliveries (Wang et al., 2014).When compared to home delivery, obtaining delivery service using smart lockers may require customers to relocate; the further the distance, the less convenient it is, and the more unlikely customers switch to use smart lockers (Chen et al., 2018).Another mooring factor, transaction costs, is identified as one of the elements that generate obstacles in switching behaviors because customers prefer to hunt for cost-optimizing solutions.Furthermore customers' intentions and behaviors are constrained when costs decrease (Campos & Mello, 2017).As demonstrated in some previous studies, the customer switching intention decreases when the conversion cost rises (Hou et al., 2011).The expenses associated with the smart locker service include (i) the cost of accessing information about the smart locker; (ii) the cost of learning how to use smart lockers; (iii) the cost of traveling to the smart locker location; and (iv) the cost of using the smart locker (Yuen et al., 2019).Low switching costs enhance the establishment of switching intention and vice versa.Customers analyze and balance the costs and benefits of switching for products and services with high switching costs (Sun et al., 2017), they are bound by current products even when they can realize the shortcomings of the current product or see the superiority of the new product.
Home delivery is one of the popular and preferred options for last-mile delivery (Buldeo Rai et al., 2021).However, errors can occur at any stage during the service delivery process, resulting in the delivery failure.All variables that disrupt service delivery or cause service failure, whether objective or subjective, have a negative influence on customers' experience and both sellers and service suppliers (Shan et al., 2021).From a business's perspective, delivery failure increases the cost of product returns, communication, and re-delivery.Besides, from the customer's view, increasing the delivery time can cause discomfort, which is unsuitable for many customers' lifestyles and working conditions (Song et al., 2009).When customers' service experience fails, they are more likely to consider changing their present buying habits (Sun et al., 2017).This finding indicates that favorable service experiences decrease switching intention, whereas insufficient service experiences promote switching intention (Chiu et al., 2011).On the other hand, customers' perceived risks for home delivery services include the risk of personal information being compromised (Cheng et al., 2019); the risk of the buyer not receiving the goods due to loss or confusion; the risk of products and goods being damaged; and the risk of responding to and returning goods (Haridasan et al., 2021).The perceived risk of home delivery is higher than the perceived risk of distribution via smart lockers, which prompts customers to switch to a low-risk solution (Marriott & Williams, 2018).
The research objective is to identify the factors that influence customers' intentions to switch from home delivery to smart lockers.To address this research gap, the following hypotheses are formed: H 1 : Pull factors (perceived convenience, privacy security, and environmental friendliness) have a positive effect on customers' switching intentions.
H 2 : Mooring factors (habit of using home delivery and perceived switching cost from a smart locker) have a negative effect on customers' switching intentions.

RESEARCH METHODOLOGY
A questionnaire was designed by forward and back-translating this instrument from English to Vietnamese and vice versa.All constructs and measurement scales were adopted from prior studies.Specifically, convenience, which includes 5 observed variables, was inherited from (ii) Questions about the factors influencing the decision to use smart lockers; and (iii) Personal information.All measurement items were assessed using a five-point Likert scale from 1 (strongly disagree) to 5 (strongly agree).
Since there was no sampling frame available, data were collected via an online survey in early 2023 using the easy-to-use snowball sampling methodology.Snowball sampling is a referral technique that enables researchers to collect data by utilizing existing social connections.This sampling method involves initial respondents suggesting their friends and acquaintances to participate in the study.The questionnaires were sent to respondents directly (28.54%), online (58.53%), and by telephone (12.93%).After 2 months of data collection, from the 700 questionnaires delivered, 581 responses were received, with 557 responses valid for analysis.Table 1 shows the demographics of the study population:  A structural equation model is used in this study to evaluate the effect of pull factors, mooring factors, and push factors on the intention to switch from home to a smart locker.The data analysis is performed using the Smart-PLS 4.0.This software package is mainly used for the application of struc-tural equation modeling based on the partial least squares method (PLS-SEM).Therefore, it is used to investigate the relationship between independent variables and dependent variables.It is also used as a moderator for quadratic.The model was evaluated in two steps using structural equation modeling.First, the measurement model was evaluated.Second, the structural model was evaluated to test research hypotheses related to the intention of switching of smart lockers in last-mile delivery.

RESEARCH RESULTS
The data analysis results ( scales for the quadratic variable.The results of the index (Table 3) show that all correlation values are smaller than the threshold of 0.85 (Henseler et al., 2015), so the scales in the model are all guaranteed with the discriminant value.
Checking the model's multicollinearity, the VIF coefficients of all variables are in the range of 1.335-2.746,smaller than the proposed level 5, which confirms that the model has no multicollinearity problem (Henseler et al., 2015).
The results of the analysis and testing of hypothesis H 1 on the positive impact of pull factors on switching intention from home delivery to using smart locker delivery are accepted (β = 0.115, t = 3.754, p = 0.000).This shows that the benefits of last-mile delivery through smart lockers, includ-ing the growth of environmental friendliness, security, and convenience, are likely to promote customers' intention to use this form rather than the traditional delivery method.Hypothesis H 2 , referring to the influence of mooring factors on customers' switching intention in this study, is confirmed (β = -0.161,t = 4.627, p = 0.000).This result demonstrated that besides the advantages of smart lockers, some constraints, such as switching costs or habits, hinder the customer's intention to switch from home delivery to using smart locker delivery.Hypothesis H 3 is accepted when finding evidence of the influence of push factors on customers' switching intention to use smart lockers (β = 0.080, t = 2.297, p = 0.022).It can be identified that home delivery failure experience and traditional home delivery risks are the factors driving customers' switching intention to use smart

DISCUSSION
The results of this study support the views of Chen and Keng (2019) and Quan et al. ( 2022), affirming the factors that attract and increase the intention to switch from home delivery to smart lockers are convenience, eco-friendliness, and privacy security.Indeed, a smart locker is an automatic device that is fixed at a location in a customer's living area (apartment, office building, parking lot...).When a customer purchases goods and chooses to receive goods through a smart locker, the customer is provided with an electronic code for the smart locker via email or phone (Lemke et al., 2016).The solution of smart lockers contributes convenience for both customers and delivery staff not only by lowering the pressure of errors on delivery addresses and incorrect delivery times but also by enabling a customer to proactively receive goods (Van Duin et al., 2020).Furthermore, smart lockers enhance convenience and flexibility by shortening shopping and delivery time at home (Zarei et al., 2020).Another advantage of a smart locker is that it protects products and minimizes the risk of leaking information when products are transferred to the destination without observation, even if the goods are lost or not received.Moreover, this solution assists customers to return items and products in a convenient and secure way (Huong & Thiet, 2020).The eco-friendliness of smart lockers is proven to have the lowest level of impact.However, it also shows that customers are interested in the "green" values of smart lockers.Therefore, the communication for the smart locker service not only focuses on the function but also the environmental friendliness value of this delivery method, especially under the overloaded condition of last-mile delivery.
This finding is consistent with some previous studies (Featherman & Pavlou, 2003;Yang & Lin, 2015), highlighting that risks and home delivery service failure experiences have a negative impact on the intention to switch from home delivery to smart locker.Potential threats increase, which motivates customers to switch behaviors.This study presents evidence of the factors boosting customers to switch to receiving products via smart lockers, such as home delivery failure and unsafe receiving experiences.Over the years, home delivery has been a popular last-mile delivery method in Vietnam.However, home delivery has increased the negative experience and risk for the customer.The customer is required to arrange time and be present at the location and time they have chosen; otherwise, the delivery will fail.The home delivery failure experience is a common case as the delivery address is a restricted access area, or the deliveryman informally changes the delivery location, which causes the customer to lose time and cost.Customers who receive goods via traditional channels may choose to collect at a convenient place.However, this method brings certain risks, such as getting the wrong product, losing items, having problems returning the product, and revealing personal information about the product.These are the basic limitations of the home delivery method, reducing customers' satisfaction with last-mile delivery.The study also indicates that switching costs and habits are barriers preventing a customer's switching intention to use smart lockers.
This conclusion supports the findings of Chen and Keng (2019) and Tsai and Tiwasing (2021).Smart locker is an essential cost-cutting and operational efficiency solution for last-mile delivery services (Van Duin et al., 2016).Nevertheless, applying technology to this service requires customers to skillfully man- age the integrated technology on the smart locker.Additionally, home delivery has become a habit of many customers in shopping online; a customer can directly receive, check the goods and pay the bill in cash.In the case of using a smart locker, a customer has to pay for the supplier of smart locker, and all transactions are online payments.Furthermore, customers are frequently concerned about how switching behavior may affect their efforts in terms of time, cost, process, and existing advantages.
Some previous studies suggested that the moderating role of anchoring factors with the other two factors was insignificant (Liu et al., 2021).This study contributed evidence about the negative regulatory role of mooring factors in the link between pull factors and the consumer's switching intention from home delivery to using smart lockers.In particular, the greater expense of switching to smart lockers and the more well-defined in-person pick-up habits will influence and weaken the bond between pull factors and switching intentions.Therefore, along with the outstanding functional values of the smart locker, it is vital to design easy-to-use smart lockers so as to ensure accessibility for customers and save time and switching costs.Additionally, communication of smart lockers should be widely spread to promote sustainable consumption.
This study also pointed out that customers' intentions to switch from home delivery to using smart locker delivery are influenced by demographic factors such as gender, age, and delivery frequency.Female consumers, in particular, have a higher switching intention than male customers, older customers, and customers who receive items on a regular basis have a higher switching intention to use smart lockers.In fact, smart lockers are a new technology solution in last-mile delivery services with new advantages and convenience that will easily attract customers having a high frequency of buying and receiving goods.Generation X and Y customers with a more stable income than Gen Z customers will have higher switching intentions.

CONCLUSION
This study aims to enhance understanding of the impact of pull, push, and mooring factors on consumer intention to switch from home delivery to smart locker delivery.Specifically, it examines the influence of mooring factors on the relationship between pull factors, push factors, and a customer's switching intention to use a smart locker.The findings of this study offer important theoretical contributions.Firstly, it is confirmed that in the context of last-mile delivery services in Vietnam, the structure of factors influencing the customer's switching intention is pull and push factors, respectively.Secondly, switching costs and habits are mooring factors that reduce the attractiveness of smart locker solutions.Finally, the study also confirms that the awareness of the environmental friendliness of service has a role in pulling the customer's switching intention.
The findings have several managerial implications for both the last-mile delivery service and the smart locker suppliers, in particular, designing and operating service regarding pull-push factors to attract switching intention along with solving mooring factors.Firstly, service suppliers should consider increasing convenience through network expansion and the installation of smart lockers in residential areas, offices, etc., for customers to easily access.Enhancing reliability by providing accurate service and minimizing errors in smart locker delivery; improving and strengthening the security features of new technology to fully secure customers' information in order to contribute more benefits than traditional ways.Secondly, smart locker service suppliers must raise customer awareness of the smart locker service in last-mile delivery as well as develop communication channels about the benefits, superiority, and environmental friendliness of smart lockers, in comparison with traditional solutions.As a result, customers gradually switch to using smart lockers.Finally, service suppliers should put more effort into minimizing the cost of switching services through discounts, subsidies, and promotions in order to balance the values for customers.Additionally, a professional support service for the smart locker system should be established, which helps reduce the negative consequences of smart locker service through feedback and service commitment.
The omnichannel approach is one of the latest trends in last-mile delivery.Although home delivery is still one of the most common solutions in last-mile delivery in Vietnam, and it is difficult to find an absolute alternative, the adoption of a smart locker allows for partial replacement, diversifying choices for customers, and assisting businesses in maximizing efficiency, competitive advantage toward sustainable development.Thus, the discussions in this study have significant implications for businesses in the last-mile delivery service business.However, there are still some limitations: (i) The study solely examines the intention to switch; however, it does not go extensively into the types of intention to switch temporarily, partially, or fully; (ii) Some analysis of differences in switching intention by groups or specific customers has not been mentioned in this study.Therefore, it is necessary for future researchers to further consider the following topics: (i) multi-channel integration models in last-mile delivery; (ii) the level of the customer's intention to change behavior; and (iii) the differences in customer groups when considering the switching intention in last-mile delivery.

H 3 :
Push factors (perceived delivery failure experience and risk) from home delivery have a positive effect on customers' switching intention to use smart lockers.H 4a,b : The relationship between pull factors, push factors, and customers' switching intention to use smart lockers is moderated by mooring factors.

Table 1 .
Research sample characteristics

Table 2 )
reveal that the model's first-order items provide reliability and validity.Cronbach's alpha coefficient of first-order variables includes convenience, eco-friendliness, security, switching cost, habit, risk perception, and delivery failure experience that are all greater than 0.65, composite reliability (CR) of the scales in the research model is satisfactory with the achieved values all greater than 0.70.The least average variance extracted (AVE) of the factors in the model reached a value greater than 0.588.Thus, all variables ensure reliability.

Table 2 .
Results of evaluating composite reliability, outer loadings, and AVE of first-order items

Table 2 (
cont.).Results of evaluating composite reliability, outer loadings, and AVE of first-order items

Table 4 .
Hypothesis testing results