Key findings:

Key findings from this research on e-commerce livestreaming include: aesthetic appeal, functional layout, financial security, and verbal communication directly influence customer engagement; customer engagement affects functional, hedonic, and social value creation, which in turn impacts customer loyalty; nonverbal communication and service skills showed no significant impact on customer engagement. Livestreaming enhances visual appeal, personalization, and communication skills, contributing to customer loyalty.

What is known and what is new?

The known aspect in this abstract is the growing importance of e-commerce livestreaming and its potential impact on customer loyalty. The new contribution is the application of the SOR (Stimulus-Organism-Response) theory to explore the relationships between livestreaming e-commerce, customer engagement, value creation, and loyalty, providing a comprehensive framework for understanding and optimizing these factors in the e-commerce industry.

What is the implication, and what should change now?

The implication of this study is the critical role of livestreaming e-commerce in enhancing customer engagement, value creation, and loyalty. Changes needed include incorporating aesthetic appeal, functional layout, financial security, and verbal communication into livestreaming strategies to optimize customer experience and loyalty, while also investing in staff communication skills to further improve engagement and value creation.

Over the past few years, the e-commerce industry has witnessed an extraordinary boom and metamorphosis. This can be credited to the progressions in technology and the rise of online platforms, both of which have propelled businesses into the digital realm, providing customers with unprecedented levels of convenience and accessibility. However, amidst this rapid expansion, e-commerce companies find themselves confronted with a multitude of challenges including intense competition, ever-changing consumer tastes, and intricate market dynamics. In the rapidly evolving world of online commerce, businesses need to make wise and deliberate strategic choices to not only survive but also succeed. The significance of effective decision-making cannot be emphasized enough, as it directly affects the growth, expansion, and financial success of organizations in the e-commerce industry. Recognizing the crucial role of strategic decision-making, businesses are constantly looking for innovative ways to improve their planning processes, aiming to outshine their competitors in this dynamic and ever-changing market. Data below show the worldwide growth of internet users from following Table 1.

      Table 1 World Internet Usage and Population Statistic, June 30, 2022

The data in Table 1 indicates that there has been a significant increase in internet users in Africa and Asia over the past 23 years, with a growth rate close to double digits from the year 2000 to 2023. The penetration rate of internet usage worldwide has reached 67.9% of the total population, which is quite impressive. This growth has had a major impact on people's economic behavior, both before and following the COVID-19 pandemic, and has had significant national implications in each country. Presently, internet users are actively participating in various economic activities online, particularly online shopping. As a result, there has been a rapid surge in online sales, highlighting the importance of e-commerce businesses for achieving success in the market. Figure 1 demonstrates a positive trend in the growth of e-commerce sales, as shown below:

Figure 1 Retail E-Commerce Sales Worldwide, Statista 2023

E-commerce sales growth is predicted to continue increasing until 2026, driven by factors such as increasing internet and smartphone penetration, the growth of the middle class, and the shift to online transactions due to the impact of the COVID-19 pandemic. While Indonesia's e-commerce market is competitive and dynamic, with the entry of new players, scenario planning becomes essential to deal with future uncertainties. Scenario planning is an important strategy in responding to change. It involves developing a set of possible future outcomes and formulating strategies to deal with each scenario. The use of a multi-analytic framework for customer decision-making can support scenario planning in the Indonesian e-commerce market. By leveraging data and analytical techniques, this framework helps understand customer behavior and the key factors that influence their decisions. Strategies can be developed to deal with different scenarios, such as customers who are more price-sensitive or who are more willing to pay more for convenience. Thus, businesses can be better prepared and can make informed decisions in the face of future market changes.

One of the key business issues for the role of Stimulus-Organism-Response (SOR) in e-commerce strategic decision-making for sustainability is the lack of data and understanding of how customers respond to sustainable stimuli. E-commerce businesses need to be able to understand how their customers perceive and react to sustainable initiatives, such as using recycled materials in packaging, offering carbon-neutral shipping options, or donating a portion of profits to environmental causes. This information is essential for making informed strategic decisions about how to allocate resources and develop sustainable business practices. However, many e-commerce businesses do not have access to the data and expertise they need to fully understand their customers' SOR to sustainable stimuli. This can lead to businesses making strategic decisions that are not aligned with their customers' values, or that are not effective in promoting sustainability. Another business issue is the complexity of measuring the impact of sustainable initiatives on customer behavior.

It is difficult to isolate the impact of sustainable initiatives on customer behavior from other factors, such as price, convenience, and product quality. This makes it challenging for e-commerce businesses to measure the return on investment (ROI) of their sustainable initiatives and to make informed decisions about where to allocate resources. Despite these challenges, there is a growing number of e-commerce businesses that are recognizing the importance of sustainability and are investing in sustainable initiatives. The role of SOR in e-commerce strategic decision-making for sustainability is becoming increasingly important, as businesses strive to meet the growing demand for sustainable products and services.

Refers to research questions, the research objective is to Provide practical guidance for e-commerce businesses on how to use SOR to promote sustainability. This guidance could include specific strategies for developing sustainable marketing campaigns, designing sustainable products and services, and measuring the impact of sustainability initiatives.

Problem Exploration

The connection between the "direct broadcast + energy industry" has gained increased focus from major e-commerce sales platforms. The rapid growth of product sales can be attributed to three types of direct broadcast: "e-commerce platform + direct broadcast", "direct broadcast platform + e-commerce", and "platform live streaming". E-commerce live streaming has become the primary catalyst for economic promotions, transforming the traditional e-commerce model and improving the shopping experience and overall quality of life for consumers. As a result, it is essential to have a thorough discussion on the subject of e-commerce live streaming.

SOR (Stimulus–Organism–Response) Model

This study uses the SOR (stimulus-organism-response) model, which is a commonly used theory to analyze online consumer behavior frameworks [1]. Explored impulsive online buying and shopping behavior in reaction to virtual environment cues (content, design, and navigation) using the SOR model [2]. Argued that creating an infectious consumer atmosphere (stimulus) in an online commerce environment would result in impulse buying behavior (response) [2]. Ma et al., (2022) [3] verified the purchase intention of online consumers when stimulated with e-commerce live-streaming scenes through the SOR model. Showed that the entertainment, informativeness, credibility, incentives, and celebrity endorsements related to sponsored Instagram ads are conducive to enhancing the effectiveness of advertising stimuli, which in turn increases consumers' cognitive and affective advertising engagement and flow, affecting their purchase behavior [1] showed that social media use moderates the relationship between scarcity messages and impulse buying in Indonesia. In this study, SOR theory was used as a framework to achieve the following research objectives: (1) to explore the relationship between physical cues (aesthetic appeal, functional layout, and financial security) and social cues (nonverbal communication, verbal communication, and service skills) in e-commerce live streaming services; (2) to evaluate the relationship between functional, hedonic and social values in customer engagement and value co-creation; (3) to explore the relationship between functional, hedonic and social values with customer loyalty; and (4) explore the relationship between functional, hedonic and social values [4].

E‑commerce Live Streaming Services and Customer Engagement

Ambient conditions refer to the various factors that create the sensory environment within a service setting. These factors encompass elements such as temperature, air quality, background music, and lighting. Each of these factors plays a significant role in influencing the overall perception and comfort of customers within the service environment. Space layout and function, on the other hand, focus on the physical arrangement and design of the service setting. This includes aspects such as the placement of furniture, the distribution of facilities, and the overall spatial organization. The arrangement of these elements not only affects the efficiency and functionality of the service environment but also contributes to the overall aesthetic appeal and comfort for both customers and employees. As stated by Bitner (1992) [5], the notion of "services cape" encompasses various elements that contribute to the overall atmosphere and experience of a service environment. 

Customer Engagement and Value Creation

Customer engagement is a natural aspect of communication that reflects how customers perceive and emotionally process their interactions with a brand or company [6-7]. According to Brodie et al. (2011) [8], customer engagement is primarily psychologically intrinsic and includes both individual and group aspects. Customer involvement encompasses participation, satisfaction, and cognitive, emotional, and behavioral reactions as part of the customer's role. Customer readiness is supported by organizational support, socialization directed by the organization, and the brand or company's interactions and connections in social networks to promote their products or services. Other factors that contribute to engagement include user-generated content, involvement, interactivity, service quality, brand experience, the service ecosystem, customer trust, customer identity, virtual service landscape quality, online product reviews, video content, and emotions.

Customer Engagement, Value Co‑Creation and Customer Loyalty

Customer loyalty can be defined as the ongoing commitment of customers to continue purchasing or supporting their preferred products and services in the future [9]. According to Ha & Park (2013) [10], this loyalty is rooted in the deep psychological connection and dedication that consumers have towards a particular brand or company, driven by their own personal preferences and encompassing cognitive, emotional, intentional, and behavioral aspects. Further elaborate that customer loyalty Srinivasan et al. (2002) [11] is reflected in the extent to which consumers prioritize and favor a specific platform, leading to repeated actions and behaviors. Building on this, Van et al. (2010) [12] argue that factors such as word-of-mouth recommendations, interactive experiences, and active engagement contribute to customer engagement behaviors that extend beyond mere transactions. Moreover, value co-creation plays a significant role in this process, whereby consumers utilize their own resources in conjunction with the resources provided by businesses to actively solve problems and generate value [4]. The concept of value in the context of the internet refers to the benefits that customers derive from using it to access, share, or utilize information in a specific situation. This value is not static but is built over time through ongoing communication and interaction between customers and the online platform. This personalized experience encompasses three dimensions of value: functional, hedonic, and social.

The method utilized for gathering data is an essential aspect of any research undertaking. The choice of a data collection procedure depends on the research methods employed. Generally, data can be classified into two types: Primary data and Secondary data. Primary data refers to information that is directly obtained by data collectors or researchers themselves. This kind of data can be collected through surveys, observations, experiments, questionnaires, personal interviews, and similar methods. Conversely, Secondary data is research data that is not acquired directly by data collectors or researchers, but rather originates from external sources. In this study, the author utilizes both primary and secondary data. The data collection is used to provide evidence for the hypothesis [4].

In order to assess physical cues in e-commerce live streaming situations, we employed 17 criteria from Harris and Goode (2010) and Wang (2022) [13-14] during the development of our predictive questionnaires, utilizing exploratory factor analysis. One particular dimension displayed cross-loading and conflicting factors, leading to the removal of four items such as aesthetic demand, functional layout, and financial security. Our questionnaire was constructed using a Likert scale consisting of five points, ranging from strongly agree to strongly disagree. The main structural model was examined using potential variable path analysis to test a series of hypotheses, including H1a, H1b, H1c, H1d, H1e, H1f, H2a, H1b, H2c, H3a, H3b, and H3c. The researchers also investigated the mediating effects of H4, H5, and H6.

In conducting this research, the author used primary data and secondary data. Data collection is shown below this table: 2, 3, 4, 5, 6

Table 2 Primary data through a survey of 1,282 respondents

Table 3 Primary data was obtained from 11 interviews, here's the list

Path Diagram 11 Variable

Table 4 Path Diagram (Pre-Modeling, per variable, after FLR dropdown <0.5) (N=1282)

No	Syntax	Path Diagram (Pre-Modeling, per variable)
1	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES AA  SAMPLE SIZE 1282 RELATIONSHIPS X1 - X4=AA OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
2	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES LF SAMPLE SIZE 1282 RELATIONSHIPS X5 – X10=LF OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
3	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES FS SAMPLE SIZE 1282 RELATIONSHIPS X11 – X14=FS OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
4	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES VC SAMPLE SIZE 1282 RELATIONSHIPS X15 – X17=VC OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM  MODEL CONFIRMATION
5	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES NC SAMPLE SIZE 1282 RELATIONSHIPS X18 – X19=NC OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM	N/A
6	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES SK SAMPLE SIZE 1282 RELATIONSHIPS X20 – X21=SK OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM	N/A
7	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES CE SAMPLE SIZE 1282 RELATIONSHIPS X22 – X25=CE OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
8	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES FV SAMPLE SIZE 1282 RELATIONSHIPS X26 – X28=FV OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
9	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES HV SAMPLE SIZE 1282 RELATIONSHIPS X29 – X31=HV OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
10	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES SO SAMPLE SIZE 1282 RELATIONSHIPS X32 – X34=SO OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
11	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES CL SAMPLE SIZE 1282 RELATIONSHIPS X35 – X38=CL OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM

Validity and reliability 11 Variables

Table 5 Summary of Validity and Reliability Test Results (N=1282, 11 variable)

According to Thompson et al. (2016) [2], in general, the rule of thumb used in the convergent validity test is recommended to have a loading factor value between 0.5 and 0.6, and an AVE value greater than 0.5. As shown in Table 5, based on the processed data from 1282 respondents, it can be seen that 2 (two) variables do not have loading factors so it is not possible to calculate CR and AVE values. In addition, there is a factor loading value below 0.5, which indicates the need to conduct a new simulation without including this specific question item.

Path Diagram 9 Variable

Table 6 Path Diagram (Pre-Modeling, per variable, after FLR dropdown <0.5) (N=1282)

	Syntax	Path Diagram (Pre-Modeling, per variable)
1	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES AA  SAMPLE SIZE 1282 RELATIONSHIPS X1 - X4=AA OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
2	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES LF SAMPLE SIZE 1282 RELATIONSHIPS X5 – X10=LF OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
3	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES FS SAMPLE SIZE 1282 RELATIONSHIPS X11 – X14=FS OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
4	MODEL CONFIRMATION FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES VC SAMPLE SIZE 1282 RELATIONSHIPS X15 – X17=VC OPTIONS:SC SS EF RS PATH DIAGRAM END OF PROGRAM  MODEL CONFIRMATION
5	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES CE SAMPLE SIZE 1282 RELATIONSHIPS X22 – X25=CE OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
6	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES FV SAMPLE SIZE 1282 RELATIONSHIPS X26 – X28=FV OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
7	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES HV SAMPLE SIZE 1282 RELATIONSHIPS X29 – X31=HV OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
8	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES SO SAMPLE SIZE 1282 RELATIONSHIPS X32 – X34=SO OPTIONS: SC SS EF RS PATH DIAGRAM END OF PROGRAM
9	FACTOR: LISREL TIAS OBSERVED VARIABLE X1-X38 COVARIANCE MATRIX FROM FILE D:\DC.COV LATENT VARIABLES CL SAMPLE SIZE 1282 RELATIONSHIPS X35 – X38=CL OPTIONS:SC SS EF RS PATH DIAGRAM END OF PROGRAM

Validity and reliability 9 Variable

Table 7 Summary of Validity and Reliability Test Results (N=1282, 9 variable)

Table 7 shows that all variables in the questionnaire are reliable. Based on the findings of validity and reliability, it can be concluded that the questionnaire can be used as a reliable research measurement tool. Therefore, it is recommended to distribute the questionnaire to the entire sample to conduct further research. After the data was re-evaluated by eliminating the question items whose FLR values were lower than 0.5, the results are presented in Table 3 above. The table shows that the minimum FLR value surpasses the threshold of 0.5, which indicates that the data is feasible to proceed to the next phase of testing. 

To address common potential methodological biases, multilayer testing was conducted using cross-sectional data. First, Harman's one-way test for common method bias was used [14]. This involved performing factor analysis on all items, resulting in the extraction of 9 constructs with characteristic roots greater than 1 from the unrotated exploratory factor analysis results. However, the first factor AVE was found to be 50.4 percent, higher than the recommended value of 50 percent [14]. In addition, the unmeasured potential method (ULMC) factor test was also used [14]. 

The results showed that Factor Model I (CFI=0.91, RMSEA=0.19, GFI=0.91) with 3 factors and Factor Model II (CFI=0.93, RMSEA=0.13, GFI=0.94) did not significantly improve the fitting index when compared (A model is said to be a good fit if it has a CFI value ≥ 0. 9 and is said to be marginal fit if 0.8 ≤CFI ≤ 0.9) According to Brown & Cudeck (1993), as cited by Wijanto (2008) [15-16], an RMSEA value ≤ 0.05 indicates a close fit, while a value of 0.05 < RMSEA ≤ 0.08 indicates a good fit. According to Wang et al. (2022) [13], an RMSEA value ≤ 0.05 indicates model fit, and the GFI value varies between 0 (indicating a poor fit) and 1 (indicating a very good fit). A GFI value of 0.90 or higher is considered a good fit, while a GFI value between 0.80 and 0.90 is commonly referred to as a marginal fit). This indicates that there is no significant common method bias.

Structural model analysis and hypothesis testing

Lisrel 8.8 and IBM SPSS Statistic 26 were used for data analysis, while path analysis was used to investigate the relationship between the research concepts to confirm the research hypotheses. (Table 8 and 9)

Table 8 Correlations between constructs (N=1282)

Table 9 Path Coefficient (N=1282)

Figure 2 The Role of Stimulus-Organism-Response (SOR) in E-Commerce Strategic Decision

The result of the Lisrel SEM model, for the hypothesis testing The Role of Stimulus-Organism-Response (SOR) in E-Commerce Strategic Decision Making for Sustainability, can be summarized as follows

Hypothesis 1

The first hypothesis examines the relationship between physical cues (aesthetic appeal (H1a), layout and functionality (H1b), financial security (H1c), social cues (verbal communication (H1d), nonverbal communication (H1e), service skills (H1f) and consumer engagement. Based on the bootstrapping result, it is found that H1a is supported because the p-value (0.0,032) is lower than 0.05 and the T-Statistics value (16,47) is higher than 1.960. Similarly, H1b is also supported as the p-value (0.041) is less than 0.05 and the T-Statistics value (18,93) is higher than 1.960. Additionally, H1c is also supported as the p-value (0,031) is less than 0,05 and the T-statistic value (2,44) higher than 1,960. 

For Social cues, H1d is also supported as the p-value (0,027) is less than 0,5 and the T-statistic value (30,96) is higher than 1,960. On the other hand, nonverbal communication (H1e) and service skills (H1f) were categorized as not supported because the factor loading value didn’t appear, so it cannot be calculated. This data also matched with interview results from several employees from Shopee, Tokopedia, and TikTok. Ensuring a good user experience and making it easier for users is crucial, as it allows people of all groups and ages to benefit from it. Additionally, it plays a significant role in attracting e-commerce users. They also inform that in real life these values are reflected in several programs and their KPI and their customer program. Here are several programs designed to provide physical prompts to customers: 

- Spread bugs (spread profit coupons), spread according to certain standards, then you will get coupons/discounts 

- Exciting tasks (complete tasks and then get coupons/discounts) 

- Tap the box (just click, spin the wheel), some stores only offer discounts 

- Arrange packages, the prizes are coupons and discounts Tokopedia cashback, and seller discounts 

- Top-ads: Popular searches (bidding system) is the highest bid, taken for popular searches) 

- Plant rice to encourage customers to come shopping, 

- Shake at specific times to get coupons and/or coins, 

- Finally, there is the "Crack the Egg" egg-breaking program to get coupons and/or free shipping.

However, the differences in data results may be influenced by other factors, such as the sellers' communication methods in Indonesia has many kinds of type for the level of interaction between the streamer and the audience during the live streaming, such as how the streamer greets the audience and presents the products.

This result aligns with the argument made by Wang et al. (2022) [13] that emphasizes the stronger impact of synchronicity on social presence. Live streaming, being a unique feature, allows for direct responses from the audience. Synchronicity, in this context, refers to the immediate feedback provided to potential buyers or the audience during a live-streaming session. Live streaming has been found to enhance trust between the audience and sellers. Research by Chandrruangphen et al. (2022) [18] explains that live streaming can also influence the audience's trust, their intention to watch the live stream, and their intention to purchase the products. Additionally, customer trust has been identified as one of the factors that impact customer engagement, which in turn is likely to influence purchase intention (Wang et al., 2022) [13].

Hypothesis 2

The second hypothesis suggests that variables related to Organism (customer engagement) to Value co-creation (function value (H2a), Hedonic Value (H2b), Social Value (H2c). After analysis, it was found that the P-value for H2a, H2b, and H2c is less than 0.05, and the T-statistic is greater than 1.960 so they are all categorized as supported. This finding aligns with previous interviews, which indicates that individuals are more likely to purchase a product based on recommendations and referrals from others rather than relying on commercials or unique selling points provided by the company itself [4]. Furthermore, customers are inclined to place more trust in the information about recommended products when it originates from individuals who possess similar perspectives, backgrounds, a positive reputation, and the ability to persuade others to make a purchase [1].

This finding aligns with the previous interview, they focus on a visually appealing and user-friendly UX experience that enhances the ease of application usage. Personalized product categorization options such as beauty, mom and baby, and automotive to cater to individual preferences they encouraged to make all users happy. Tokopedia offers a convenient feature where deliveries are directly fulfilled from their warehouse, ensuring safe and appealing packaging. Additionally, they provide discounts, cashback, and loyalty rewards through accumulated points. Moreover, customers with certain loyalty levels can benefit from Top Priority CS.

As mentioned before, Tokopedia offers a Top Community program that values our sellers and their contribution to society. Additionally, we have a voice-of-customer program that allows us to directly address complaints from our customers. We also have a Tokopedia Care bestie program, which aims to appreciate and foster a sense of community among our loyal users. This program also encourages them to educate others about transactions on Tokopedia.

Hypothesis 3

The third hypothesis suggests that there is a positive relationship between value co-creation (function value (H3a), Hedonic Value (H3b), Social Value (H3c). to customer loyalty. Based on the bootstrapping result, H3a, H3b, and H3c have a P-value less than 0.05, and a T-statistic greater than 1.96. This supports the hypothesis. 

According to Wongkitrungrueng & Assarut (2018) [16], the hedonic value experienced in live streaming, such as enjoying the showcased products, plays a significant role in building trust towards both the products and sellers. Furthermore, the symbolic value of customer engagement also contributes to trust formation in live streaming. In conclusion, based on the respondents' feedback, the author concludes that customers purchase beauty products through live streaming in social commerce not solely for enjoyment, entertainment, or the experience itself, but rather because they have a genuine interest and need for these products to fulfill their necessities.

To enlarge customer loyalty, the company aims to create a positive and rewarding experience for its loyal customers, thereby encouraging them to continue their patronage. By providing personalized rewards and offering specialized customer service, the company can further solidify its relationship with its loyal customer base. To enhance customer loyalty, the company can implement various strategies that focus on providing exceptional customer service and addressing customer concerns. One effective approach is to actively seek feedback from customers, demonstrating the company's willingness to listen and attend to their needs. By actively addressing complaints and offering effective solutions, the company can build trust and loyalty among its customer base. Additionally, the company can show appreciation for customer loyalty by implementing a tiered rewards system. This system recognizes and rewards customers based on their level of loyalty and engagement with the company. For instance, loyal customers may receive special offers or exclusive discounts that are not available to the general public. Furthermore, the company can establish dedicated channels for customer service, specifically catering to the different tiers of customer loyalty.

Implication to Management

After conducting a survey and analyzing the Stimulus-Organism-Response model in the e-commerce industry from the customer's perspective, it would be beneficial for the author to consider conducting interviews to gain insights from companies. This will help understand how they can adapt and thrive in the rapidly changing industry and ultimately impact their strategic management decisions for achieving sustainability goals. Adopted from the interview results, the stimulus-organism-response (SOR) framework in e-commerce and its business implications for management regarding sustainability can be understood through the following key points:

1. Aesthetic Appeal and Layout 

Aesthetic appeal holds significant importance in order to cater to diverse age groups and users’ categories. AP said “Ensuring a good user experience and making it user-friendly greatly contributes to attracting e-commerce users”. The e-commerce industry recognizes the significance of attractive design and organization in its user interface (UI) and user experience (UX). Although there is no set timetable for updating the UI UX, changes are often prompted by customer feedback, such as difficulty in locating items or suggestions for improving the e-commerce platform.

1. Layout and Functionality

The company's strategy to increase functional value involves conducting Quality Assurance Testing, as well as testing the user interface (UI) and user experience (UX). They use a team swiping method based on user requests, typically from customers or the top community, to gather feedback and ideas for improving the product. This information is then shared with the business unit to implement the necessary changes.

1. Financial security

LP said “A PRD or Product Requirements Document is always prepared before launching a product on Tokopedia. This document includes objectives, features, and RCA (root cause analysis) to ensure the product's safety before it is launched”. As part of this effort, the company has introduced various insurance features, such as shipping, travel, automotive, non-suitability, and beauty protection, to ensure customers feel secure and protected. These features help mitigate potential risks and prevent customers from incurring additional expenses in case of any issues with their purchased goods. Tokopedia offers a Resolution Center to handle cases where customers receive inappropriate, lost, or missing goods. In such cases, Tokopedia acts as a middleman between the buyer and seller, providing insurance claims and refunds based on evidence provided by both parties.

1. Verbal communication

Utilizing live streaming platforms can boost marketing prospects for sustainable products, resulting in enhanced consumer trust and involvement. This approach facilitates authentic and interactive connections with a large audience, making it a powerful tool for marketing. It is recommended that businesses integrate live streaming into their marketing strategies to promote brand authenticity and foster greater customer engagement. On the other hand, “Engage with customers through Instagram live streams to address any inquiries or concerns about transactions on Tokopedia. Furthermore, reach out to users via phone calls to offer additional support for unresolved issues that were not resolved through live chat with customer service.” said AZ

1. Non-verbal communication

The company recognizes the significance of non-verbal communication methods such as project blueprints, correspondence, and historical data. They will ensure proper storage of these for future enhancements. They have developed a chatbot to improve customer service by addressing common queries and filtering out issues before involving human support. If the chatbot fails to resolve the problem, the customer service team will step in. Furthermore, all conversations will be evaluated and utilized to enhance e-commerce services in the future. Verbal Communication

1. Service skills

To create a strong sense of commitment and dedication, Tokopedia consistently reinforce their core values and service vision through activities like monthly quizzes and town hall events. They also have programs in place that involve and appreciate their customers, such as Tokopedia Care's Bestie, National Customer Day celebrations, Monthly Quiz on social media, and CX Summit. These programs aim to foster customer loyalty, highlight the quality of their services, and show appreciation to customers who help answer questions on social media and support the company's initiatives.

1. Customer Engagement

Customer engagement can be achieved by implementing an affiliate program on Shopee, where both the individuals who share a product and those who purchase it through the shared link are rewarded. The company values customer feedback and is eager to receive it. We are also committed to attentively addressing any complaints and providing effective solutions.

1. Function value

This ecommerce platform allows for personalized interactions based on customer preferences, making everyone feel like they are being catered to individually. Products and information are categorized and tailored to specific focus groups. The integration of consumers and organizations throughout the entire product process is crucial. The company understands that they need to embrace open innovation, incorporating knowledge and skills from various sources, in order to gain a competitive advantage.

1. Hedonic Value

LD said “To improve the customer experience in online shopping, the company has introduced different initiatives like gamification and providing customers with free shipping coupons, coins, and additional discounts”. “The use of virtual gifts on live streaming platforms is seen as a thoughtful strategy to encourage customers to make purchases and engage with the company. This approach has the potential to promote sustainable consumption and align with environmental responsibility” said AD. Live streaming commerce has been found to greatly enhance the shopping experience by allowing for more interaction online. This increased engagement can be used to influence customer behavior to receive more gifts and foster greater loyalty.

1. Social Value

Tokopedia offers a Top Community program that values its sellers and prioritizes customer feedback through the voice of customer program. They also have the Tokopedia Care Bestie program, which fosters a sense of appreciation and community among users. Loyal Tokopedia users are also encouraged to educate others about transactions on the platform. On the other hand, TikTok has an influencer community and live streamer community, with special programs that allow big brands to partner with live streamers to promote and sell their products online.

1. Customer Loyalty

Various programs are implemented in order to ensure that customers continue to make purchases on e-commerce platforms. These programs include special promotional events, updates that incorporate gamification elements, collaborations with popular artists or influencers, and the provision of rewards based on customer loyalty levels. “By implementing this tiering system, customers are encouraged to remain loyal to the company. For instance, loyal customers are offered exclusive deals and dedicated customer service channels that cater to their specific loyalty tier.” Said AC

Structural equation modeling is employed to validate these research frameworks:

1. For H1a, H1b, H1c, H1d, accepted as evidenced by the t count being greater than the t table.  Based on research, the author discovered that factors such as aesthetics, layout, functionality, financial security, and verbal communication positively influence customer engagement. However, nonverbal communication and service skills had a negative impact due to limited survey questions.

    

2. H2a, H2b, and H2c have been confirmed as valid because the t statistic exceeds the critical value. It has also been found that customer engagement is closely linked to the creation of value, encompassing functional, hedonic, and social aspects.

    

3. In addition, H2a, H2b, and H2c have been proven to be valid due to the t statistic being higher than the critical value. Additionally, it has been discovered that customer engagement is strongly connected to value creation, including functional, hedonic, and social elements.

First, the study used a case-crossover design, which has limitations in determining causality. Therefore, it is not possible to definitively establish the exact causal relationship between structures. To obtain more reliable results, future research should utilize long-term investigation methods and ensure that each variable is measured through multiple questions to determine its influence.

Second, the author used their personal social media accounts, specifically WhatsApp and Instagram with ads-on, to share the survey link and also relied on friends who were more likely to share it. This helped them gather a sample size of 1282 within four days. This method can be used in other research to increase participant involvement. In the future, the survey could be distributed through different methods, like events and online communities, to compare data with the current findings. 

Third, this study emphasizes the importance of real-time e-commerce services in building customer loyalty. The author proposes adopting a method from China's e-commerce industry and examines its potential application in Indonesia. However, it is crucial to acknowledge that various countries may have distinct business practices. To enhance the research, future studies could involve samples from multiple countries for comparative purposes.

Fourth, it is worth mentioning that real-time e-commerce services can have negative effects, such as negative feedback on product functionality. These negative effects may reduce the influence of real-time e-commerce services on customer loyalty. Future research can further investigate the negative aspects of real-time e-commerce services based on the findings of this study.

Fifth, for an e-commerce platform to become a leader in sustainability innovation, they need to consider the business implications of sustainability in terms of physical cues (aesthetic appeal, layout and functionality, financial security), social cues (nonverbal communication, verbal communication, service skills), customer engagement, value co-creation (function value, hedonic value, and social value) and last but not least, customer loyalty based on these research and specific programs from the company can help in understanding sustainable strategies and making strategic decisions. To remain competitive in the e-commerce industry, businesses should consider incorporating live streaming features, which allow for interactive conversations, price negotiations, and exclusive offers during live events.

Funding: No funding sources.

Conflict of interest: None declared.

Ethical approval: The study was approved by the Institutional Ethics Committee of Bandung Institute of Technology.

1. Zhang, Lu, et al. "The influence of greenwashing perception on green purchasing intentions: The mediating role of green word-of-mouth and moderating role of green concern." Journal of Cleaner Production 187 (2018): 740-750.https://doi.org/10.1016/j.jclepro.2018.03.201        
2. Thompson, Arthur, et al. EBOOK: Crafting and executing strategy: The quest for competitive advantage: Concepts and cases. McGraw hill, 2013.https://books.google.com/books?hl=en&lr=&id=7MovEAAAQBAJ&oi=fnd&pg=PR2&dq=Thompson,+A.,+Peteraf,+M.,+Gamble,+J.,+%26+Stricland,+A.+J.+(2016).+Crafting+and+Executing+Strategy:+The+Quest+for+Competitive+Advantage.+New+York,+USA:+McGraw-Hill+Education.&ots=pBLHeNLk7j&sig=ZJtL3e_9vIa4vmcx1XDGamBDR6Q 
3. Ma, Linye, Shuqing Gao, and Xiaoyan Zhang. "How to use live streaming to improve consumer purchase intentions: evidence from China." Sustainability 14.2 (2022): 1045.https://doi.org/10.3390/su14021045 
4. Ye et al. Journal of Innovation and Entrepreneurship (2023) 12:41 https://doi.org/10.1186/s13731-023-00310-2
5. Bitner, Mary Jo. "Servicescapes: The impact of physical surroundings on customers and employees." Journal of marketing 56.2 (1992): 57-71. https://doi.org/10.1177/002224299205600205
6. Bowden, Jana, and Abas Mirzaei. "Consumer engagement within retail communication channels: an examination of online brand communities and digital content marketing initiatives." European Journal of Marketing 55.5 (2021): 1411-1439.https://www.emerald.com/insight/content/doi/10.1108/EJM-01-2018-0007/full/html 
7. Bowden, Jana Lay-Hwa. "The process of customer engagement: A conceptual framework." Journal of marketing theory and practice 17.1 (2009): 63-74. https://doi.org/10.2753/MTP1069-6679170105
8. Brodie, Roderick J., et al. "Customer engagement: Conceptual domain, fundamental propositions, and implications for research." Journal of service research 14.3 (2011): 252-271. https://doi.org/10.1177/1094670511411703
9. Oliver, Richard L. "Whence consumer loyalty?." Journal of marketing 63.4_suppl1 (1999): 33-44. https://doi.org/10.1177/00222429990634s105
10. Ha, Young Wook, and Myeong Cheol Park. "Antecedents of customer satisfaction and customer loyalty for emerging devices in the initial market of Korea: An equity framework." Psychology & Marketing 30.8 (2013): 676-689. https://doi.org/10.1002/mar.20637
11. Srinivasan, Srini S., Rolph Anderson, and Kishore Ponnavolu. "Customer loyalty in e-commerce: an exploration of its antecedents and consequences." Journal of retailing 78.1 (2002): 41-50. https://doi.org/10.1016/S0022-4359(01)00065-3
12. Van Doorn, Jenny, et al. "Customer engagement behavior: Theoretical foundations and research directions." Journal of service research 13.3 (2010): 253-266. https://doi.org/10.1177/1094670510375599
13. Harris, Lloyd C., and Mark MH Goode. "Online servicescapes, trust, and purchase intentions." Journal of services marketing 24.3 (2010): 230-243. https://doi.org/10.1108/08876041011040631
14. Wang, Cenlan, et al. "How learner engagement impacts non-formal online learning outcomes through value co-creation: An empirical analysis." International Journal of Educational Technology in Higher Education 19.1 (2022): 32.https://link.springer.com/article/10.1186/s41239-022-00341-x 
15. Mw, Browne. "Alternative ways of assessing model fit." Testing structural equation models (1993).https://cir.nii.ac.jp/crid/1572261550791372416 
16. Vukadin, Ana, Apiradee Wongkitrungrueng, and Nuttapol Assarut. "When art meets mall: Impact on shopper responses." Journal of Product & Brand Management 27.3 (2018): 277-293.https://www.emerald.com/insight/content/doi/10.1108/JPBM-01-2017-1406/full/html 
17. Wijanto, Setyo Hari. "Structural equation modeling dengan Lisrel 8.8." Yogyakarta: Graha Ilmu (2008).

18. Chandrruangphen, Earth, Nuttapol Assarut, and Sukree Sinthupinyo. "The effects of live streaming attributes on consumer trust and shopping intentions for fashion clothing." Cogent Business & Management 9.1 (2022): 2034238.https://doi.org/10.1080/23311975.2022.2034238