Adoption of AI in retailing: Efficient operations and privacy concerns

Learning objectives and key issues

The case will support a wide range of learning objectives focussed on gaining an understanding of the following:

• Recognize the importance of digital technology and its impact on traditional businesses.

The problem

This case highlights a key ethical issue in the application of artificial intelligence (AI) technology: how shopping malls can harness the convenience and value that AI offers while safeguarding customers' rights and privacy. Striking a balance between technological advancement and privacy protection is not merely a technical challenge but a pivotal factor in building customer trust and ensuring sustainable business growth.

Company background

SKMall is a large-scale shopping mall that integrates various facilities, including shopping, leisure, culture, entertainment, and dining. With over 30 years of operational history, SKMall has expanded across six major cities in China, operating more than 20 stores. SKMall values the customer shopping experience and loyal customers and is especially eager to meet customers’ needs for high-quality items. By 2024, SKMall had established a membership base exceeding 3.5 million, with member transactions accounting for over 80% of the company’s total revenue. Connie, a manager of loyalty marketing, noted:

‘With the support of our effective customer loyalty program, our shopping malls offer an extensive and luxurious shopping experience, creating a warm and welcoming environment for our customers. We believe that if we understand the needs of our customers, we will have business opportunities. As a result, we seek to build a personalized marketing experience for each member that we could propose the right product, at the right time, to the appropriate customers’.

Over the past 25 years, SKMall has primarily relied on offline retail environments. The key to achieving exceptional sales performance has been the expertise of its counter staff, who possess extensive experience in managing memberships. These staff members are adept at understanding customer needs through product recommendations and excel in building and maintaining strong customer relationships.

Understanding in-store customer wants and needs

For SKMall’s management, counter staff are considered a vital core asset of the company. They play an essential role not only in promptly identifying customer needs but also in fostering long-term customer loyalty through consistent interaction. This contribution has been a key driver of SKMall’s strong sales performance.

In addition to relying on counter staff to understand customer needs through explanations and recommendations, SKMall aims to adopt more comprehensive digital tools to gain deeper insights into customers’ shopping motivations during their visits. This approach is intended to support individual brand counters in enhancing their sales performance. As a result, SKMall has investigated employing marketing technologies to acquire information about potential customers, analyse their behaviour, and better understand their preferences (Lemon and Peter, 2016). For example, while the company uses social media and QR code scanning tools to promote customer interaction, the results have not as anticipated as this technique requires precise customer profile, sufficient incentives, and alignment with customer preference to enhance customer engagement.

Furthermore, customers may be hesitant to disclose accurate personal information out of concern for their privacy. As a result, the firm lacks the capability to tailor and align these incentives to each brand and target customer group and design attractive reward schemes to collect customer information effectively. SKMall had also planned to use innovative technical solutions to gather potential customers’ profiles, stay time in various stores, and movement patterns inside the venue. For example, SKMall employs GPS positioning in malls and allows customers to activate mobile app navigation to assist them find the stores they wish to visit, or it uses Wi-Fi positioning technology to collect information about customers’ locations in malls. In addition to tracking customer locations, these technologies enable SKMall to get a more thorough picture of customer behaviour and optimize marketing campaigns, although the actual outcomes have been less than predicted. Sean, a manager of the digital development department, stated:

‘With the goal of guiding customers to spend money through digital technology, we investigated employing GPS or Wi-Fi to track the geolocation information of customers’ mobile phones. This would allow us to study their movement routes and the amount of time they spend in shopping malls. However, the reliability of the data was affected by things like customers' mobile phone habits and the mall's signal reception limitations’.

Apparently, SKMall has yet to successfully leverage technological means to understand customers’ needs upon visiting the store. In other words, SKMall’s understanding of and service to customers still relies heavily on the proactive efforts of brand counter staff, who create sales opportunities through professional advising and guiding. As a result, one of the critical challenges for SKMall is how to maximize insights into customers’ motivations for visiting the store and continue driving sales through online channels after they leave the store. According to Wang, a store manager:

‘We are constantly striving to understand our customers' motives and behaviour, as well as to discover their needs and wants. Are there any needs that we are unable to fulfil while competitors can? We must meet their demands and exceed their expectations with our services and products’.

A pivot to AI technology in retail

SKMall has previously drawn inspiration from smart retailing trends and the innovative practices of other industry players. By leveraging cutting-edge technologies such as big data and AI, SKMall has upgraded its traditional retail operations (Bradlow et al., 2017; Grewal et al., 2020). In 2017, when Amazon Go launched unmanned convenience stores using computer vision, deep learning, and sensor fusion technology, cameras and sensors in the store could track customer behaviour in real time. AI technology not only partially replaced traditional checkout services but also provided personalized product recommendations by analysing customers’ shopping behaviour and preferences (Cao, 2021; Fu et al., 2023; Guha et al., 2021; Polacco and Backes, 2018). This innovative model has prompted SKMall to further explore the potential of AI technology in the retail sector. Sean, a manager of the digital development department, said:

‘We are amazed and intrigued by the retail industry's combination of AI and computer vision technologies to identify customers' shopping behaviour in-store. In 2018, we learned through our suppliers' products that China's technology in this area is accelerating, and it is starting to be possible to scan faces for payment through cameras, as well as recognizing faces or voices to take a pre-designed response to the service. That is, when a customer walks into the store, we may be able to know who is coming...’.

Thus, Sean, a manager of the digital development department, introduced such ideas to several internal departments such as store management, merchandising, loyalty marketing, customer services, IT, and operations departments. In addition, Sean started to understand the challenges and potential needs from these departments and begin to seek appropriate solutions in the Chinese market. Of course, for SKMall, which emphasizes member management and shopping experience, what it hopes to do is to manage all aspects of mall planning, crowd monitoring, and public safety with the help of digital technology (Chen and Wang, 2023; Mokayed et al., 2023). David, a manager of the IT department, responded:

‘We have been using infrared sensor technology to track the real-time traffic in the malls, and using this data to calculate the shopper conversion rate, which is a key performance indicator (KPI) in retail. However, due to human interference and old equipment, the data is not accurate enough, affecting the business performance measurement’.

Jane, a manager of customer services, responded:

‘We began to focus on the mall’s space utilization and security. As a result, we urgently need a system that could objectively assess passenger flow, as well as better security surveillance within the mall so as to improve staff efficiency and our response time to emergencies’.

According to SKMall’s CRM analytics data, many customers initially visit mobile app malls to browse products before adding them to their shopping list. Surprisingly, approximately 82% of these customers later return to brick-and-mortar malls to experience the products before making a purchase. This implies that customers still require direct interaction with tangible products. Leo, a manager of the merchandising department, stated:

‘In the app, online malls can recommend products to customers through AI technology, which improves the matching efficiency. Is it possible to achieve the same level of efficiency in physical malls? For instance, assisting customers in identifying potential offers and desired items could be beneficial. Items that customers browse through an online mall but don't buy are usually missing the actual experience, like seeing, touching, or trying them out. Then, when the customer is at the mall, we need to provide that opportunity to the customer as precisely as possible, shortening the time she spends exploring and saving her the cost of consideration’.

SKMall, which has long prioritized the in-store experience, has made significant strides in its digital transformation over the past 5 years, developing a robust e-commerce platform. At SKMall, a common practice is the seamless integration of online and offline experiences. Examples include ordering from the online mall and picking it up in person or ordering from the physical mall and having it delivered home quickly. Currently, they are considering ways to integrate AI technology at its physical mall to better match people’s needs and goods in order to reach an optimal state. Sean, a manager of the digital development department, observed:

‘We can use technology to keep track of what products customers are interested in as they browse the mobile app mall. This helps us match the right goods with the right customers as precisely as possible. In a physical mall, on the other hand, we currently can't keep track of how customers browse. Because of this, we are thinking about adding AI to our physical mall, like computer vision eyes that can see and know what buyers want. We also want to use technologies that allow for personalized service across all channels. Customers can have the same ease of shopping and happiness whether they are in our physical mall or on our mobile app mall’.

A few weeks after the digital development department shared its thoughts on new technologies for retail, various departments expressed their opinions. Among these are pressing challenges that must be addressed.

How can privacy and technology coexist when machines are replacing humans?

SKMall leadership team believes that the key to digital transformation is to use digital technologies to improve current core capabilities rather than pursuing new technologies blindly. Adopting AI technology to understand customer demands is undoubtedly a technologically enabled business tool capable of efficiently addressing a wide range of needs (Guha et al., 2021; Inman and Nikolova, 2017). However, installing equipment in shopping malls to collect personal information for verification, identification, or analysis of facial recognition technology requires processing people’s personal bioinformation, which may violate their rights and interests (Chen and Wang, 2023; Middleton et al., 2022; Pizzi and Scarpi, 2020; Wang et al., 2024).

The legal department and external specialists have expressed concerns about the AI-powered passenger flow analysis solution.

‘Multiple cameras installed in shopping malls can record customer characteristics such as gender, age, clothing, items carried, and behaviour, primarily for data analysis on people's flow and operational efficiency enhancement. However, these cameras passively gather customers' faces and physical features, potentially breaching their personal privacy. Prior notification, customer approval, and controlled use are required when using such data for commercial purposes’.

In response to these concerns, the consultant asserted that the solution met the department’s sales and operational requirements and that customer privacy concerns would not arise if the technologies were used only for statistical and analytical purposes. To address the customers’ privacy concerns, customers will be mosaiced during AI data processing. As long as the data is ‘de-identified’, there is no privacy concern. Furthermore, if the processing of face information requires prior notification to gain consent from customers, the system will display a sign informing them of the purpose and extent of the data collection.

Despite these attempts, SKMall believes that the supplier’s ‘de-identification’ measures and notification strategy are only partial solutions to privacy issues. Even if the data is ‘de-identified’, there remains the possibility of re-identification. Moreover, merely notifying customers through signage fails to meet their expectations for choice. This passive acceptance approach could further erode customer trust in the mall. Jane, a manager of customer services, stated:

‘Simply putting up a sign saying “You are recorded by cameras” is not enough. Customers want to have the right to choose whether the mall collects their data, rather than being forced to accept it. Such a mandatory approach might make customers feel that their privacy is being disregarded’.

Jane, a manager of customer services, added:

‘Respecting customers' right to choose is the key to earning their trust. Passive notifications alone cannot meet the expectations of today’s consumers!’

After numerous meetings, Connie, a manager of loyalty marketing, sat at her desk surrounded by various feasibility reports. She was searching for a solution that could both enhance the customer shopping experience and safeguard their privacy. Deep in thought, she reflected:

‘If customers feel that our data collection methods violate their privacy, that trust could crumble quickly. We need a way to apply technology that makes customers feel respected’.

Meanwhile, Sean, a manager of the digital development department, was also contemplating the situation. As the digital lead, he constantly strives to balance cutting-edge technology with practical business needs. His role involves not only assessing technical feasibility but also addressing legal and ethical concerns. Over the past few weeks, Sean and Connie had been in frequent discussions, trying to find a breakthrough.

‘With the advancements in AI technology, we now have the ability to track customer shopping behavior and even make more efficient purchase recommendations within the mall. But if we can’t secure customers’ explicit consent, using this technology might create more problems than it solves’.

The core challenge for SKMall lies in balancing technological innovation with customer rights. While AI can create business value by analysing passenger flow, improving operational efficiency, and offering personalized services, its implementation may lead to a trust crisis if it fails to adequately respect customer privacy and informed consent. Specifically, collecting facial and behavioural data without prior notification and explicit consent could be perceived as infringing on customers’ control over their personal information (Adanyin, 2024; Guha et al., 2021; Wang et al., 2024). Indeed, in non-business operating public spaces, personal data might be collected to ensure public safety. According to the consulting company’s proposal, people visiting the mall would be informed that they are being recorded by cameras via a notice. However, such an approach does not provide customers the option to decline.

While these ideas sounded feasible, practical, and are being adopted by other industry players, the leadership team remained hesitant. As of now, no final decision has been made by the company.

Discussion questions

Questions specific for information systems students

Using Technology Acceptance Models (TAM) (Davis, 1989) or similar models to discuss:

(7) How might the mall’s employees and customers perceive the usefulness and ease of use of AI technology in their shopping experience?

Using EKB Model (Engel et al., 1968) of customer decision making that includes five stages (need recognition, information process, evaluation of alternatives, purchase, and post-purchase evaluation): to discuss:

(9) How can AI and computer vision technology enhance customer experiences within the shopping mall’s physical and digital spaces?