Technological Forecasting & Social Change 195 (2023) 122778

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

journal homepage: www.elsevier.com/locate/techfore

From my Perspective

Technology roadmap of AI applications in the retail industry

Hsi-Peng Lu a, Hsiang-Ling Cheng a, Jen-Chuen Tzou a, Chiao-Shan Chen b, *
a
Department of Information Management, National Taiwan University of Science and Technology, Taipei City, Taiwan
b
Department of Health Care Management, National Taipei University of Nursing and Health Sciences, Taipei City, Taiwan

A R T I C L E I N F O A B S T R A C T

Keywords: New retail, an online-merge-offline system with artificial intelligence (AI) applications, became popular after the
Artificial intelligence (AI) COVID-19 pandemic. This research introduces a novel technology roadmap methodology for new retail after
Technology roadmap studying 30 smart retail cases. Upon conducting panel discussions, this study identified 16 service applications,
Retail
10 technology applications, 4 application types, and 4 AI development stages. Based on the correlation analysis,
Technology development trend
it was found that new retail should: (1) aim towards the fourth stage of AI development, such as image gener­
ation technology; (2) develop a seamless shopping process; and (3) formulate cloud integration platforms
integrating data on consumer shopping behavior. Overall, the findings obtained could be promising reference for
future studies in the retail industry.

1. Introduction applications are initiated without understanding consumer needs. A

technology roadmap for the consumer shopping process is also needed.
Artificial intelligence (AI) is a significant contributor to many in­ New retail industries often fail to develop because of confusion. The
novations that are fundamentally reshaping the retail industry (Shankar retail industry needs to analyze consumer shopping behavior through AI
et al., 2021), especially during the COVID-19 outbreak. A change in to address this issue and help the retail industry to implement artificial
consumer spending habits has altered the dynamics of our global intelligence to enhance efficiency.
economy and indirectly affected human perspectives, forcing retailers to This study explored the development of AI-based service and tech­
rethink market needs. The retail sector is transforming as a result of nology applications during the epidemic through a case study approach
technological advances. Among the examples is the “click-and-mortar” to understand the futuristic trend of AI within the retail industry. The
business model, where business operations are online and offline (Yeh technology roadmap methodology was used to conduct case study
et al., 2020). Integrating online and offline operations will undoubtedly analysis, encode cases, and help retailers achieve AI integration. The
revolutionize the retail industry and allow consumers to shop differ­ aims of this study are (1) to present a more comprehensive service and
ently. The technology predicts consumer demand and provides retailers technology relationship methodology for AI applications in new retail
with goods and services that help them create value (Pillai et al., 2020). based on a technology roadmap; (2) to explore the development of
Moreover, Wadhwani and Gawande (2023) predicted that retail AI ap­ service and technology applications of AI in stages of the different
plications will surpass USD 100 billion by 2032. shopping processes; and (3) to investigate AI development recommen­
The integration between online and offline operations will un­ dations in service and technology roadmap. Overall, the findings
doubtedly bring radical changes to retail industries. As traditional re­ demonstrated that the research methodology obtained could be a
tailers busily and rapidly arrange their distribution channels, they also promising reference for future studies in the retail industry.
neglect to optimize the shopping experience for their consumers (Zhai
et al., 2019) and the enormous threat posed by the Internet (Hart et al., 2. Literature review
2000). To break out of this dilemma, retailers have taken the initiative to
adopt AI technology to clear any unsatisfied consumer expectations This study examines the technology roadmap for AI applications in
during their shopping experience and emphasize a new form of value- the emerging retail industry. Our literature review consists of three main
added shopping experience. However, many services and technology sections: (i) the retail shopping process and the concept of new retail, (ii)

* Corresponding author at: Department of Health Care Management, National Taipei University of Nursing and Health Sciences, No. 365, Mingde Rd., Beitou Dist.,
Taipei City 112303, Taiwan
E-mail address: [email protected] (C.-S. Chen).

https://doi.org/10.1016/j.techfore.2023.122778
Received 5 May 2022; Received in revised form 10 May 2023; Accepted 3 August 2023
Available online 26 August 2023
0040-1625/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

artificial intelligence, and (iii) the theoretical basis for the technology integration of several (McCarthy et al., 2006). Rich (1983) defines AI as
roadmap. a study investigating how to use computers to do things better than
humans (Ertel, 2018). According to Duncan, vice president at Gartner,
2.1. New retail who has extensive experience in data-driven digital transformation and
analysis of changes in consumer behavior, including retail and financial
China and many Western countries are experiencing retail industry services. He indicated that data is the key to intelligence in the digital
changes. There were 12,200 brick-and-mortar store closings in the U.S. world. Pettey (2018) also emphasized the importance of AI-conforming
in 2020 (Wahba, 2021). PricewaterhouseCoopers and Kantar (2020) data management in building professional knowledge. Previous studies
predicted that new and smaller stores would emerge as the global retail confirmed that data is the foundation of AI development and AI has
market moves towards seamless new retail. Nevertheless, new retail transformed the global economy and begun a new phase.
could create value-added products and services using big data analysis Lee (2019) stated that AI development would undergo four waves.
and cloud computing. Innovative technology applications can help re­ The first wave would be Internet intellectualization, in which businesses
tailers predict consumer needs and recommend products based on their yield profit from data analysis via the Internet. The second wave is
preferences. It also can help suppliers optimize inventory management, business intellectualization, which uses existing data from enterprises to
logistics, and transportation (Shankar et al., 2021). Most studies create commercial value. A third wave is perception AI, in which facial
analyzed retail consumers or supply chains through mathematical recognition, voice recognition, and other sensory devices are used to
modelling (Liu et al., 2023; Wu, 2021). However, there was a lack of collect data. The fourth wave is autonomous AI, in which machines
comprehensive analysis of the industry. This study focused on con­ operate independently to replace intensive labor. Additionally, Lu
sumers using a technology roadmap by identifying the key elements (2018) also a framework that defined the stages of AI development. AI
such as service applications, technology applications, application types, development encompasses four primary areas - the Internet, Internet of
and the four stages of AI development to generate a technology roadmap Things (IoT), recognition technology, and interaction and gambling.
of AI applications in the retail industry. This paper discusses future interactions between AI technology and
humans or the environment. Our study further expands upon Lu (2018)‘s
2.1.1. Retail shopping process framework to help understand how AI is evolving in the new retail
Consumer shopping is divided into three phases by Lemon and sector.
Verhoef (2016): pre-purchase, purchase, and post-purchase (Hoyer During the first stage of AI development, immense amounts of data
et al., 2020). The study used checkout behavior as the dividing point of a are collected from users' Internet footprints, analyzed, and generalized
shopping process and formulated enterprise responses to different pha­ to generate recommendations and decision-making applications. The
ses of consumer purchase behavior. Upon encountering the product, second stage moves from purely online to offline. AI collects data from
consumers enter the pre-purchase phase until they pay. Customers are sensors so that it can “communicate” between objects and generate
motivated to purchase products through product searches, price com­ personalized services. The third stage uses non-figurative data to build a
parisons, and channel strategies. According to Rhee et al. (2009), database using IoT technology. Alternatively, deep learning algorithms
“search” is an online event that fulfills a specific shopping desire, which can be utilized for repeated self-training. The fourth stage of AI devel­
can be facilitated through image recognition, keyword search, or opment features service applications based on image and voice recog­
personalized advertising recommendations. “Comparison” refers to nition. AI can then interact actively with the environment to predict the
finding and comparing a specific product that fulfills the consumer's user's needs. In summary, retailers should provide services and tech­
needs and then looking for similar products or services (Li et al., 2020). nology applications to consumers at different AI stages under different
“Channel” refers to brick-and-mortar stores, online shopping malls, and conditions (Chari et al., 2017).
mobile applications (Hoyer et al., 2020) to allow consumers to shop
when and where they want. All of the above are pre-purchase actions 2.3. Technology roadmap
taken by consumers prior to checkout.
During the purchase phase, consumers pay and receive their goods, Technology roadmaps are popular for predicting technology trends
and enterprises can optimize the checkout process for checkout, pay­ and managing research and development (R&D) (Amadi-Echendu et al.,
ment, and collection. In addition to offering consumers diverse payment 2011). Technology roadmaps can accelerate technology commerciali­
options and multiple channels for collecting purchased goods, enter­ zation by analyzing market needs and current technologies (Islam et al.,
prises use image detection and self-checkout to reduce time spent on 2020). Hence, businesses can determine their medium- and long-term
checkout (Leng and Wee, 2017). In the post-purchase phase, customer business strategies and goals (Lu and Weng, 2018) to achieve market
service specialists and virtual assistants automate the return and ex­ penetration objectives and technological advancement.
change process after checkout (Cui et al., 2017) to automate the overall Technology roadmaps usually consist of three main layers. The top
post-checkout process. Madakam et al. (2019) demonstrate how logistics layer includes trends, overall objectives, and factors behind market de­
robots and robotic arms reduced transportation times and improved mand. The middle layer consists of products and services developed
business efficiency by replacing traditional staff. based on the elements in the top layer. Moreover, the bottom layer in­
Next, this study will investigate the impact of AI on consumers at the volves internal/external resources and technologies (Phaal et al., 2010).
purchase stage through pre-purchase, purchase, and post-purchase. Since enterprises often lack hands-on application guidance when
developing technology roadmaps, Phaal et al. (2004) developed the T-
2.2. Artificial intelligence plan process to stimulate technology roadmap development by identi­
fying the priority of new products and technologies. Technology road­
In 1956, the Dartmouth Conference established the birth of AI (Soni maps have continued to be widely used in business and management
et al., 2020); in addition, the world economic discourse also describes it since Walsh (2004) demonstrated their relevance through a case study
as the centerpiece of the world's current technological revolution as it, (Huang et al., 2018). Tierney et al. (2013) proposed a new technology
along with a host of complementary new technologies (Moore et al., roadmap after multiple trials to compensate for the previous focus on
2022). The technology lays the foundation for radical business advances single-root technology.
that enhance the consumer shopping experience and streamline delivery Antonopoulos et al. (2020) identified AI-based forecasting as a po­
services, while facilitating the creation of new business models. With its tential critical solution for determining strategic needs and under­
innovative applications, the retail industry can be transformed intel­ standing technology trends. Studies in the past tended to focus on the
lectually. However, AI is not a single technology but rather an medical field (Ramos et al., 2022; Tierney et al., 2013), yet, it is

2
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

important to consider intelligent services (Han and Geum, 2022). see how AI applications are used in various shopping processes. A
Despite the importance of AI in retail, few reports exist on the topic. consensus is reached after repeated inventory and cross-checking.
Thus, the technology roadmap is used in this paper to forecast AI-based Finally, 16 service applications, 10 technology applications, 4 applica­
technology and services. With the help of a technology roadmap, this tion types, and 4 AI development stages have been categorized.
paper projects AI application development across various stages of the The third stage aims to develop a technology roadmap for AI appli­
retail process. The key elements are service applications, technology cations in new retail. Cases were analyzed from multiple perspectives:
applications, application types, and AI development stages. The results service application, technology application, type of application, and AI
suggest strategies for retailers, services, and technology providers. stage. The number of services and technology applications with partic­
ular key technologies that appeared in the same case was analyzed. Data
3. Research methodology from these sources determine whether a technology or service is widely
used. Then, a comprehensive technology roadmap of AI applications in
There are single-case and multiple-case studies (Yin, 2009), with the new retail was generated by connecting the appendix items (see Ap­
latter offering more robust research (Silva and Mercês, 2018). In this pendix D and Appendix E). A strategy for developing new retail based on
multi-case study, 30 cases are examined from different perspectives to AI applications is then proposed according to the technology roadmap
explore the phenomenon of AI applications in the retail industry. results.
Additionally, Lu and Weng's (2018) technology roadmap was adopted to
extend the framework and apply AI technology to new retail services 4. Results and analysis
and technologies. Eisenhardt (1989) argued that case study researchers
could assist in fact-checking. Therefore, four experts were invited Various perspectives were examined in this paper. In this section, we
(including a project manager with up to ten years of retail experience list 16 service applications, 10 technology applications, 4 application
and three researchers with two to four years of retail experience) to re- types, and 4 AI development stages to conduct a detailed analysis and
examine and confirm analysis results to ensure coherence. A total of 16 show in order the overview of developments in the new retail industry
service applications, 10 technology applications, 4 application types, (including the analysis of AI development stages and application types)
and 4 AI stages have been uncovered. and details of technical implementation (including the analysis of new
retail industry cases, service and technology applications).

3.1. Research process 4.1. An overview of developments in the new retail industry

This study proposed three stages, as illustrated in Fig.1. A list of 30 The experts counted the number of AI development stages and ser­
real-world new retail cases (Appendix A) incorporating AI technologies vice and technology applications in the shopping process to understand
and services is selected for data collection and validation (first stage). the current status of AI in the retail industry.
Most cases were gathered from press releases and public statements by
government agencies or prominent international exhibitions. The 4.1.1. Analysis of AI development stages
background and technical details of each case were described through Table 1 shows a summary review of the 30 cases based on the
digital media coverage. Through the integration of online and offline
services, each case focuses on improving the customer experience. The
Table 1
cases were selected because of (1) the involvement of AI applications in
A review of AI development stages.
the real world; (2) active in the market for at least one year during the
COVID-19 epidemic, demonstrating consumer acceptance; and (3) Number of Number of Total number of
domestic cases foreign cases cases collected
gradually accepted by consumers, suggesting that the services and collected collected
technology applications offered may change accordingly with techno­
The First Stage of AI 2 2 4
logical advances and consumer demand. This study also provides a
development
preliminary classification and definition of the technology and service The Second Stage of 1 1 2
applications related to these cases. AI development
The second stage focused on reviewing and analyzing key elements The Third Stage of AI 7 17 24
development
of the new retail sector. Four new retail experts encoded and categorized
The Fourth Stage of 0 0 0
the research cases separately to prevent blind spots or personal subjec­ AI development
tivity. As part of this step, experts examined how often the key elements - Total number of 10 20 30
service applications (Appendix B), technology applications (Appendix cases collected
C), application types, and AI development stages - appear in each case to

Fig. 1. Service and technology correlation methodology.

3
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

previously mentioned four stages of AI development. The results show less varied and frequently used. Hence, retailers should prioritize
that the first and second stages are relatively mature. In the third stage, introducing logistics robots, virtual assistants, or robotic arms to mini­
most cases are based on IoT or deep learning applications and expanded mize consumer wait time for product delivery.
to serve consumers. Although some applications integrate image By analyzing the shopping process (Fig. 3), this study found that
recognition and virtual assistants, none have achieved genuine AI- most technology applications cover three phases of the shopping pro­
driven interactive gambling. The fourth stage is considered a potential cess. More than half of the cases were analyzed using big data, cloud
market for AI solutions, focusing on image and voice recognition. computing, and machine learning. Still, these items are based on existing
technologies. Machine learning and deep learning are AI core technol­
4.1.2. Analysis of application types ogies. Despite the success of machine learning in the post-purchase
After examining checkout behavior, this study divided the shopping phase, deep learning has yet to be developed. As a result, this study
process into three phases: pre-purchase, purchase, and post-purchase. suggests that AI-related technologies should be fully integrated into
Each case may cover more than two (or more) shopping processes. post-purchase phases when developing innovative applications.
Thus, a case could involve multiple shopping phases simultaneously. Retailers can use several technology applications to provide con­
Fig. 2 shows that most cases are still in the pre-purchase phase. Further, sumers with innovative technology services. Contrarily, the purchase
most cases only applied AI solutions to the pre-purchase phase and and post-purchase phases employ fewer service and technology appli­
primarily to a single shopping phase. Integrating the pre-purchase and cations. Automated purchase phases and electronic payment methods
purchase phases was the most common combination in integrative create centralized customer data management. For the post-purchase
shopping. Enterprises could use consumer shopping behavior data to phase, this study advocates introducing logistics robots and mechani­
recommend personalized products and services. By offering diverse cal arms to replace labor needed for repetitive tasks or heavy-lifting jobs
payment options, enterprises can better understand consumer prefer­ such as transporting heavy cargo. Overall, new retail should focus on the
ences. Therefore, a “seamless” shopping strategy should include inte­ Third Stage of AI development by focusing on pre-purchase applications
grating the three stages of the shopping process, extending after-sales from the perspective of AI development stages and application types.
service after checkout, and reducing shipping times.
Fig. 2 shows the analysis of service applications from the perspective
4.2. Details of technical implementation
of the overall shopping process. Pre-purchase applications were found to
be the most common. Specifically, enterprises optimized pre-checkout
This study examines the technological implementation and relevance
behavior increased chances of contacting product consumers and
of the applications to obtain a more detailed analysis.
enabled consumers to locate products fast based on their needs. While
there are few service applications in the purchase phase, self-checkout
4.2.1. Analysis of new retail industry cases
and electronic payments are found in many cases, suggesting their
This study examined the number of times service and technology
popularity has gradually increased. However, post-purchase services are
applications were used across 30 cases. The cloud integration platform is

Fig. 2. Application type review and service application type analysis diagram.

4
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

Fig. 3. Technology application type analysis diagram.

the most popular among service applications (Fig. 4). Without the cloud 4.2.2. Analysis of new retail service and technology applications
integration platform, retailers cannot collect consumer data. Therefore, Appendix D illustrates the number of times different service and
service providers must prioritize the development of cloud integration technology application items were applied to the same case. Cloud
platforms. This study recommends providing consumers with the op­ integration platforms (mostly paired with big data analysis, 25 times)
portunity to experience innovative technology services through logistics have the highest service correlation count, indicating they are core ap­
robots, virtual and augmented reality applications. plications. A total of 25 cases were applied to cloud integration plat­
Fig. 5 shows a similar analysis from the perspective of technology forms and big data analysis, cloud computing (23 times), and machine
providers. Big data analysis was the most frequently applied technology learning (18 times). Thus, the study suggests that service providers of­
and a key technology at the core. Technology providers interested in fering cloud integration platforms should collaborate with providers of
expanding can use big data analysis to develop new technology appli­ the above three technologies to develop a solid foundation for cloud
cations and create new shopping experiences. For retailers, partnering integration platforms. Appendix E shows the number of times different
service providers are sought for purchases, collaborations, or indepen­ technologies corresponded in the same case. Big data analysis as the core
dent development. technology had the highest correlation count. Other technologies that
Based on an analysis of services and technologies, new retail has are often used in conjunction include cloud computing (21 times), ma­
focused more on acquiring large amounts of consumer data than chine learning (17 times), and deep learning (16 times), suggesting the
diversifying service applications. The third stage of artificial intelligence importance of big data analysis in developing new retail. Hence, the
development focuses on pre-purchase applications. Consumer shopping collaboration between technologies used for big data analysis is
behavior data is used to create a database, which is then trained possible. Image generation technology, in contrast, has the lowest total
repeatedly to improve recognition accuracy and predict what consumers count when combined with deep learning (1 time), image recognition (1
want so that an enterprise can deliver a more enjoyable shopping time), and machine learning (1 time). This may be because image gen­
experience. Currently, fewer service applications are readily accessible eration technology needs to mature and its application is less visible in
to consumers due to immature technology and consumers' unwillingness the market, with fewer research cases available.
to embrace such applications.

Fig. 4. The number of times service applications appeared in the cases.

5
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

Fig. 5. The number of times technology applications appeared in the cases.

5. Conclusion and discussion Nevertheless, its application should not be pushed too hard too quickly.
Customers will prevent them from easing into it or may even trigger a
This paper identified 16 service applications, 10 technology appli­ rejection response, causing the business to lose its chances. Technology
cations, 4 application types and 4 AI stages, eventually leading to a providers should use image and voice processing technologies to launch
technology roadmap. Aside from suggesting 3 possible sequences for a interactive applications and image processing services in the future,
crucial development order, it also proposes a development order for which will gradually lead retailers and service providers towards the
potential key technologies. Furthermore, the study provides recom­ fourth stage of AI development.
mendations regarding AI integration in New Retail, which can serve as a
reference for retailers and service/technology providers. 5.2. Proposals for the development of AI service and technology roadmap

5.1. Proposals for the development of AI integration in new retail In summary, this study extends a service and technology roadmap of
AI applications (Fig. 6) for retailers, service and technology providers to
5.1.1. New retail should move towards the Fourth Stage of AI development consider which applications should be prioritized to lay the foundation
As the COVID-19 pandemic continues, it is clear that the first and for the new retail industry.
second stages of AI development have matured. In most cases, domestic
and international retailers fall under the third stage, which uses IoT and 5.2.1. The Cloud integration platform is fundamental to service applications
deep learning to personalize services. In the Fourth Stage, service ap­ In new retail, accumulating consumer data is more important than
plications are based on image/voice recognition technologies. As a developing service applications (Geiger and Gross, 2021). Cloud inte­
result of these findings, this study contends that the new retail industry gration platforms have become indispensable service applications to
is currently in the third stage of AI development. Developing image/ collect data on consumer features or purchase information across the
voice recognition technologies and interactive applications will be the entire shopping process. The priority should be introducing big data
next step in developing the fourth stage as a potential market. analysis, cloud computing, and machine learning when developing
cloud integration platforms. There is a correlation between big data
5.1.2. New retail should focus on the purchase and post-purchase phases of analysis and cloud computing. Without the data collection, analysis, and
development processing performed by big data analysis, it would be impossible to
By applying AI-based technologies and services to retail shopping, compute, access, and share data statistics or perform prediction and
consumers can have a smoother shopping experience (Pillai et al., analysis on such data and share the results with other users, even if
2020). In most cases, services and technology applications are still in the having a mature cloud application environment (Gavrila and de Lucas
pre-purchase phase, concentrating on collecting consumer shopping Ancillo, 2021).
data. The purchase phase is less diverse, but electronic payment and self-
checkout are increasingly popular, with over USD 60 billion in trans­ 5.2.2. Big data analysis, cloud computing, and machine learning are
actions per year (Xiao et al., 2022). In the post-purchase phase, there are essential to technology applications
relatively few service and technology applications, and AI-based tech­ Big data analysis, cloud computing, and machine learning are the
nologies are still in their infancy. The reason is that there are fewer cases most commonly adopted key technologies during shopping. Big data
of landing applications post-purchase, where enterprises seek to analysis and cloud computing are technologies that are most closely
improve their development. The advent of AI forces traditional retailers related. In contrast, big data analysis and cloud computing are the
to adjust their development strategies and experience a virtual-real technologies most correlated with machine learning. Thus, technology
“seamless” integration of a comprehensive shopping experience providers must develop three essential technologies: big data analytics,
(Ogunjimi et al., 2021). cloud computing, and machine learning, if partnering with service
Retailers should prioritize assisting their growth to address imbal­ providers or retailers to create innovative consumer experiences.
anced development in the current shopping process and treat the pur­
chase and post-purchase phases as the primary focus. The study also 5.2.3. Image generation technology is key to developing technology
recommends gradually incorporating AI-based services and technology applications in potential new retail markets
applications. In addition to introducing consumers to innovative tech­ New retail can leverage image recognition technology for interaction
nology (Shankar et al., 2021), AI technology can also improve the effi­ and gambling as a potential market for the future. The study found that
ciency and quality of different shopping stages (Bughin et al., 2017). image generation technology had the lowest correlation count, which

6
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

Fig. 6. Service and technology roadmap of AI applications in the new retail industry.

placed it in the fourth stage of AI development. Technology providers to collect data on consumer behavior and create a new shopping expe­
should reference Fig. 6 and integrate image generation technology into rience for consumers while optimizing their pre-purchase process when
the pre-purchase phase to provide consumers with innovative experi­ shopping.
ences (Simoni et al., 2022). An overview of 30 AI application case studies in the global market is
Cloud integration platforms are a top priority for retailers and service presented in this paper, which focuses on the latest developments in AI
providers working with AI technology for the first time or with limited application cases. It is possible that each case is limited by the number of
budgets. On the other hand, technology providers should prioritize media reports available and may not apply to other industries due to the
developing more mature technologies such as big data analysis, cloud expert's bias when selecting cases for study. Four experts were recruited
computing, and machine learning. The application of AI to traditional for this study to perform repeated examinations and verifications of the
retail is enhanced with the ability to collect shopping process data and paper's proposed operational definitions and content descriptions to
predict consumer behavior. COVID-19 has led to government re­ eliminate any subjective bias or prejudice.
strictions on out-of-home purchases and social distancing, increasing
consumer acceptance of online shopping and innovative technologies Data availability
(Pantano et al., 2020). This peculiar period presents an opportunity for
retailers who desire to introduce image-generation technologies to their I have shared my data at the Attach Files step.
businesses to develop innovative technology applications. The reason is

Appendix A. Case selection

No. Domestic/foreign Case name

1.1 Domestic VesCir Ltd. Skin Analyzer Device - défiderm

1.2 FamilyMart Technology Concept Store
1.3 PXGo! PX-Mart Online Shopping
1.4 Appier Inc. CrossX User Acquisition Platform
1.5 Fruitpay Smart Fruit Vending Machine (2nd Gen)
1.6 A.S⋅O Dynamic Foot Pressure Measurement
1.7 SkyREC Business Intelligence System
1.8 7-ELEVEN X-store
1.9 Viscovery Visual Checkout System
1.10 Aiello Intelligent Voice Assistant for Hospitality
2.1 Foreign Amazon Echo Smart Speaker
2.2 Amazon GO Store
2.3 Aspectiva Product Review Analysis
2.4 FedEx SameDay Bot
2.5 Google Home Smart Speaker
2.6 Kroger EDGE Digital Shelves
(continued on next page)

7
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

(continued )
No. Domestic/foreign Case name

2.7 Lawson Reji Robo Automated Checkout System

2.8 Modiface Augmented Reality
2.9 Mujin Intelligent Piece Packing Robot
2.10 Samsung Family Hub 2.0 Smart Fridge
2.11 Stitch Fix Personal Styling Service E-commerce Platform
2.12 Vaak AI Theft-detection System
2.13 ViSenze ViSearch API Search
2.14 Wal-Mart Distribution and Logistics System
2.15 Wasteless Dynamic Pricing Tags
2.16 Tmall Genie Smart Speaker
2.17 MoBagel Decanter AI Big Data Analysis Engine
2.18 JD.com 7Fresh Supermarket
2.19 HaiDiLao Robotic Smart Hotpot Restaurant
2.20 Hema Fresh Food Supermarket

Appendix B. Service applications and definitions

No. Item Operational definition

S1 Passenger Flow Analysis Images are obtained from cameras in brick-and-mortar stores and then used for consumption data analysis and evaluation; results should
provide insight and reference for future business activities (Günther, 2017).
S2 Self-checkout In a brick-and-mortar store, consumers are able to complete the checkout process and pay for their purchase without requiring a cashier to
operate the checkout register at a brick-and-mortar store (Fernandes and Pedroso, 2017).
S3 Mobile Application (App) Consumers are provided with software services for their mobile devices, such as a smartphone or a tablet, enabling them to, at any time,
shop online, access their membership information, or benefit from utilizing features developed by the store to fulfill specific needs (Karch,
2020).
S4 Logistics Robot Physical robots that are employed in the supply chain process for automated cargo storage and logistics to reduce the consumer's waiting
time for order delivery (Shankar et al., 2021).
S5 Service Robot A human-like robot device that is capable of responding differently towards different changes in the environment. (Brengman et al., 2021).
S6 Personalized Recommendation Businesses collect large amounts of consumer data for prediction and analysis and create personalized marketing content for individual
consumer needs (Chiranjeevi and Vishwanath, 2016).
S7 Product Customization Offering exclusive and customized product design services based on each consumer's individualized preferences and needs, providing
consumers with personalized product options that fit the individual's specific needs (Díaz-Martín et al., 2019).
S8 Virtual Assistant An AI-based software service that helps enterprises grow their business using natural language processing (Rai et al., 2019).
S9 Virtual Reality Application (VR) Creates a virtual 3D experience on the computer to give the user an immersive experience (Bonetti et al., 2019).
S10 Cloud Integration Platform Manages data from different systems and devices in a shared network environment to reduce the response time for solving consumer
problems (Rai et al., 2019).
S11 Electronic Payment The exchange of non-cash transaction information between consumers, enterprises, and financial institutions via the Internet, banks, or
their counterparts (Chaudhry and Khan, 2016).
S12 Electronic Shelf Label (ESL) The replacement of traditional paper labels with a physical, electronic display device that utilizes diverse information and dynamically
updated features to display product prices and related information in real-time (Yang and Lee, 2018).
S13 E-Tag Reads data through a physical media on device carriers with built-in Radio Frequency Identification (RFID) or Near Field Communication
(NFC), whose chip passwords cannot be copied, given its high-security protection (Fernández-Caramés and Fraga-Lamas, 2018).
S14 Security Monitoring System While storing video recordings from surveillance cameras, the system analyzes the image contents simultaneously. It initiates sending
warning messages to the supervisors upon identifying abnormal images, thus providing a quicker response to the crisis at hand (Casola
et al., 2019).
S15 Robotic Arm Applications An automatic control device that imitates the function of a human arm and can perform laborious or tedious tasks independently (Nag
et al., 2018).
S16 Augmented Reality Applications Overlaying virtual images in the natural environment allows consumers to see an augmented virtual and real image (Bonetti et al., 2019).
(AR)

Appendix C. Technology applications and definitions

No. Item Operational definition

T1 Big Data Analysis Processing large amounts of diverse data to create valuable data or make predictive decisions (Hofmann and Rutschmann, 2018).
T2 Natural Language Processing Using computer programs to recognize, understand, and generate language, including the use of automatic question-answering systems, voice
(NLP) recognition, and machine translation (Kumar et al., 2016).
T3 Deep Learning Using artificial neural networks as the basic framework to run multi-level processing on a large number of disordered signals, converting them
into usable, featured data as a problem-solving approach (Wei et al., 2020).
T4 Barcode Scanning Identifying QR codes and barcodes using a barcode scanner or smartphone to obtain information such as the price of a product (Wei et al.,
Technology 2020)
T5 Cloud Computing Utilizing the Internet for computational operations, accessing shared hardware/software resources, and sharing information with other end-
users on demand (Rai et al., 2019).
T6 Sensor Technology Utilizing sensory hardware to detect and collect data; converting such data into signals that can be processed by hardware for the purpose of
equipment control and data processing (Liu et al., 2019).
T7 Remote Application Monitoring or controlling devices' connection in different places through the Internet (Stark et al., 2020).
Technology
T8 Image Generation After accessing and reading an icon image through the camera, the computer program will analyze the different angles and positions of the
Technology reading and then display the analyzed image on the screen (Wei et al., 2020).
(continued on next page)

8
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

(continued )
No. Item Operational definition

T9 Image Recognition After repeated attempts and training, the deep learning module extracts and analyzes the target features through image processing or sensor
Technology technologies (Chen et al., 2020).
T10 Machine Learning The machine uses the analyzed data or past experiences to compute the closest-matching answer. The goal is for the algorithm to train itself to
improve its performance, elevate the standard for self-learning, and make predictions (Tramèr et al., 2016).

Appendix D. Service application correlation matrix

T4. Total count

T1. Big T2. Natural T7. Remote T8. Image T9. Image T10.
T3. Deep Barcode T5. Cloud T6. Sensor of key
Data Language Application Generation Recognition Machine
Learning Scanning Computing Technology service
Analysis Processing Technology Technology Technology Learning
Technology correlation
S1. Passenger Flow
5 1 4 6 7 6 3 0 6 4 42
Analysis
S2. Self-Checkout 6 2 4 5 6 7 6 0 8 4 48
S3. Mobile
7 2 3 5 8 6 5 0 5 4 45
Application
S4. Logistics Robot 0 0 0 0 0 1 0 0 1 1 3
S5. Service Robot 3 2 0 2 2 2 3 0 2 2 18
S6. Personalized
13 4 9 3 10 6 2 1 4 9 61
Recommendation
S7. Product
2 1 0 1 0 2 2 0 1 0 9
Customization
S8. Virtual Assistant 7 7 5 2 5 7 3 0 2 5 43
S9. Virtual Reality
1 1 0 1 1 1 1 0 1 1 8
Application
S10. Cloud
25 8 16 11 23 16 8 1 13 18 139
Integration Platform
S11. Electronic
10 5 7 7 10 10 7 0 7 6 69
Payment
S12. Electronic
4 1 2 5 5 4 3 0 4 4 32
Shelf Label
S13. E-tag 4 2 1 3 2 4 4 0 4 2 26
S14. Security
2 0 2 1 1 1 1 0 2 1 11
Monitoring System
S15. Robotic Arm
2 2 1 2 1 2 2 0 2 2 16
Applications
S16. Augmented
1 1 0 0 0 0 0 1 1 1 5
Reality Applications

Appendix E. Technology application correlation matrix

T1 T2 T3 T4 T5 T6 T7 T8 T9 T10

T1. Big Data Analysis - 8 16 9 21 13 8 1 12 17

T2. Natural Language Processing 8 - 5 2 6 7 3 0 2 6

T3. Deep Learning 16 5 - 4 14 9 4 1 9 14

T4. Barcode Scanning Technology 9 2 4 - 10 8 6 0 6 5

T5. Cloud Computing 21 6 14 10 - 12 6 0 10 17

T6. Sensor Technology 13 7 9 8 12 - 7 0 9 10

T7. Remote Application Technology 8 3 4 6 6 7 - 0 7 4

T8. Image Generation Technology 1 0 1 0 0 0 0 - 1 1

T9. Image Recognition Technology 12 2 9 6 10 9 7 1 - 9

T10. Machine Learning 17 6 14 5 17 10 4 1 9 -

Total count of key technology
105 39 76 50 96 75 45 4 65 83
correlation

References Bonetti, F., Pantano, E., Warnaby, G., Quinn, L., Perry, P., 2019. Augmented reality in
real stores: empirical evidence from consumers’ interaction with AR in a retail
format. In: Augmented Reality and Virtual Reality: The Power of AR and VR for
Amadi-Echendu, J., Lephauphau, O., Maswanganyi, M., Mkhize, M., 2011. Case studies
Business, pp. 3–16.
of technology roadmapping in mining. J. Eng. Technol. Manag. 28 (1–2), 23–32.
Brengman, M., De Gauquier, L., Willems, K., Vanderborght, B., 2021. From stopping to
https://doi.org/10.1016/j.jengtecman.2010.12.002.
shopping: an observational study comparing a humanoid service robot with a tablet
Antonopoulos, I., Robu, V., Couraud, B., Kirli, D., Norbu, S., Kiprakis, A., Flynn, D.,
service kiosk to attract and convert shoppers. J. Bus. Res. 134, 263–274.
Elizondo-Gonzalez, S., Wattam, S., 2020. Artificial intelligence and machine learning
Bughin, J., Hazan, E., Ramaswamy, S., Chui, M., Allas, T., Dahlstrom, P., Trench, M.,
approaches to energy demand-side response: a systematic review. Renew. Sust.
2017. Artificial Intelligence: The Next Digital Frontier?.
Energ. Rev. 130, 109899 https://doi.org/10.1016/j.rser.2020.109899.

9
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

Casola, V., De Benedictis, A., Rak, M., Villano, U., 2019. Toward the automation of threat Ogunjimi, A., Rahman, M., Islam, N., Hasan, R., 2021. Smart mirror fashion technology
modeling and risk assessment in IoT systems. Internet Things 7, 100056. for the retail chain transformation. Technol. Forecast. Soc. Chang. 173, 121118
Chari, M.D., Luce, K., Thukral, I., 2017. Mastering the market intelligence challenge. MIT https://doi.org/10.1016/j.techfore.2021.121118.
Sloan Manag. Rev. Pantano, E., Pizzi, G., Scarpi, D., Dennis, C., 2020. Competing during a pandemic?
Chaudhry, I.A., Khan, A.A., 2016. A research survey: review of flexible job shop Retailers’ ups and downs during the COVID-19 outbreak. J. Bus. Res. 116, 209–213.
scheduling techniques. Int. Trans. Oper. Res. 23 (3), 551–591. https://doi.org/10.1016/j.jbusres.2020.05.036.
Chen, W., Wang, W., Wang, K., Li, Z., Li, H., Liu, S., 2020. Lane departure warning Pettey, C., 2018. How to get artificial intelligence “right”. Gartner. https://www.gartner.
systems and lane line detection methods based on image processing and semantic com/smarterwithgartner/how-to-get-artificial-intelligence-right.
segmentation: a review. J. Traffic Transp. Eng. (English Edition) 7 (6), 748–774. Phaal, R., Farrukh, C.J., Probert, D.R., 2004. Technology roadmapping—a planning
Chiranjeevi, T.N., Vishwanath, R.H., 2016. PRS: personnel recommendation system for framework for evolution and revolution. Technol. Forecast. Soc. Chang. 71 (1–2),
huge data analysis using porter stemmer. ICTACT J. Soft Comput. 6 (3). 5–26. https://doi.org/10.1016/s0040-1625(03)00072-6.
Cui, L., Huang, S., Wei, F., Tan, C., Duan, C., Zhou, M., 2017. Superagent: a customer Phaal, R., Farrukh, C., Probert, D., 2010. Roadmapping for Strategy and Innovation.
service virtual assistant for e-commerce websites. In: Proceedings of ACL 2017, Aligning Technology and Markets in a Dynamic World. University of Cambridge.
System Demonstrations, pp. 97–102. https://doi.org/10.18653/v1/P17-4017. Pillai, R., Sivathanu, B., Dwivedi, Y.K., 2020. Shopping intention at AI-powered
Díaz-Martín, A.M., Roche, I.C., Suárez, M.G., García, M.Q., Schmitz, A., 2019. automated retail stores (AIPARS). J. Retail. Consum. Serv. 57, 102207 https://doi.
Transformación digital en distribución: Soluciones tecnológicas y estrategias org/10.1016/j.jretconser.2020.102207.
competitivas de las empresas minoristas españolas. Rev. Ibér. Sist. Tecnol. Inf. E24, PricewaterhouseCoopers and Kantar, 2020. Retailing 2020: winning in a polarized
151–167. world. https://rasci.in/downloads/2012/Retailing_2020_Winning_Polarized_World.
Eisenhardt, K.M., 1989. Building theories from case study research. Acad. Manag. Rev. pdf.
14 (4), 532–550. Rai, A., Constantinides, P., Sarker, S., 2019. Next generation digital platforms: toward
Ertel, W., 2018. Introduction to Artificial Intelligence. Springer. human-AI hybrids. MIS Q. 43 (1), iii–ix.
Fernandes, T., Pedroso, R., 2017. The effect of self-checkout quality on customer Ramos, A.G., Daim, T., Gaats, L., Hutmacher, D.W., Hackenberger, D., 2022. Technology
satisfaction and repatronage in a retail context. Serv. Bus. 11, 69–92. roadmap for the development of a 3D cell culture workstation for a biomedical
Fernández-Caramés, T.M., Fraga-Lamas, P., 2018. A review on the use of blockchain for industry startup. Technol. Forecast. Soc. Chang. 174, 121213 https://doi.org/
the internet of things. Ieee Access 6, 32979–33001. 10.1016/j.techfore.2021.121213.
Gavrila, S.G., de Lucas Ancillo, A., 2021. Spanish SMEs’ digitalization enablers: E-receipt Rhee, H.S., Riggins, F.J., Kim, C., 2009. The impact of product type and perceived
applications to the offline retail market. Technol. Forecast. Soc. Chang. 162, 120381 characteristics of the web on multifaceted online shopping behavior. J. Organ.
https://doi.org/10.1016/j.techfore.2020.120381. Comput. Electron. Commer. 19 (1), 1–29.
Geiger, S., Gross, N., 2021. A tidal wave of inevitable data? Assetization in the consumer Shankar, V., Kalyanam, K., Setia, P., Golmohammadi, A., Tirunillai, S., Douglass, T.,
genomics testing industry. Bus. Soc. 60 (3), 614–649. Hennessey, J., Bull, J., Waddoups, R., 2021. How technology is changing retail.
Günther, W.A., Mehrizi, M.H.R., Huysman, M., Feldberg, F., 2017. Debating big data: a J. Retail. 97 (1), 13–27. https://doi.org/10.1016/j.jretai.2020.10.006.
literature review on realizing value from big data. J. Strateg. Inf. Syst. 26 (3), Silva, L.A.G.P.D., Mercês, N.N.A.D., 2018. Multiple case study applied in nursing
191–209. ISO 690. research: a case report. Rev. Bras. Enferm. 71, 1194–1197. https://doi.org/10.1590/
Han, M., Geum, Y., 2022. Roadmapping for data: concept and typology of data- 0034-7167-2017-0066.
integrated smart-service roadmaps. IEEE Trans. Eng. Manag. 69 (1), 142–154. Simoni, M., Sorrentino, A., Leone, D., Caporuscio, A., 2022. Boosting the pre-purchase
https://doi.org/10.1109/tem.2020.3013295. experience through virtual reality. Insights from the cruise industry. J. Hosp. Tour.
Hart, C., Doherty, N., Ellis-Chadwick, F., 2000. Retailer adoption of the internet – Technol. 13 (1), 140–156. https://doi.org/10.1108/JHTT-09-2020-0243.
implications for retail marketing. Eur. J. Mark. 34 (8), 954–974. https://doi.org/ Soni, N., Sharma, E.K., Singh, N., Kapoor, A., 2020. Artificial intelligence in business:
10.1108/03090560010331441. from research and innovation to market deployment. Proc. Comput. Sci. 167,
Hofmann, E., Rutschmann, E., 2018. Big data analytics and demand forecasting in supply 2200–2210. https://doi.org/10.1016/j.procs.2020.03.272.
chains: a conceptual analysis. Int. J. Logist. Manag. 29 (2), 739–766. Stark, E., Kučera, E., Haffner, O., Drahoš, P., Leskovský, R., 2020. Using augmented
Hoyer, W.D., Kroschke, M., Schmitt, B., Kraume, K., Shankar, V., 2020. Transforming the reality and internet of things for control and monitoring of mechatronic devices.
customer experience through new technologies. J. Interact. Mark. 51, 57–71. Electronics 9 (8), 1272.
https://doi.org/10.1016/j.intmar.2020.04.001. Tierney, R., Hermina, W., Walsh, S., 2013. The pharmaceutical technology landscape: a
Huang, Y., Porter, A.L., Cunningham, S.W., Robinson, D.K., Liu, J., Zhu, D., 2018. new form of technology roadmapping. Technol. Forecast. Soc. Chang. 80 (2),
A technology delivery system for characterizing the supply side of technology 194–211. https://doi.org/10.1016/j.techfore.2012.05.002.
emergence: illustrated for Big Data & Analytics. Technol. Forecast. Soc. Chang. 130, Tramèr, F., Zhang, F., Juels, A., Reiter, M.K., Ristenpart, T., 2016. Stealing machine
165–176. https://doi.org/10.1016/j.techfore.2017.09.012. learning models via prediction {APIs}. In: 25th USENIX Security Symposium
Islam, N., Marinakis, Y., Majadillas, M.A., Fink, M., Walsh, S.T., 2020. Here there be (USENIX Security 16), pp. 601–618.
dragons, a pre-roadmap construct for IoT service infrastructure. Technol. Forecast. Wadhwani, P., Gawande, S., 2023. Artificial Intelligence (AI) in Retail Market Size, By
Soc. Chang. 155, 119073 https://doi.org/10.1016/j.techfore.2017.09.016. Component (Solution [Chatbot, Customer Behavior Tracking, Customer Relationship
Karch, M., 2020, March 9. A Beginner’s Guide to Mobile Apps. Retrieved from Lifewire: Management (CRM), Inventory Management, Price Optimization, Recommendation
https://www.lifewire.com/what-are-apps-1616114. Engine, Supply Chain Management, Visual Search], Service [Professional Service,
Lee, K., 2019. AI-Superpowers: China, Silicon Valley und die neue Weltordnung [E-book]. Managed Service]), By Technology (Machine Learning, Natural Language Processing
Campus Verlag GmbH. (NLP), Computer Vision), By Application (Automated Merchandising, Programmatic
Lemon, K.N., Verhoef, P.C., 2016. Understanding customer experience throughout the Advertising, Market Forecasting, In-Store AI & Location Optimization, Data Science),
customer journey. J. Mark. 80 (6), 69–96. COVID-19 Impact Analysis, Growth Potential, Regional Outlook, Competitive Market
Leng, H.K., Wee, K.N.L., 2017. An examination of users and non-users of self-checkout Share & Forecast, 2023–2032. Global Market Insights Inc.
counters. Int. Rev. Retail Distrib. Consum. Res. 27 (1), 94–108. https://doi.org/ Wahba, P., 2021. A record 12,200 U.S. stores closed in 2020 as e-commerce, pandemic
10.1080/09593969.2016.1221842. changed retail forever. Fortune. https://fortune.com/2021/01/07/record-store-clo
Li, K., Kim, D.J., Lang, K.R., Kauffman, R.J., Naldi, M., 2020. How should we understand sings-bankruptcy-2020/.
the digital economy in Asia? Critical assessment and research agenda. Electron. Walsh, K., 2004. Interpreting the impact of culture on structure: the role of change
Commer. Res. Appl. 44, 101004. processes. J. Appl. Behav. Sci. 40 (3), 302–322.
Liu, K., Cheng, Y., Li, X., Gao, Y., 2019. Microwave-sensing technology using orbital Wei, Y., Tran, S., Xu, S., Kang, B., Springer, M., 2020. Deep learning for retail product
angular momentum: overview of its advantages. IEEE Veh. Technol. Mag. 14 (2), recognition: challenges and techniques. Comput. Intell. Neurosci. 2020 https://doi.
112–118. org/10.1155/2020/8875910.
Liu, Y., Guo, C.X., Zhou, H., Chen, X.Y., 2023. Pre-sale ordering strategy based on the Wu, Q., 2021. Research on pricing strategy of online and offline supply chain based on
new retail context considering bounded consumer rationality. J. Indust. Manag. channel preference in the context of new retail. Complexity 2021, 1–9. https://doi.
Optim. 19 (2), 1015–1043. org/10.1155/2021/5211642.
Lu, H.P., 2018. “New AI” era: species explosion. Tsinghua Business Review Z2. Xiao, M., Zhou, Q., Lu, L., Tao, X., He, W., Zhou, Y., 2022. Applying deep learning-based
Lu, H.P., Weng, C.I., 2018. Smart manufacturing technology, market maturity analysis personalized item recommendation for mobile service in retailor industry. Mob. Inf.
and technology roadmap in the computer and electronic product manufacturing Syst. 2022 https://doi.org/10.1155/2022/2364154.
industry. Technol. Forecast. Soc. Chang. 133, 85–94. https://doi.org/10.1016/j. Yang, J.Y., Lee, S.R., 2018. A study on introduction of IoT infrastructure based on BSC
techfore.2018.03.005. and AHP: focusing on electronic shelf label. J. Soc. e-Bus. Stud. 22 (3).
Madakam, S., Holmukhe, R.M., Kumar Jaiswal, D., 2019. The future digital work force: Yeh, C.H., Lin, H.H., Gau, Y.L., Wang, Y.S., 2020. What drives customers’ purchase
robotic process automation (RPA). J. Inf. Syst. Technol. Manag. 16, 1–17. https:// behavior in a click-and-mortar retailing context: a value transfer perspective.
doi.org/10.4301/s1807-1775201916001. J. Enterp. Inf. Manag. https://doi.org/10.1108/jeim-10-2019-0344.
McCarthy, J., Minsky, M.L., Rochester, N., Shannon, C.E., 2006. A proposal for the Yin, R.K., 2009. Case Study Research: Design and Methods, Vol. 5. Sage.
Dartmouth summer research project on artificial intelligence, August 31, 1955. AI Zhai, Q., Cao, X.J., Zhen, F., 2019. Relationship between online shopping and store
Mag. 27 (4), 12. shopping in the shopping process: empirical study for search goods and experience
Moore, S., Bulmer, S., Elms, J., 2022. The social significance of AI in retail on customer goods in Nanjing, China. Transport. Res. Record J. Transport. Res. Board 2673 (11),
experience and shopping practices. J. Retail. Consum. Serv. 64, 102755 https://doi. 38–47. https://doi.org/10.1177/0361198119851751.
org/10.1016/j.jretconser.2021.102755.
Nag, A., Menzies, B., Mukhopadhyay, S.C., 2018. Performance analysis of flexible printed
Hsi-Peng Lu is a Professor of Information Management at the National Taiwan University
sensors for robotic arm applications. Sensors Actuators A Phys. 276, 226–236.
of Science and Technology. He received his Ph.D. in Industrial Engineering from the
University of Wisconsin, Madison. His researches encompass e-commerce, digital

10
H.-P. Lu et al. Technological Forecasting & Social Change 195 (2023) 122778

transformation, managerial decisionmaking, and information systems management. His Jen-Chuen Tzou is a part-time Assistant Professor at the National Taiwan University of
articles have appeared in journals such as International Journal of Information Manage­ Science and Technology. She received her Ph.D. in Information Management from the
ment, Technological Forecasting and Social Change, Information & Management, Omega, National Taiwan University of Science and Technology. Her researches encompass e-
Journal of Business Research, Future Generation Computer Systems, Computers in Human commerce, digital marketing, and innovative technology management. Her articles have
Behavior, Internet Research, European Journal of Operational Research, and Journal of appeared in journals such as Behavior & Information Technology, Journal of Internet
Computer Information Systems. Technology, and several international academic conferences.

Hsiang-Ling Cheng is currently a Ph.D. candidate in Information Management at the Chiao-Shan Chen is an Assistant Professor in Health Care Management at National Taipei
National Taiwan University of Science and Technology. Her researches encompass e- University of Nursing and Health Sciences. She received her Ph.D. in Information Man­
commerce, online customer behavior, and innovative technology management. agement from the National Taiwan University of Science and Technology. Her researches
encompass the smart city, technology roadmap, service innovation design, decision-
making, and medical information.

11