1 s2.0 S026840122200069X Main

International Journal of Information Management 66 (2022) 102535
Contents lists available at ScienceDirect
International Journal of Information Management
journal homepage: www.elsevier.com/locate/ijinfomgt
How does artificial intelligence create business agility? Evidence

from chatbots
Xuequn Wang a, Xiaolin Lin b,*, Bin Shao c
a
School of Business and Law, Edith Cowan University, Joondalup, WA 6027, Australia
b
Department of Information Systems and Business Analytics, College of Business, California State University, Sacramento, CA, USA
c
Department of Computer Information and Decision Management, Paul and Virginia Engler College of Business, West Texas A&M University, 2501 4th Ave, Canyon, TX
79016, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Artificial intelligence Artificial intelligence (AI) is gaining increasing attention from business leaders today. As a primary AI tool,
Chatbot chatbots have seen increasing use by companies to support customer service. An understanding of how chatbots
Customer service are used is essential for improving customer service. Based on the relevant literature, this study examined the
Agility impacts of chatbot-enabled agility (namely, internal and external chatbot agility) on customer service perfor-
Routine use
mance and explored the antecedents from the perspective of information technology use (both routine and
Innovative use
innovative use). We collected data from 294 U.S. marketing employees from various industries, using a survey for
Dynamic capabilities
the assessment of our research model. The results showed that both routine and innovative use of chatbots were
positively related to internal and external agility. In particular, the innovative use of chatbots plays an important
role in creating business agility. Moreover, internal and external agility are positively related to customer service
performance. Through a close look at chatbots and their use, our study provides insight into the role of AI in
creating business agility. Practically speaking, this study suggests that both the routine and the innovative use of
chatbots should be encouraged to create agility and develop business sustainability.
1. Introduction
companies’ ability to take advantage of their resources (e.g., technol-
Artificial intelligence (AI) has provided great opportunities for ogy) to efficiently identify and address opportunities and threats
companies to address challenges raised in today’s rapidly changing (Mathiassen & Pries-Heje, 2006). Companies must enhance business
marketplace (Borges, Laurindo, Spínola, Gonçalves, & Mattos, 2020; agility to respond to market changes by deploying proper IT. Such IT-
Duan, Edwards, & Dwivedi, 2019; Dwivedi et al., 2021; Hu, Lu, Pan, enabled agility is critical to sustaining business success (Chuang, 2020;
Gong, & Yang, 2021). For example, chatbots supported by AI have Trischler, Johnson, & Kristensson, 2020). To incorporate chatbots into
received increasing responsibility to provide effective customer customer service, practitioners need to recognize the business agility
communication (Chung, Ko, Joung, & Kim, 2020; Coombs, 2020; He, created by chatbots (Akhtar, Khan, Tarba, & Jayawickrama, 2018).
Zhang, & Li, 2021). These chatbots can process a large amount of data Such chatbot-enabled business agility can help firms to better manage
and train themselves to interact with customers (Mikalef & Gupta, the changing environment and meet customer needs (Akhtar et al.,
2021). Recent reports predict that retail sales via chatbots will reach 2018; Chuang, 2020).
$112 billion by 2023 (Williams, 2019). Chatbots constantly improve The literature has suggested that IT use can enhance business
their conversational abilities (Watson, 2019), which can help agility. For example, Chuang (2020) reported that social-information
companies respond to customer demand and market changes (Chiu & processing capability and customer cocreation build social media
Chuang, 2021). agility, which then enhances the strength of customer–firm
As such, chatbots have gained increasing popularity in creating new relationships. With the development of AI, organizations have
business agility enabled by information technology (IT) (Chung et al., increasingly used chatbots to support employees’ work, which is likely
2020; McLean & Osei-Frimpong, 2019). Business agility refers to to create new business agility. However, to the best of our knowledge,
few studies have attempted to investigate how the use of chatbots creates
business agility and supports
* Corresponding author.
E-mail addresses: [email protected] (X. Wang), [email protected] (X. Lin), [email protected] (B. Shao).
https://doi.org/10.1016/j.ijinfomgt.2022.102535
Received 27 May 2021; Received in revised form 26 May 2022; Accepted 27 May 2022
Available online 13 June 2022
0268-4012/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
X. Wang et International Journal of Information Management 66 (2022)
customer service. It is meaningful to study business agility by exploring

the use of chatbots, given that IT use can evolve (Petter, DeLone, & 2. Literature review and theories
McLean, 2012). Therefore, our overall research question is: How do
chatbots create business agility and help support customer services? Specif- 2.1. Relevant studies on chatbots
ically, our study aims to do the following: (a) examine how chatbot use
results in chatbot-enabled business agility and (b) examine the effects The literature has increasingly recognized chatbots as an important
of chatbot-enabled business agility on customer service quality. To technological trend in supporting customer service (Table 1). Some
achieve our research aims, we draw upon the dynamic capability view studies have investigated chatbot adoption. For example, based on the
(DCV) (Teece, Peteraf, & Leih, 2016) to develop our research model. technology acceptance model and gratification theory, Rese, Ganster,
DCV is an appropriate theory explaining how the use of chatbots can and Baier (2020) reported that users’ acceptance of chatbots can be
help organizations develop dynamic capabilities, such as sensing determined by different factors, such as the authenticity of conversation,
and seizing opportunities, which ultimately enhance customer service perceived usefulness, and perceived enjoyment. Conversely, although
performance. We conducted a survey and tested our model with 294 some prior studies have examined the adoption of chatbots in various
U.S. marketing employees. industries such as telecommunications and clothing (Etemad-Sajadi &
Our study makes two important contributions. First, our study shows Ghachem, 2015; Roy & Naidoo, 2021), few studies have examined the
how different types of chatbot use help achieve business agility. We actual use of chatbots. Among these few studies, McLean and Osei-
argue that chatbot use within organizations can be divided into routine Frimpong (2019) reported that factors such as website aesthetics and
and innovative chatbot use (Li, Hsieh, & Rai, 2013; Roberts, Campbell, perceived customization can increase chatbot use. More recently,
& Vijayasarathy, 2016). Routine and innovative use represent distinct Shumanov and Johnson (2021) showed that consumers can use chatbots
post-acceptance information systems (IS) usage behaviors (Li et al., for longer when their personality is congruent with a chatbot’s per-
2013) and were therefore selected for our study. Specifically, relying sonality. Both studies examined the use of chatbots from the consumer
only on routine use often leads to the underutilization of IT, whereas perspective and in the context of mobile services.
engaging in innovative use allows employees to explore IT’s new ap- As chatbots have become an innovative channel by which to
plications (Tams, Thatcher, & Craig, 2018). Therefore, fully compre- interact with customers (e.g., Chung et al., 2020; Claessen, Schmidt, &
hending the role of routine use versus innovative use is crucial to Heck, 2017; Etemad-Sajadi & Ghachem, 2015; Jain et al., 2018;
formulating methods of realizing the full value of chatbots. By exam- Shumanov & Johnson, 2021), some studies have started to examine
ining their routine and innovative use as separate variables, our study the impacts of chatbot use on business outcomes with a focus on
can clarify how chatbots enable agility and which type of chatbot use is customer service through identifying chatbots’ features and
more important. Because different types of IT can be used in a variety functionalities (e.g., Chung et al., 2020; Rese et al., 2020). For
of example, Schuetzler et al. (2020) suggested that the anthropomorphic
ways, the effects of routine and innovative use are probably different in character of chatbots, such as perceived humanness and partner
other contexts.1 Our study thus contributes to the literature by engagement, improves users’ experiences. Ashfaq et al. (2020)
capturing the uniqueness of chatbots and clarifying how chatbot- reported that the quality of information and services provided to
enabled agility can be achieved through routine and innovative use. customers could improve their satisfaction. McLean and Osei-Frimpong
Second, our study provides novel insights by demonstrating the (2019) suggested that chatbots can help companies provide high-
importance of chatbot-enabled agility for enhancing customer service. quality services to their customers and achieve various benefits, such
The literature has suggested that agility adjusts internal operational as word-of-mouth referrals and customer satisfaction. More recently,
changes and exploits external business environments (e.g., Akhtar et Adam et al. (2020) showed that anthropomorphism (identity, small
al., 2018; Lokshin, Belderbos, & Carree, 2008), making it essential to talk, and empathy) and fulfilling small requests increase the likelihood
include both internal and external aspects when evaluating the role of of consumer compliance.
agility in enhancing customer services. Following the literature To summarize, companies have been deploying chatbots to cus-
(Chuang, 2020), our study conceptualizes agility as internal and tomers with the goal of improving customer service and customer
external. Our results show that while both internal and external agility satisfaction. The key to achieving this goal stems from the capacities
can significantly improve customer service, external agility plays a and agility provided by chatbot integration into regular business
more important role. Our study thus contributes to the literature by operations. However, our review (Table 1) showed that few studies
describing how chatbot-enabled agility can enhance customer services. have examined chatbots from an employee perspective and assessed
Our paper could also make significant original contribution to the AI how employees’ chatbot use can result in business agility and enhanced
literature, given that it is among the very first studies to attempt to customer service. To fill this research gap, our study has focused on how
conceptualize agility and investigate its impacts within the context of chatbot use generates business agility, which in turn enhances
emerging AI tools. AI-enabled business agility will be the next frontier, customer service quality. Therefore, employees in the field of
and the results of this study provide helpful guidelines for companies to marketing were selected as the focus of this study. Next, we describe
achieve new business agility and improve customer service via agility and ways of helping transition the use of chatbots into enhanced
investment in the development and use of chatbots. customer service performance.
2.2. Business agility: a chatbot perspective
The term agility refers to the ability of a company or an organization

to rapidly react to changes and uncertainties (Akhtar et al., 2018;
Chuang, 2020). Business agility can be defined as a company’s ability to
use its human resources, technology, processes, and knowledge to effi-
ciently identify and address opportunities and threats, such as market
1
For example, customer service can use chatbots to answer customers’ changes, customer demands, and new technologies (Mathiassen &
standard questions (routine use) or to analyze customers’ data and generate Pries-Heje, 2006). Teece et al. (2016) have defined agility as “the ca-
additional insights (innovative use). On the other hand, human resources can
pacity of an organization to efficiently and effectively redeploy/redirect
use chatbots to conduct initial screening (routine use). Chatbots can also be
its resources to value creating and value protecting (and capturing)
used to conduct satisfaction surveys, after which the collected data can be
higher-yield activities as internal and external circumstances warrant”
analyzed with novel approaches to better understand employee performance
(innovative use). (p. 17). Nold, Anzengrubcr, Woclfle, and Michel (2018) argued that
2
organizational agility can be conceptualized in four dimensions:
culture,
3
Table 1
Summary of the chatbot literature.
Study Method Context Independent variable Dependent Variables Summary
variable measuring
chatbot use
Adam, Wessel, EXperiment Online banking Anthropomorphic design cues; User compliance Anthropomorphism (identity,
and Benlian
(2020) foot-in-the-door technique small talk, and empathy) and
the need to stay consistent
Ashfaq, Yun, Yu, increase compliance.
and Loureiro Survey Information quality; service Satisfaction; The need for interaction
(2020) quality; perceived enjoyment; continuance moderates the relationships
perceived usefulness; perceived intention between perceived ease of use
ease of use and perceived usefulness and
satisfaction.
Chung et al. Survey E-service agents for luXury Interaction; entertainment; Satisfaction Service agents’ marketing
(2020) brands trendiness; customization, efforts can increase customers’
problem solving satisfaction via accuracy and
credibility.
Etemad-Sajadi Survey Airline carrier; travel Utilitarian and hedonic value E-service quality Utilitarian values influence
and Ghachem agency; nine dimensions of e-service
(2015) telecommunications; rail quality, whereas hedonic
transport; furniture values influence five
retailing; health insurance dimensions of e-service
quality.
Jain, Kumar, Interview; Eight chatbots were
Kota, and Patel second-hand evaluated.
(2018) data
Lalicic and Survey values; reasons for/against value co- Chatbot use intention can be
Weismayer chatbot creation; explained by four different
(2021) behavioral combinations.
intention
McLean and Survey Mobile services Website aesthetics; perceived Use of chatbot Three items Eight variables (e.g., website
Osei-Frimpong customization; perceived ease of measuring the aesthetics, perceived
(2019) use; perceived usefulness; use of chatbot customization) motivate
perceived information quality; chatbot use.
perceived web credibility;
perceived timeliness;
dissatisfaction with experience
Pizzi, Scarpi, and EXperiment Mobile services; car rental Assistant type/initiation Choice Anthropomorphic chatbots
Pantano (2021) can reduce reactance, but also
reduce satisfaction.
Rese et al. (2020) Survey Clothing company Behavioral A technology acceptance
intention model was compared with the
use and gratification theory to
predict chatbot acceptance.
Roy and Naidoo EXperiment Hotel; smartphone; clothing Conversation style Attitude toward Consumers can have positive
(2021) company brand; purchase attitudes and higher purchase
intention intentions when their time
orientations match their
conversation type.
Schuetzler, EXperiment Conversational skills Social presence; Chatbots with high
Grimes, and perceived conversational skills can lead
Scott Giboney humanness; to higher social presence and
(2020) partner anthropomorphism.
engagement
Shumanov and EXperiment Mobile services Consumer-chatbot personality Chatbot Chatbot Congruence between
Johnson (2021) engagement; engagement consumer and chatbot
purchasing (the average personality increases
behavior interaction consumer engagement and
duration) purchasing behaviors.
leadership, systems, and people. of IT is one way firms can sustainably

Innovation represents the core of the business agility framework
and is often influenced by IT. Flexible IT infrastructure significantly
benefits agility (Tallon & Pinsonneault, 2011). Gao, Zhang, Gong, and
Li (2020) described two research streams that examined the role of IT
in creating organizational agility. The first stream deals with technical
IT capabilities and focuses on IT flexibility and IT integration. The
second stream investigates managerial IT capabilities, including
factors such as IT skills, proactive stance, and business spanning
capability. In the first stream, Ngai, Chau, and Chan (2011) showed
that both IT integration and flexibility can enhance the agility of a
supply chain. In the second stream, Overby, Bharadwaj, and
Sambamurthy (2006) explored the specific ways in which IT and
digital alternatives impact company agility and found that effective use
4
engage in sensing business environmental changes and in formulating
responses. The use of advanced IT tools can enhance business agility
because they provide companies and organizations with new ways of
increasing agility (e.g., Akhtar et al., 2018; Lu & Ramamurthy,
2011).
The DCV is a useful theoretical lens through which to understand
the business agility created by IT, which can emerge in the form of
dynamic capabilities (e.g.,Gupta & Meissonier, 2020). It suggests that
“firms whose managers have superior dynamic managerial
capabilities can adapt and change more successfully than firms whose
managers have less effective or no dynamic managerial capabilities”
(Helfat & Martin, 2015, p. 1304). More specifically, dynamic
capabilities include firms’ abilities to identify and seize opportunities,
and they comprise three main clusters: identifying and developing
technological opportunities in response to customer needs (i.e.,
sensing), mobilizing resources to deal
5
with these needs (i.e., seizing), and continual renewal (i.e., trans- 2020). Chatbots can also help achieve business agility through
forming) (Teece et al., 2016). Prior studies have demonstrated that IT external sourcing. Chatbots
can play a bridging role in developing dynamic capabilities by con-
necting organizational resources and operational business functions (e.
g., Duan, Faker, Fesak, & Stuart, 2013; Gupta, Drave et al., 2020; Zhang,
Qu, Ho, & Huang, 2011).
Therefore, sensing, seizing, and transforming are essential to
achieving agility (Teece et al., 2016), which is created by IT from a DCV
perspective. Sensing involves proactively developing new solutions
based upon customer needs and evaluating the effectiveness of these
solutions (e.g., scenario planning). Seizing deals with selecting appro-
priate solutions and implementing these solutions (e.g., developing
“slack,” re-engineering structures). Transforming includes iterative
improvement through trialing, launching, and learning. When firms
have dynamic capabilities that have emerged from IT, they can modify
organizational resources to respond to customer needs and market
changes, thereby achieving agility (Teece et al., 2016).
In addition, based on the DCV, the literature has argued that dynamic
capabilities can have internal and external sourcing and that both in-
ternal and external capabilities can strengthen relationships with cus-
tomers (e.g., Akhtar et al., 2018; Lokshin et al., 2008). In a recent
study, Chuang (2020) categorized social media agility as either
internal or external agility. Internal agility refers to firms’ flexibility in
dealing with customers’ demands in their day-to-day operations (i.e.,
internal sourcing). EXternal agility refers to firms’ flexibility in
responding to customers’ requirements stemming from external
information sources, links, and collaborative relationships (i.e.,
external sourcing).
In the digital age, AI can increase companies’ flexibility and ability to
respond to changes in the competitive business environment by under-
pinning such important business functions, such as automating business
processes, gaining insight through data analysis, and engaging with
customers and employees (Davenport & Ronanki, 2018). Chatbots
have been increasingly used by companies to support customer service.
They can effectively connect companies with their customers and
improve customer service through various technical features, such as
seamless live communication and 24/7 customer service (Ameen,
Tarhini, Reppel, & Anand, 2021; Huang & Rust, 2018; Hui, Fong, &
Jha, 2001). As such, chatbots enable companies to answer customers’
questions and address their problems more efficiently. Companies
using chatbots can also understand their customers better, identify
customer demands, and identify changes in targeted marketing based
on interaction with their customers. As such, they can better fulfill
customer needs, react to changes more efficiently and effectively, and
increase business agility. Therefore, it is clear that the merging of
chatbots and business operations has created a new way for
companies to achieve the agility to respond to both internal
customer demand and external marketing changes, which we name
chatbot-enabled agility.
Following Chuang (2020), we developed the concept of chatbot-
enabled agility as one that has two dimensions: internal and external.
Internal chatbot agility refers to the flexibility of companies enabled by
chatbots to respond to customer demand by optimizing the delivery
and offering of products and services and by resolving customers’
questions efficiently and effectively. Specifically, chatbots can help
achieve business agility through internal sourcing. Chatbots can
analyze customer data via AI algorithms (e.g., machine learning) and
understand how to address customers’ questions (i.e., sensing). These
solutions can then be implemented in chatbots to better serve
customers (i.e., seizing). Last, such a process is interactive, and
chatbots can improve their performance through ongoing learning
from their interactions with customers (i.e., transforming). Thus,
chatbots can help businesses achieve sensing, seizing, and transformation
through internal sourcing and enabling of agility (Teece et al., 2016).
EXternal chatbot agility refers to the flexibility of companies
enabled by chatbots to respond to market changes efficiently and
effectively and to identify new marketing opportunities (Chuang,
6
can analyze market trends and changes from external sources via AI information
algorithms (e.g., machine learning) and develop their estimations of
different markets (i.e., sensing). These estimations can then be inte-
grated into chatbots to better understand customers and provide
improved services (i.e., seizing). Last, given that the market keeps
changing, such a process is interactive, and chatbots can help firms
continuously adapt to the changing market (i.e., transforming). Thus,
chatbots can also help achieve sensing, seizing, and transforming via
external sourcing and achievement of agility (Teece et al., 2016).
Applying this distinction thus follows the literature, which posits
that agility can be captured as either internal or external (Da Silveira,
Borenstein, & Fogliatto, 2001; Huo, Gu, & Wang, 2018). One can
define agility using two orientations: companies’ ability to quickly
address customer needs or to address market change. Internal agility
addresses customers’ needs via internal business processes, focusing
on issues related to a company’s products and services as well as the
customers’ requirements. EXternal agility identifies and responds to
dynamics in the external business environment (e.g., the market).
2.3. Routine versus innovative use
After chatbots have been implemented into organizations, they

need to be used by marketing employees to provide customer service
and to generate business value. Previous studies have examined
consumers’ general use of chatbots (McLean & Osei-Frimpong, 2019;
Shumanov & Johnson, 2021), but few studies have taken the further
step of examining different types of chatbot use within
organizations. In the IS literature, the process of technology
implementation has been divided into siX stages (Cooper & Zmud,
1990): initiation, adoption, adaptation, acceptance, routinization, and
infusion. Routinization refers to the integration of technology into the
normal work process, and infusion refers to the deep and
comprehensive embedding of technology into the work process. After
employees accept the technology and commit to its use, routinization
and infusion can occur during the post-acceptance stage. Further, the
literature has argued that routinization and infusion do not need to
happen sequentially and can coexist in parallel (Cooper & Zmud,
1990).
In the context of business intelligence systems, Li et al. (2013) pro-
posed the concepts of routine and innovative use to reflect the routini-
zation and infusion stages, respectively. Routine use is defined as
“employees’ using IS in a routine and standardized manner to
support their work” (Li et al., 2013, p. 659). Innovative use refers to
“employees’ discovering new ways to use IS to support their work” (Li
et al., 2013, p. 659). Routine use has a standardization orientation
(Benner & Tushman, 2002). Under routine use, employees use
technology as a normal part of their work activities (Li et al., 2013).
Thus, routine use allows employees to become familiar with a certain
technology and to improve their work efficiency.
Conversely, innovative use has an innovation orientation and chal-
lenges the repetitive work process (Benner & Tushman, 2002).
Innova- tive use entails developing creative alternatives that further
realize the value of the technology (Jasperson, Carter, & Zmud, 2005).
Based on the work of Li et al. (2013), Roberts et al. (2016) identified
routine and innovative use as two types of managerial use of decision
support systems. They argued that the primary difference between
the two kinds of use lies in the different ways managers use decision
support systems to support their work. Managers may use these
systems to accomplish assigned work in a standardized manner, such
as learning about customers and sales. Managers may also use
decision support systems to accomplish assigned work in an
innovative manner, such as applying new data analysis skills and
identifying new business opportunities.
Therefore, in our study, we argue that both routine and innovative
use appear when using chatbots to support business operations and
employees’ work. Marketing employees can engage in routine use.
For example, employees can use chatbots to answer customers’
routine questions quickly and automatically to provide product
7
with web pages and tutorials, and to collect feedback from customers.
Chatbots can also be trained to better answer routine questions. Mar- product and service delivery (Pizzi et al., 2021). Without a doubt,
keting employees can participate in innovative use as well. For chatbot-supported internal agility will allow companies to
example, they can analyze conversational data with new methods to communicate with customers clearly and methodically and to decrease
help identify additional marketing opportunities. Employees can also the difficulty of maintaining relationships with customers. Therefore,
ask new questions when interacting with customers. The conversational we hypothesized the following:
data can then be used to train chatbots so they can deal with new types
H1. : Internal chatbot agility is positively related to customer service
of questions or answer questions in new ways, based on specific
performance.
contexts.
Firms can engage in both routine and innovative use of chatbots. Moreover, we argue that increased external agility, enabled by
They can integrate chatbots into routine business operations while also chatbots, can also improve customer service performance. The market
discovering new approaches to obtain additional benefits from is dynamic—especially in the digital era—and companies must present
chatbots. Whether a certain way to use chatbots is innovative depends correct and timely responses to market changes to remain competitive
on the specific context. Specifically, when a certain usage approach to (Weber & Tarba, 2014). EXpected market changes include but are not
chatbots has not been routinized at a certain point in time, it is limited to factors such as consumer demand, market capacity, industry
considered innovative use and is consistent with operationalizations of inventory, competitor pricing, product quality, financing, interest
routine and innovative use in the existing literature (Li et al., 2013). It rates, asset prices, the employment rate, and national and regional
is possible that certain approaches to innovative use may become political stability. Through interactions with customers, chatbots can
routinized later and become routine use. collect a large volume of data. Data analysis can improve
understanding of customer profiles and needs and thus improve
3. Research model and hypotheses development understanding of targeted marketing (Pizzi et al., 2021). Specifically,
chatbots rely on AI algorithms such as machine learning, which can
Our model appears in Fig. 1. Based on the literature, we argue that be trained to monitor market changes (Phansalkar, Kamat, Ahirrao, &
marketing employees can engage in both routine and innovative use of Pawar, 2019). As such, chatbots can help companies monitor market
chatbots. These two types of use can enhance both internal and dynamics and respond to their changes (Chuang, 2020). We also
external chatbot agility, which ultimately will enhance customer anticipate the appearance of new business models, taxes, and
service performance. In the following text, we will describe each regulations—as well as a change in inflation rates. If a company is
hypothesis in more detail. involved in the international market, external agility will also help it to
withstand the fluctuation of exchange rates. Therefore, we proposed
3.1. The effect of agility on customer service performance the following hypothesis:
H2. : EXternal chatbot agility is positively related to customer service
Previous studies have shown a significant positive correlation be- performance.
tween business agility and performance (Queiroz, Tallon, Sharma, &
Coltman, 2018; Tallon & Pinsonneault, 2011). Based on the DCV,
recent studies have demonstrated that dynamic capabilities, which 3.2. The effect of routine use on agility
agility re- lies on, can improve various business outcomes, such as
economic performance, to sustain the business (e.g., Gupta & It is understood that a particular type of IT use is likely to be asso-
Meissonier, 2020; Helfat & Martin, 2015). Accordingly, we argue that ciated with individuals’ perceptions of outcomes resulting from the use
increased internal agility enabled by chatbots in organizations can (e.g., Bae, 2017). For example, Chuang (2020) showed that social-
boost customer service performance. AI-enabled chatbots can use information process capability of social media can positively affect
various algorithms (e.g., natural language processing) and understand users’ perceptions of social media agility. Social-information process
how to interact with customers (Mikalef & Gupta, 2021). Therefore, capability can be considered a specific type of social media use, which
this growth in internal agility, enabled by chatbots, allows a business’s increases the agility created by social media. Business agility has been
daily operations to adjust to handle revisions in customer demands referred to as a business’s ability “to sense and respond rapidly to
(Chuang, 2020). For example, chatbots can capture customers’ feedback unpredictable events in order to satisfy changing customer demands”
more quickly, in turn allowing the research and development (Holmqvist & Pessi, 2006, p. 146). It thus deals with firms’ capabilities
department to redesign products to meet updated requirements more to adapt to a changing environment driven by factors such as changing
quickly. customer preferences and target markets (Overby et al., 2006). As pre-
With the help of chatbots, the marketing department will be able to viously discussed, internal agility can capture the capabilities of com-
determine the best times to offer promotions. Furthermore, through panies to deal with customers by optimizing internal business processes.
analysis of data from chatbots, companies can streamline the sensing In our study, we argue that routine use of chatbots can increase
and reflection of customer needs to enhance product and customer internal agility. As previously discussed, routine use refers to the
service quality (Pizzi et al., 2021). The aforementioned rise in adapt- integration of chatbots into a firm’s standardized marketing process
ability to customer demands will also facilitate a more efficient system of (Li et al., 2013). Based on the DCV, routines are important to
support dynamic
8
Fig. 1. Research model.
9
capabilities (Teece et al., 2016) because they can efficiently integrate

the functions of chatbots into organizational functions in real-time. marketing opportunities can be identified, and firms can continuously
Following the DCV (Helfat & Martin, 2015), we argue that routine use provide new offerings or promotions to meet customers’ additional
can increase companies’ dynamic capabilities, such as identifying and needs (i.e., transforming), a process that also enhances internal agility
seizing, as well as reconfiguring, organizational assets (Roberts et al., (Teece et al., 2016). Therefore, we hypothesized that:
2016). Specifically, as marketing employees become more familiar with H4a. Innovative use of chatbots is positively related to internal chatbot
chatbots and incorporate them into their work, the marketing process agility.
can be better supported. With the incorporation of chatbots into the
routine work process, they can enhance information processing within The innovative use of chatbots can also increase external agility.
organizations and help collect a large amount of customer data that Chatbots can be trained with new algorithms and can thus respond to
can be used to generate marketing insights (Akhtar et al., 2018). The market changes and address new requirements from customers (Chuang,
chatbots’ unique ability to handle high-volume data with improved 2020). Thus, the innovative use of AI can increase firms’ dynamic ca-
acces- sibility and analysis facilitates the development of dynamic pabilities related to sensing (Teece et al., 2016). Employees can also
capability, which leads to agility. Therefore, routine use is more likely ask different questions during their interactions with customers and
to provide offerings and promotions to meet customers’ needs by then train chatbots to deal with new issues from customers. Therefore,
enhancing effi- ciencies in the marketing process, thus increasing firms’ dynamic seizing capabilities can also be enhanced. Last, given
internal agility. Therefore, we hypothesized the following: that such processes are iterative and chatbots can keep monitoring
continuous changes to the market, firms’ dynamic transforming
H3a. Routine use of chatbots is positively related to internal chatbot agility.
capabilities can also be developed (Teece et al., 2016). In those contexts,
EXternal agility can capture the capabilities of companies to innovative use of chatbots can help firms and employees better identify
identify and respond to new opportunities in the external environment and respond to market changes and new opportunities and explore
(Chuang, 2020). The use of chatbots can provide companies with new growth paths, thus improving external agility. (Roberts et al.,
functions to actualize these capabilities. For example, social media 2016). In addition, according to the DCV (Helfat & Martin, 2015),
functions such as information processing and customer cocreation can employees have bounded rationality, which limits their capacity to
create social media agility (Chuang, 2020). In our study, we also argue identify new marketing opportunities. Innovative use of chatbots can
that routine use of chatbots can increase external agility. AI algorithms help employees overcome limitations and expand their capabilities,
can help chatbots keep track of customers’ needs and concerns thus allowing them to explore new opportunities. The literature has
(Phansalkar et al., 2019). For example, by conducting routine analysis also shown that innovative use can increase managers’ abilities to
with conversational data between chatbots and customers, employees sense opportunities for organizational innovation (Roberts et al.,
can better understand whether customer preferences have changed, 2016), which can also contribute to marketing expansion. Therefore,
enhancing sensing capabilities. By integrating chatbots into standard we hypothesized the following:
marketing procedures, employees can also examine whether customers H4b. Innovative use of chatbots is positively related to external chatbot
request certain types of products or focus more on certain features of agility.
products in response to targeted marketing, increasing seizing
capabilities. Altogether, routine use of chatbots provides insights that
enhance an organization’s dynamic capabilities, such as sensing and 3.4. The mediating effects of internal and external agility
seizing (Teece et al., 2016), to assist in reconfiguring resources in
response to market changes. As a result, external agility also increases. In the previous discussions, we have hypothesized that routine use
In other words, employees who use chatbots as part of their routine work of chatbots has a positive effect on internal and external agility, in turn
processes can acquire knowledge about the market and thus seize increasing customer service performance. Based upon DCV, we further
marketing opportunities (Roberts et al., 2016). To summarize, routine argue that internal and external agility mediate the effect of routine use
use allows employees to be more aware of customer demand and of chatbots on customer service performance (Teece et al., 2016). Spe-
market changes. Therefore, we hypothesized the following: cifically, routine use of chatbots can support organizations’ standardized
marketing processes, and enhancing these processes can generate dy-
H3b. Routine use of chatbots is positively related to external chatbot agility.
namic capabilities (Teece et al., 2016) dealing with sensing, seizing,
and transforming organizational assets (Roberts et al., 2016).
3.3. The effect of innovative use on agility Therefore, incorporating chatbots into routine workplaces should
improve companies’ understanding of customer needs and their
The innovative use of chatbots can also increase internal agility. tracking of marketing changes by processing massive amounts of
Innovative use refers to new ways of using chatbots to support customer data (Akhtar et al., 2018), generating both internal and
customer service (Li et al., 2013, p. 659). Innovative use helps external agility. Based on DVC, this enhanced business agility allows
organizations achieve the full benefits of IT (Tams et al., 2018). With companies to provide better customer services (e.g., Gupta &
innovative use, employees can also benefit from the features of IT Meissonier, 2020; Helfat & Martin, 2015). Therefore, we hypothesize:
more extensively and expend more effort applying IT in innovative
H5. : Internal (H5a) and external chatbot agility (H5b) mediate the
approaches (Nambisan, Agarwal, & Tanniru, 1999). The DCV argues effect of routine use of chatbots on customer service performance.
that dynamic capabilities are not solely based on routines, and
innovative thinking and actions are also vital for developing dynamic Lastly, the previous hypotheses state that innovative use of chatbots
capabilities (Teece et al., 2016). Consistent with the DCV (Helfat & positively impacts internal and external agility, which then enhance
Martin, 2015), innovative use can also increase organizations’ ability customer service performance. Based on DCV, we further argue that
to sense opportunities and then reconfigure resources accordingly internal and external agility mediate the effect of innovative use of
(Roberts et al., 2016). For example, marketing employees may let chatbots on customer service performance (Teece et al., 2016). In this
chatbots ask customers new questions and then try new approaches to mediated relationship, the innovative use of chatbots is important to
serving them. Thus, marketing employees can challenge existing generate dynamic capabilities in addition to routine use (Teece et al.,
processes and develop creative approaches to responding to 2016). Specifically, the innovative use of chatbots also strengthens
customers’ needs while also enhancing internal agility. Next, new companies’ capabilities of sensing and seizing opportunities (Roberts
methods can be used to analyze conversational data resulting from et al., 2016). For instance, marketing employees may add new
interactions between chatbots and customers. As a result, new questions into chatbots and later use new methods to analyze
1
customer data chatbots collect. These new approaches of using
chatbots can help
1
companies further understand customer preferences and market

changes, enabling them to offer better customer services (e.g., Gupta & Table 2
Meissonier, 2020; Helfat & Martin, 2015). We therefore hypothesize: Demographic information of subjects.
Category Sample (N = 294)
H6. : Internal (H6a) and external chatbot agility (H6b) mediate the
effect of innovative use of chatbots on customer service performance. Gender Female: 37.8%
Age
18–24 7.5%
4. Research method 25–34 38.4%
35–44 37.4%
4.1. Data collection and sample 45 or above 16.7%
Company size in number of employees
1–200 22.5%
We hired a professional research firm to recruit employees who 201–500 16.4%
work for companies that use chatbots, and the firm conducted a survey 501–1000 20.4%
to collect data about employees’ perceptions of the use of chatbots and 1001–3000 17.9%
its consequences. The survey contains items that measure our > 3000 22.9%
Education
constructs as well as open-ended questions to help us further understand
High school or below 9.5%
the impact of chatbot use. Therefore, our study is mainly quantitative, Some college education or bachelor’s degree 61.2%
intended to test the model as developed, complemented by answers Graduate degree 29.3%
from qualitative questions. The research firm maintains national panels, Industry
Health care 9.9%
and recruits and screens participants in a variety of industries. Our
Manufacturing 13.3%
study focuses on employees who are familiar with the use of chatbots Education 4.8%
within organizations and who regularly interact with customers. Higher education 2.0%
Thus, our subjects are Banking/Finance 15.0%
U.S. employees who work in marketing departments or other similar Insurance 3.7%
Wholesale and distribution 4.1%
departments. Data collection occurred over two weeks around the
Transportation 5.1%
beginning of 2021. The link to the survey was sent to potential partici- Government 2.4%
pants via email. The survey contains quality control questions (e.g., Retail 11.9%
“Please answer this question by selecting ‘strongly disagree’”) to detect Hospitality 1.4%
careless answers. The sampling approach was systematic sampling. For IT and telecommunication 6.5%
Other 20.0%
example, if we planned to collect 200 participants and the national
panels had about 100,000 individuals, then the survey invitations were
sent out with the interval of 500 individuals. We included several and are consistent with the item “My company has established good
screening questions, such as “Does your company use chatbots to sup- relationships with customers” from Chuang (2020). Therefore, this item
port customer services?” and “What are your job responsibilities?” was selected in our study to reflect enhanced customer-firm relation-
Participants would not qualify for our study if they selected “no” in ships as a measure of customer service performance.2 We also used the
response to the first question or did not select “marketing (using chat- item “My company has provided better customer service than ever” to
bots to interact with customer)” in response to the second question. In represent overall customer service performance. This item is consistent
the survey, we also asked participants for the name of the chatbot they with “achieving overall satisfaction” from Setia et al. (2013).
used. In total, we received 323 responses. After removing careless an- We adapted items measuring routine and innovative use of
swers in response to the quality control questions, a total of 294 valid chatbots from Li et al. (2013) and modified them to fit our research
responses remained (valid rate: 91.02%). We tried to eliminate the context for chatbots. These items were selected because they reflected
nonresponse bias through several approaches (Rosenthal & Rosnow, employees’ actual usage behaviors. We used items that Akhtar et al.
2008): (a) offering incentives (bonus points provided by the survey (2018) and Chuang (2020) designed to measure internal and external
company, which could be converted to money) to encourage partici- agility, with modifications to fit the current research context. These
pation; (b) sending reminders to enhance participation; and (c) ensuring items were selected because they focused on the measurement of
that participation was anonymous and no personal identification in- agility enabled by IT (e.g., social media). Adding “Because of the use of
formation would be collected. We also compared gender, age, and ed- chatbots” to the questions ensured that participants focused on agility
ucation between early and late participants, and there were no enabled by chatbots, also consistent with Chuang (2020). In other
significant differences. Therefore, a nonresponse bias should not be an words, we were able to measure chatbot-enabled agility, rather than
issue in our study. On average, the respondents had worked for their business agility in general. As such, it can provide further insights
current companies for 7.86 years. The companies involved used chatbots into the measures of business agility that is created by a particular
such as SnatchBot and Bold360. Participants came from a variety of type of IT. Removing this text would lead the items to measure general
industries, such as bank/finance, health care, IT and telecommunica- internal and external agility, which can be influenced by
tion, retail, and hospitality, consistent with a recent report on the organizations’ differing technologies.
chatbot market (Mordor Intelligence, 2021). Demographic information
appears in Table 2.
4.3. Data analysis and results
4.2. Measures We used SmartPLS 3.2.8, which offers the bootstrap resampling
method (using 5000 samples), to test our model. Our measures were not
Our measures were adapted from prior literature (see Appendi X A). normally distributed (i.e., Shapiro–Wilk tests were significant).
This study used 7-point Likert scales, with anchor 1 indicating Following Hair, Hult, Ringle, and Sarstedt (2016), it is appropriate to
“strongly disagree” and anchor 7 indicating “strongly agree.” We analyze unusually distributed data with PLS. Further, PLS focuses on
measured customer service performance using items modified from predicting and maximizing explained variance (Henseler, Ringle, &
Chuang (2020) and Setia, Setia, Venkatesh, and Joglekar (2013). In Sarstedt, 2015), consistent with our aim of predicting customer service
particular, Setia
et al. (2013) used items such as “retaining existing customers” and
“attracting new customers” to measure customer service performance. These items reflect a firm’s strengthened relationships with customers
1
2
We thank one reviewer for providing very helpful comment regarding the
measures of customer service performance.
1
performance. Last, PLS requires a smaller sample size (Henseler et al.,

2015) and would therefore allow us to conduct post-hoc analysis with Table 4
subsamples (Henseler et al., 2015). Therefore, we chose SmartPLS for Correlation between constructs and square root of AVEs (on Diagonal).
this study. 1 2 3 4 5
1 Routine use of chatbots 0.90

4.3.1. Measurement model testing 2 Innovative use of chatbots 0.65 0.89
We assessed the measurement model using both convergent and 3 Internal agility 0.58 0.68 0.88
discriminant validity criteria for the variables. First, the loadings of 4 EXternal agility 0.57 0.68 0.81 0.88
5 Customer service performance 0.59 0.64 0.72 0.81 0.94
each item placed above.70. Items also had acceptable reliabilities and
the average variance extracted (AVE) for each construct exceeded.50
(Table 3), both supporting convergent validity. Second, as shown in use of chatbots is also positively related to external agility, as stated in
Table 4, the square root of each variable’s AVE exceeded correlations H4b, which is also supported (β =0.53, p < .001). The two types of
between that variable and all other variables, which supports discrimi- chatbot use explain 49.37% of the variance based on internal agility
nant validity. In addition, all the correlations between constructs fell and 48.79% of the variance based on external agility. The two types of
below the quality criterion of 0.85, further confirming the discriminate agility explain 66.68% of the variance from the perspective of customer
validity of our measurement model (Brown, 2015). We conducted service performance.
additional analyses by assessing cross loadings and the heterotrait– Furthermore, the results are robust after controlling for firm size.
monotrait ratio of correlations. Again, the results (Appendi X B) After adding firm size as the control variable and testing the
supported discriminant validity. Last, we calculated variance inflation hypotheses,
factors (VIF). The values were 1.74 and 2.86 for routine/innovative use the effects of both internal (β = 0.17, p < .05) and external agility
and internal/external agility, respectively. All values were below the 0.61, p < .001) remain significant, and the firm size has a positive
(β =
threshold of 3.3 (Petter, Straub, & Rai, 2007), so multicollinearity was effect on customer service performance (β= 0.12, p < .05). Overall,
not an issue. Thus, this study had acceptable psychometric properties. these results provide strong support for our model.
Finally, we tested the mediating effects of internal and external
4.3.2. Structural model testing agility on bootstrapping by using the PROCESS macro (Hayes, 2017).
We assessed path coefficients and R2 and tested our hypotheses. Constructed scores generated from SmartPLS were used to conduct the
The results appear in Fig. 2. They support H1, which states that analysis. When zero was not within the 95% confidence interval, the
internal agility is positively related to customer service performance
= indirect effects were significant (i.e., the mediating effects were signif-
(β 0.20, icant). The results (Table 5) show that external chatbot agility
p < .05). H2, which states that external agility is positively related to mediated the relationship between routine use of chatbots and
customer service performance, is also supported (β = 0.65, p < .001). customer service performance. Both internal and external chatbot
H3a proposes that the routine use of chatbots is positively related to agility mediated the relationship between innovative use of chatbots
internal agility. This hypothesis is also supported (β=0.23, p < .001). and customer service performance. These results further support our
H3b, arguing that the routine use of chatbots is positively related to model.
external agility, is supported as well (β =0.22, p < .001). H4a states
that the innovative use of chatbots is positively related to internal 4.3.3. Common method bias
agility. This hypothesis is supported (β = 0.53, p < .01). The We also assessed common method bias (CMB) (Podsakoff, MacK-
innovative enzie, Lee, & Podsakoff, 2003) because all the variable collection
occurred in one survey. First, the results of Harman’s single factor
Table 3 analysis showed three factors, and the first of these explained 39.64%
Item descriptive statistics. of the total variance. Second, we created a common method factor
Construct Item Mean SD Loading Alpha CR AVE including all items (Podsakoff et al., 2003). We then calculated the
variance explained by both the focal factor and the method for each
Routine use of RTN1 5.42 1.15 0.89 0.88 0.92 0.80
chatbots item following (Liang, Saraf, Hu, & Xue, 2007). The results show that
(Mean: 5.40; the average variance explained by the focal factor was.78, whereas the
SD: 1.09) variance explained by the method factor was.003, with a ratio of 281:1.
RTN2 5.35 1.18 0.89
Lastly, we assessed variance inflation factors (VIFs). According to one
study (Kock, 2015), CMB could be an issue when VIFs are greater than
SD: 1.16) RTN3 5.43 1.28 0.91
Innovative use 0.86 0.92 0.78 INV1 5.15 1.33 0.90
of chatbots
(Mean: 5.18;
SD: 1.23)
INV2 5.22 1.36 0.89
INV3 5.19 1.43 0.86
Internal agility 0.90 0.93 0.77 IA1 5.22 1.25 0.89
(Mean: 5.27;
SD: 1.08)
IA2 5.35 1.18 0.86
IA3 5.22 1.30 0.87
IA4 5.32 1.24 0.88
EXternal agility EA1 5.26 1.22 0.88 0.90 0.93 0.78
(Mean: 5.26;
SD: 1.13)
EA2 5.28 1.24 0.89
EA3 5.22 1.30 0.86
EA4 5.32 1.32 0.89
Customer CSP1 5.47 1.25 0.94 0.87 0.94 0.88
service
performance
(Mean: 5.43;
1
approaches to the use of chatbots, we divided our sample into two
3.3. Our analyses revealed that the VIFs for
CSP2 5.41 subsamples based on firm size. This post-hoc analysis provided further
1.23
routine and innovative use were both 1.74, and
insights for our research aims and helped us understand whether the
0.94 the VIFs for internal and external chatbot agility
strengths of the relationships presented in our model differed across
were both 2.86. All of these values were below
firm sizes. Seven participants did not report their firm sizes and were
3.3. Therefore, CMB seemed unlikely to be a
dropped from this analysis. According to one study (Dilger, 2017),
concern and unlikely to influence the research
firms with fewer than 500 employees are typically considered small-to-
results of this study.
medium firms. Therefore, we chose to divide our sample into small-to-
medium firms with fewer than 500 employees and large firms with 500
4.3.4. Post-hoc analysis
or more employees. We reran our model and compared the path
Because firms of different sizes may have
coefficients between two subsamples following Keil et al. (2000), using
different business processes and different
the following formula to calculate the t-value and evaluate the
1
Fig. 2. Model Results.
Table 5
The mediating effect of internal and external agility. Table 6
Post-hoc analysis (small-to-medium versus large firms).
Independent
Mediator Dependent Bootstrap analysis Hypothesis Small-to- Large Firms T-value Diff.
variable
variable Medium Firms (N = 177) Sig.?
Indirect Percentile
effect (N = 110)
confidence
H1: Internal chatbot 0.18 0.24** — 4.58 < ** *
Lower Upper
agility → Customer
Routine use of Internal Customer 0.08 —0.01 0.18 service performance
chatbots chatbot service H2: EXternal chatbot 0.67*** 0.61*** 5.33 > ** *
agility performance agility → Customer
EXternal 0.34 0.23 0.46 service performance
chatbot H3a: Routine use of 0.37*** 0.17* 18.51 > ** *
agility chatbots → Internal
Total 0.42 0.32 0.53 chatbot agility
Innovative use Internal Customer 0.10 0.01 0.23 H3b: Routine use of 0.37** 0.17* 17.38 > ** *
of chatbots chatbot service chatbots → EXternal
agility performance chatbot agility
EXternal 0.40 0.27 0.54 H4a: Innovative use of 0.41*** 0.58*** — 16.88 < ** *
chatbot chatbots → Internal
agility chatbot agility
Total 0.50 0.37 0.65 H4b: Innovative use of 0.41** 0.58*** — 14.95 < ** *
chatbots → EXternal
chatbot agility
significance levels of the differences:
*p < 0.05, **p < 0.01, ***p < 0.001, Diff. Sig. = different significantly.
t[ =
Path coefficientGroup1 — Path coefficentGroup2
] [ ]
√̅̅(̅m̅—̅̅̅1̅)̅2̅̅̅ ̅̅̅̅ ̅̅̅ ̅̅̅2̅̅̅̅̅̅̅̅̅̅̅ ̅̅̅̅̅̅(̅n̅—̅1̅)̅2̅̅̅̅ ̅̅̅̅̅ ̅̅̅ ̅̅̅2̅̅̅̅̅̅̅̅̅ √̅ ̅̅̅ ̅̅̅̅̅ ̅
1 1 m Chatbots share some counterwork. For example, they provide customers
×SE + (m+n—×SE ×
Group2 + n
with precise andcost
personalized
(m+n—
2)
Group1
2)
saves the labor of the callconsulting
center butservices, which
also makes usenot only greatly
of autonomous
where SEGroupi is the standard error of path in the structural model of interpretation.
group i, m is the sample size of Group 1 and n is the sample size of Group Two respondents remarked:
2. In the existing literature, this approach has been demonstrated as a
valid technique for testing subsample differences (e.g., Chiu, Wang,
Fang, & Huang, 2014; Lin & Wang, 2020; Zhang, Ma, Xu, & Xu, 2019).
The results (Table 6) show that all hypotheses were still significant
except for H1 for small-to-medium firms. Internal chatbot agility had a
stronger effect on customer service performance for large firms, whereas
external chatbot agility had a greater effect on small-to-medium firms.
Routine use of chatbots produced stronger effects on internal/external
chatbot agility for small-to-medium firms, whereas innovative use of
chatbots had greater effects on large firms. These research findings
offered some interesting insights into the use of chatbots and the impacts
across firms of different sizes. Nevertheless, the effects of routine and
innovative use on internal and external agility were significantly
different when comparing small-to-medium firms and large firms.
Future studies could further examine these differences.
4.4. Additional exploratory research
To test and substantiate the proposed research model with some

additional insights, we gave participants an opportunity to express
their views about the use of chatbots and its impacts through open-
ended questions. Below we summarize some responses and our
1
X. learning
Wang et technology to analyze users’ preferences and habits, actively seek International Journal of Information Management 66 (2022)
and identify potential needs, and improve user engagement and conver-
sion rate.
The data provided by Chatbot will also be used as an aspect of the
company’s business reference. Managers will need this data to do
business scheduling analysis through the data.
The responses indicate that routine business processes
incorporate chatbots (i.e., routine use of chatbots, as defined in the
Section 2.3 Routine Versus Innovative Use), which can enhance
business agility. Respondents also highlighted the role of the
innovative use of chatbots in creating agility. For example, one
participant highlighted the following:
Compared with the traditional marketing methods, we can understand
the users’ feedback through the data and call recording after natural lan-
guage processing and constantly optimize the marketing scheme, which
is very good.
This response further confirms that data collected by chatbots can
be analyzed by advanced methods to help identify new marketing
opportunities (i.e., innovative use of chatbots as defined in the
Routine Versus Innovative Use section).
Regarding the impacts of agility on customer service performance,
one participant stated the following:
1
For mobile banking and credit-card-exclusive software, it not only could The effect of routine use is stronger for small-to-medium firms than large
do the company’s business but also bring it to more areas that can be
consumed with the chat robot, increasing customer experience, cohesion,
and loyalty.
Another participant also highlighted the following:
In the initial stage of product promotion, users can be effectively main-
tained, the retention rate of users can be improved, and the experience of
users can be improved.
These responses helped to explain the important role of chatbots in
improving customer service performance. In summary, the responses
highlighted how the use of chatbots can improve firms’ flexibility (i.e.,
enhance agility) in responding to customers’ requirements and to mar-
keting changes, which would, in turn, improve customer service
performance.
5. Discussion
This study aimed to examine the effects of chatbot-enabled

business agility by exploring its antecedents and consequences. We
identified two antecedents: routine and innovative use of chatbots. We
explained how these two types of chatbot use could create business
agility, consisting of internal and external agility. In addition, we
argued that both internal and external agility can improve customer
service performance. Based on the survey data collected from 294
employees, our results provide strong support for our model.
Our results show that both the routine and the innovative use of
chatbots increase business agility. Roberts et al. (2016) reported that
the innovative use of decision support systems enhances managers’
sensing abilities, which refers to the capacity to identify opportunities
for organizational innovation. Altschuller, Gelb, and Henry (2010) also
reported that innovative investment and use of IT can enhance firm
performance. These studies provided insights into how traditional IS
could bring benefits to business; however, they cannot be generalized
to newly developed IS such as those that use AI. In our study, the focus
on chatbot context allowed us to develop an understanding (Hong,
Chan, Thong, Chasalow, & Dhillon, 2014) of how emerging AI tools
can enhance business agility and thus benefit businesses. Our study
also extends the reach of the literature by demonstrating additional
benefits such as agility, which is an essential digitally enabled
capability for business sustainability in the era of competitive
marketing (Ashrafi, Ravasan, Trkman, & Afshari, 2019; Kappelman et
al., 2019).
In addition, Roberts et al. (2016) reported that routine use of deci-
sion support systems does not enhance managers’ sensing abilities.
However, our study shows that routine use of chatbots can increase
internal and external agility, although its effects are smaller than those
brought by the innovative use of chatbots. In other words, once inte-
grated into marketing employees’ work processes, chatbots can help
employees understand market changes and provide offers efficiently.
Therefore, it is possible that the effect of the routine use of IS could
evolve as long as the technology advances (Petter et al., 2012). In the
context of traditional IT, routine use may not be a good approach for
improving business operation and outcomes because of its relatively
limited capabilities. Emerging technologies empowered by new
features and functions, such as AI, are likely to address those issues
and can thus change the way information systems are used as a routine
process for achieving business benefits. The results of this study
support this idea. As such, our study expands the literature by
clarifying how different types of IT use increase internal and external
agility in newly developed IT such as AI. This is also consistent with
the research of Sambamurthy, Bharadwaj, and Grover (2003), which
showed that IT investment and capabilities can enhance
organizations’ agility.
Our post-hoc analysis further showed that the effect of routine versus
innovative use on internal and external agility differs across firm sizes.
1
firms, whereas the effect of innovative use is greater for large firms chatbots in reaping business benefits (i.e., business agility) as well as
than for small-to-medium firms. It is possible that the business the role of
processes of small-to-medium firms are simpler due to having few
resources; therefore, there are not many options for chatbots to
integrate into these businesses. In contrast, larger firms have more
complex business processes and can implement innovative ways to
use chatbots. Because our analysis was exploratory, future studies are
needed to further examine how firms of different sizes adopt
chatbots.
Last, our results reveal that internal and external agility can ulti-
mately improve customer service performance. This is consistent with
the findings of Chuang (2020), who reported that social media agility
is positively related to the strength of customer–firm relationships.
Customer service is a primary element in business sustainability (Huang,
Niu, & Pan, 2021). Our study provides further insights into how
customer service can be enhanced by digitally enabled agility created
by emerging AI tools. Our research results support this relationship
between agility and customer service based on data collected from
marketing employees who are knowledgeable about customer
service. Our post-hoc analysis further showed that internal chatbot
agility has a stronger effect on customer service performance for large
firms than for small-to-medium firms, whereas external chatbot
agility has a greater effect on small-to-medium firms than on large
firms. Small-to-medium firms, which have fewer resources, become
more vulnerable to market changes, so increased external agility
would greatly improve their ability to respond to market changes and
serve customers. On the other hand, larger firms usually follow
standardized procedures to deal with customers, so internal agility,
which better supports these standardized procedures, would
tremendously enhance customer services for large firms. Our findings
entail important theoretical and practical implications, which are
presented below.
5.1. Theoretical implications
Our study is among the very first studies to contribute to the

literature of emerging AI by exploring employees’ use of chatbots
and related outcomes. This study makes two primary theoretical
contributions. First, it describes the effect of chatbot use on business
agility. Based on the literature (Li et al., 2013; Roberts et al., 2016),
we argued that employees can engage in both routine and innovative
use of chatbots. Our study thus extends the literature of routine and
innovative use to the context of chatbots. Such conceptualization is
important to help explain how chatbots can be used in organizations.
Further, Helfat and Martin (2015) summarized that dynamic
capabilities rely on managerial cognition, social capital, and human
capital. By showing that IT can also contribute to agility, our study
extends the literature on the DCV and highlights an additional avenue
for the development of dynamic capabilities.
Although both routine and innovative use of chatbots can increase
both internal and external agility, our results show that innovative use of
chatbots plays a more important role in creating agility than does
routine use. Our results highlight the importance of innovative use in
creating business agility in the context of advanced technologies. The
literature has suggested that individuals can vary regarding how they
use this technology (e.g., features utilized, time invested) after tech-
nology adoption (Nambisan et al., 1999). Thus, innovative use is often
proactive and initiated by employees, a finding consistent with the
perspective that innovation is a human-driven phenomenon (Jero
´nimo, Henriques, de Lacerda, da Silva, & Vieira, 2020; Lee &
Raschke, 2020). As such, innovative use is vital because it allows
companies to gain additional benefits from their IT investment
(Venkatesh, Brown, Mar- uping, & Bala, 2008) that cannot be
achieved through routine use (Roberts et al., 2016). Failure to engage
in innovative use could thus result in the underutilization of IT
(Venkatesh et al., 2008), preventing
companies from achieving the full potential value of IT. Our study has
therefore highlighted the essential role of the innovative use of
1
the routine use of chatbots. It has also further confirmed the

importance of conceptualizing different types of use of newly Our study also highlights the fact that both routine and innovative
developed IT tools, such as AI, which requires further attention. use of chatbots can enhance agility. Therefore, firms should not only
Overall, our study contributes to the literature by explaining how require marketing employees to integrate chatbots into standardized
chatbot uses lead to chatbot-enabled business agility. marketing processes, but also encourage them to try new approaches
Besides this, our study extends the agility concept to the emerging to use chatbots to generate new insights about customers. Since routine
context of AI by conceptualizing chatbot-enabled agility using two di- use of chatbots can enhance agility, companies should integrate
mensions: internal and external chatbot agility. This is a valuable chatbots into routine processes of customer services. For example,
contribution because agility plays a critical role in identifying new ways chatbots can answer common questions from customers and collect
that IT can support business strategies and improve performance (Kar- their feedback to help identify customers’ needs and market changes.
ahanna, Xu, Xu, & Zhang, 2018; Queiroz et al., 2018). An enhanced Firms can also encourage employees to use chatbots innovatively to
understanding of chatbot agility could provide insight into how the generate further insights from their interactions with the chatbots and
intersection of business strategies and advanced technology can create customers. Our results show that innovative use has a stronger effect
value. It could also help us rethink the advantages of technological ad- than routine use. Therefore, innovative use should be more
vances, thus motivating the design and deployment of emerging tech- emphasized when chatbots are incorporated to support employees’
nologies and changing the landscape of IT use within organizations jobs and business operations. For example, firms can allow marketing
(Petter et al., 2012). employees to hold regular meetings and discuss different approaches
Second, our results highlight the fact that chatbot-enabled business to chatbot use, which will generate new insights and allow for the
agility can significantly improve customer service performance. These learning of new usage methods from peers. The literature has also
findings suggest that chatbot-enabled agility is critical for companies found that intrinsic motivation can support innovative use (Li et al.,
to create sustained competitive advantages through responding to 2013). Therefore, managers need to support employees’ intrinsic
customer preference and market change (e.g., Chuang, 2020). There- motivation. For example, managers can help employees set up
fore, this study brings new insights into IT-enabled business agility reasonable performance objectives for customer service. They can also
and allows future studies to explore the role of emerging technologies allow some autonomy and encourage employees to try different features
in business (e.g., Nishant, Kennedy, & Corbett, 2020; Sipior, 2020). or approaches to the analysis of customers’ data. The purpose of this
For example, future studies could explore how innovative and routine would be to support employees to go beyond routine chatbot use to
use of chatbots (or other AI tools) can lead to different organizational achieve further benefits for the company.
outcomes (e.g., Gursoy, Chi, Lu, & Nunkoo, 2019). Although our results show that the effect of innovative use is
weaker for small-to-medium firms, we hesitate to suggest that small-
5.2. Practical implications to-medium firms should focus less on innovative use. It is possible that
small-to- medium firms have simpler business processes and fewer
Our study also has important practical implications. First, our resources; therefore, they may find it difficult to use chatbots
study shows that both internal and external chatbot agility relate to innovatively. Such innovative use, however, may have tremendous
customer service performance. This finding provides practitioners with impact once identified because the literature has shown that innovative
an enhanced understanding of the important role of chatbot-enabled use can allow firms to generate further benefits from IT (Venkatesh et al.,
agility in improving customer service. As consumers spend more time 2008). Therefore, we suggest that even small-to-medium firms try to
in digital environments facilitated by AI, companies are investing explore innovative approaches to using chatbots.
heavily in the development of AI tools and moving into this new
environment (Ameen et al., 2021). As such, our study could assist 5.3. Limitations and suggestions for future studies
business leaders with improving customer service via leveraging
chatbots into business operations and creating business agility. Our study has several limitations. First, a survey company recruited
Based on the relationships validated in this study, business leaders our sample. Although our participants came from different companies, it
may develop or choose efficient chatbot tools that enable them to is still possible that our sample was biased. Second, we focused on U.S.
create an innovative digital environment wherein they can better employees, so it is possible that the results might not be generalizable
interact with customers, identify customer demand, and respond to to other cultural backgrounds. For example, employees of certain back-
market changes. Businesses can thus improve their customer service grounds might be more likely than those from the U.S. to strictly follow
performance, which is likely to promote their sustainability and orders from supervisors and thus engage in only the routine use of
produce many long-term benefits for them (Lee & Raschke, 2020). For chatbots. By focusing on routine and innovative use of chatbots, our
example, to improve customer experience and satisfaction, companies study captures the uniqueness of chatbots because varied types of
could choose chatbots that create internal agility through features such technology are used differently. For example, unless combined with
as automatically answering customers’ questions, resolving customers’ chatbots or other service assistants, social media cannot automatically
problems, and identifying customers’ preferences. To better serve answer consumers’ standard questions. Nevertheless, future studies
customers and remain competitive, companies could choose chatbots could examine how specific features of chatbots lead to agility and
that create external agility with data analysis to identify changes in improve customer service.
customer demand and the market. These practical approaches help Our paper used items focusing on enhanced customer-firm re-
managers determine the type of chatbot agility needed. lationships and overall performance to measure customer service per-
In addition, the results of our study indicate external agility has a formance. This could be a limitation as they do not capture all aspects
larger effect on customer service performance than internal agility. of customer service performance (Setia et al., 2013). In future studies,
Some companies may want to focus more on external agility and utilize re- searchers could consider adopting more comprehensive items for
chatbots to interpret market changes, especially whether marketing measuring customer service performance when assessing chatbots and
changes are what hinder the company in establishing excellent customer other AI tools. Future studies could also attempt to create new scales
services performance. These benefits provide meaningful reasons for for measuring customer services for different contexts.
business leaders to incorporate chatbots into business strategies and In our study, we focused on marketing employees who used
maximize their competitive advantages. Alternatively, the research re- chatbots to interact with customers. Unlike customers, who use
sults of this study could provide initial implications for businesses to chatbots to receive service, marketing employees use chatbots to
leverage other AI tools in business practices. provide customer service. Therefore, it was appropriate to collect data
from marketing employees because they understand how product
1
offers are provided and to what extent chatbots enhance customer
service performance.
1
Customers could provide another perspective to understand the effec-

tiveness of chatbot services, and we suggest that future studies enabled agility. The data collected came from a survey of 294
examine this important issue. marketing employees who had experience with the use of chatbots
Future studies could extend our study in several ways. First, future within organizations. The results show that the effect of innovative use
studies could test our model with different types of chatbots to assess on internal and external agility is stronger than that of routine use.
which type of chatbot is more effective in other contexts, such as Further, both internal and external agility enhance customer service
human resource management. For example, chatbots that support performance. Our results demonstrate how companies can achieve
employees and those that support customers will probably be used in business agility by effectively integrating chatbots into business
different ways. Second, field experiments could be conducted in future operations based on different types of use. Our study thus provides
studies to examine the optimal combination of chatbot use contexts useful insights regarding ways of achieving business benefits (i.e.,
and its effect on chatbot agility. Third, other consequences of the use of enhanced customer service) through the incorporation of chatbots into
chatbots could be examined. For example, future studies could business operations. Future studies could extend our work regarding
examine firm performance or employee productivity. Fourth, studies how certain organizational variables could moderate the relationship
could examine how certain variables (e.g., organizational culture) between chatbot use and its outcomes. Future studies could also
moderate the relationship between chatbot use and its consequences. examine work- versus non-work- related benefits for employees and
Last, future studies could also examine factors that influences additional business outcomes resulting from the use of chatbots
continuance intention in the context of chatbot use. within organizations.
6. Conclusion CRediT authorship contribution statement
Chatbots have been increasingly used to support customer service. Xuequn Wang: Conceptualization, Methodology, Investigation,
This study aimed to examine how chatbot-enabled agility improved Formal analysis, Writing— original draft, Writing review & editing,
customer service performance and to explore the antecedents of chatbot- Xiaolin Lin: Conceptualization, Methodology, Investigation, Formal
analysis, Writing — original draft, Writing — review & editing, Bin Shao:
Conceptualization, Writing — original draft, Writing — review &
editing.
Appendix A. Measures
Routine Use of Chatbots (Li et al., 2013)

RTN1 The use of chatbots has been incorporated into our regular business operations.
RTN2 The use of chatbots is pretty much integrated as part of our normal business operations.
RTN3 The use of chatbots is now a normal part of our business operations.
Innovative Use of Chatbots (Li et al., 2013)
INV1 My company has discovered new uses of chatbots to enhance business operations.
INV2 My company has used chatbots in novel ways to support business operations.
INV3 My company has developed new applications, based on chatbot use, to support business operations.
Internal Chatbot Agility (Akhtar et al., 2018; Chuang, 2020)
Because of the use of chatbots in my company, the following is true:
IA1 The reliability of my company’s offerings (i.e., products and services) has increased.
IA2 My company’s day-to-day operations have been flexible for customized demand.
IA3 My company’s offerings (i.e., products and services) have been more cost efficient than those of our competitors.
IA4 My company has delivered our offerings (i.e., products and services) more quickly.
External Chatbot Agility (Akhtar et al., 2018; Chuang, 2020)
EA1 My company has responded very reliably to market changes.
EA2 My company has had greater flexibility in our offerings to adapt to market changes.
EA3 My company has efficiently redesigned our offerings to adapt to market changes.
EA4 My company has been very quick to adapt to market opportunities.
Customer Service Performance (Chuang, 2020) (Setia et al., 2013)
CSP1 My company has established good relationships with customers.
CSP2 My company has provided better customer service than ever.
Appendix B. Additional analysis of discriminant validity
See AppendiX Tables B1 and B2.

This appendiX presents additional analysis results. Table B1 shows that all loadings are higher than cross-loadings. Table B2 shows that all values of
Heterotrait-Monotrait ratio of correlations are below 0.90 (Teo, Srivastava, & Jiang, 2008). These results provide additional support for
discriminant validly.
1
Table B.1
Loadings and cross-loading. Coombs, C. (2020). Will COVID-19 be the tipping point for the intelligent automation of
work? A review of the debate and implications for research. International Journal
Routine Innovative Internal EXternal Customer of Information Management, 55, Article 102182.
Use of Use of Chatbot Chatbot Service Cooper, R. B., & Zmud, R. W. (1990). Information technology implementation research: a
Chatbots Chatbots Agility Agility Performance technological diffusion approach. Management Science, 36(2), 123–139.
Da Silveira, G., Borenstein, D., & Fogliatto, F. S. (2001). Mass customization: Literature
RTN1 0.89 0.54 0.49 0.50 0.53
review and research directions. International Journal of Production Economics, 72(1), 1–
RTN2 0.89 0.62 0.53 0.51 0.51
13.
RTN3 0.91 0.58 0.53 0.52 0.56
Davenport, T. H., & Ronanki, R. (2018). Artificial intelligence for the real world. Harvard
INV1 0.59 0.90 0.61 0.61 0.61 Business Review, 96(1), 108–116.
INV2 0.57 0.89 0.59 0.60 0.56 Dilger, R. J. (2017). Small Business Size Standards: A Historical Analysis of
INV3 0.56 0.86 0.60 0.59 0.54 Contemporary Issues. Congressional Research Service, July, 21, 7–5700.
IA1 0.52 0.63 0.89 0.73 0.68 Duan, J., Faker, P., Fesak, A., & Stuart, T. (2013). Benefits and drawbacks of cloud-based
IA2 0.52 0.58 0.86 0.67 0.60 versus traditional ERP systems. Proceedings of the 2012–13 course on Advanced
IA3 0.48 0.60 0.87 0.72 0.61 Resource Planning.
IA4 0.49 0.57 0.88 0.70 0.63 Duan, Y., Edwards, J. S., & Dwivedi, Y. K. (2019). Artificial intelligence for decision
EA1 0.50 0.58 0.70 0.88 0.70 making in the era of Big Data–evolution, challenges and research agenda.
EA2 0.52 0.55 0.74 0.89 0.72 International Journal of Information Management, 48, 63–71.
EA3 0.47 0.59 0.68 0.86 0.70 Dwivedi, Y. K., Hughes, L., Ismagilova, E., Aarts, G., Coombs, C., Crick, T., &
Williams, M. D. (2021). Artificial Intelligence (AI): Multidisciplinary perspectives on
EA4 0.51 0.66 0.72 0.89 0.72
emerging challenges, opportunities, and agenda for research, practice and policy.
CSP1 0.54 0.60 0.70 0.76 0.94
International Journal of Information Management, 57, Article 101994.
CSP2 0.57 0.61 0.65 0.76 0.94 https://doi.org/ 10.1016/j.ijinfomgt.2019.08.002
Etemad-Sajadi, R., & Ghachem, L. (2015). The impact of hedonic and utilitarian value of
online avatars on e-service quality. Computers in Human Behavior, 52, 81–86.
Gao, P., Zhang, J., Gong, Y., & Li, H. (2020). Effects of technical IT capabilities on
organizational agility: The moderating role of IT business spanning capability.
Table B.2 Industrial Management & Data Systems, 120(5), 941–961.
Heterotrait-monotrait ratio of correlations. Gupta, S., Drave, V. A., Dwivedi, Y. K., Baabdullah, A. M., & Ismagilova, E. (2020a).
Achieving superior organizational performance via big data predictive analytics: A
1 2 3 4 5 dynamic capability view. Industrial Marketing Management, 90, 581–592.
Gupta, S., Meissonier, R., Drave, V. A., & Roubaud, D. (2020b). EXamining the impact of
1 Routine use of chatbots
Cloud ERP on sustainable performance: A dynamic capability view. International
2 Innovative use of chatbots 0.75
Journal of Information Management, 51, Article 102028.
3 Internal agility 0.65 0.77 Gursoy, D., Chi, O. H., Lu, L., & Nunkoo, R. (2019). Consumers acceptance of artificially
4 EXternal agility 0.64 0.77 0.88 intelligent (AI) device use in service delivery. International Journal of Information
5 Customer service performance 0.61 0.66 0.69 0.88 Management, 49, 157–169.
Hair, J. F., Jr., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2016). A Primer on Partial Least
Squares Structural Equation Modeling (PLS-SEM). Sage Publications,.
References Hayes, A. F. (2017). Introduction to Mediation, Moderation, and Conditional Process
Analysis: A Regression-based Approach. Guilford Publications,.
He, W., Zhang, Z. J., & Li, W. (2021). Information technology solutions, challenges, and
Adam, M., Wessel, M., & Benlian, A. (2020). AI-based chatbots in customer service and
suggestions for tackling the COVID-19 pandemic. International Journal of Information
their effects on user compliance. Electronic Markets, 1–19.
Management, 57, Article 102287.
Akhtar, P., Khan, Z., Tarba, S., & Jayawickrama, U. (2018). The Internet of Things,
Helfat, C. E., & Martin, J. A. (2015). Dynamic managerial capabilities: Review and
dynamic data and information processing capabilities, and operational agility.
assessment of managerial impact on strategic change. Journal of Management, 41(5),
Technological Forecasting and Social Change, 136, 307–316.
1281–1312.
Altschuller, S., Gelb, D. S., & Henry, T. F. (2010). IT as a resource for competitive agility:
Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing
an analysis of firm performance during industry turbulence. Journal of International
discriminant validity in variance-based structural equation modeling. Journal of the
Technology and Information Management, 19(1), 39–59.
Academy of Marketing Science, 43(1), 115–135.
Ameen, N., Tarhini, A., Reppel, A., & Anand, A. (2021). Customer experiences in the age
Holmqvist, M., & Pessi, K. (2006). Agility through scenario development and continuous
of artificial intelligence. Computers in Human Behavior, 114, Article 106548.
implementation: a global aftermarket logistics case. European Journal of Information
https:// doi.org/10.1016/j.chb.2020.106548
Systems, 15(2), 146–158.
Ashfaq, M., Yun, J., Yu, S., & Loureiro, S. M. C. (2020). I, Chatbot: Modeling the
Hong, W., Chan, F. K., Thong, J. Y., Chasalow, L. C., & Dhillon, G. (2014). A framework
determinants of users’ satisfaction and continuance intention of AI-powered service
and guidelines for context-specific theorizing in information systems research.
agents. Telematics and Informatics, 54, Article 101473. https://doi.org/10.1016/j.
Information Systems Research, 25(1), 111–136.
tele.2020.101473
Hu, Q., Lu, Y., Pan, Z., Gong, Y., & Yang, Z. (2021). Can AI artifacts influence human
Ashrafi, A., Ravasan, A. Z., Trkman, P., & Afshari, S. (2019). The role of business
cognition? The effects of artificial autonomy in intelligent personal assistants.
analytics capabilities in bolstering firms’ agility and performance. International
International Journal of Information Management, 56, Article 102250.
Journal of Information Management, 47, 1–15.
Huang, M.-H., & Rust, R. T. (2018). Artificial intelligence in service. Journal of Service
Bae, S.-M. (2017). The relationship between the type of smartphone use and smartphone
Research, 21(2), 155–172.
dependence of Korean adolescents: National survey study. Children and Youth
Huang, P.-Y., Niu, B., & Pan, S. L. (2021). Platform-based customer agility: An integrated
Services Review, 81, 207–211.
framework of information management structure, capability, and culture.
Benner, M. J., & Tushman, M. (2002). Process management and technological
International Journal of Information Management, 59. https://doi.org/10.1016/j.
innovation: A longitudinal study of the photography and paint industries.
ijinfomgt.2021.102346
Administrative Science Quarterly, 47(4), 676–707.
Hui, S. C., Fong, A., & Jha, G. (2001). A web-based intelligent fault diagnosis system for
Borges, A. F., Laurindo, F. J., Spínola, M. M., Gonçalves, R. F., & Mattos, C. A. (2020).
customer service support. Engineering Applications of Artificial Intelligence, 14(4), 537–
The strategic use of artificial intelligence in the digital era: Systematic literature
548.
review and future research directions. International Journal of Information
Huo, B., Gu, M., & Wang, Z. (2018). Supply chain flexibility concepts, dimensions and
Management, 57, Article 102225. https://doi.org/10.1016/j.ijinfomgt.2020.102225
outcomes: an organisational capability perspective. International Journal of
Brown, T. A. (2015). Confirmatory Factor Analysis for Applied Research. Guilford
Production Research, 56(17), 5883–5903.
Publications,.
Jain, M., Kumar, P., Kota, R., & Patel, S. N. (2018). Evaluating and informing the design of
Chiu, C., Wang, E., Fang, Y., & Huang, H. (2014). Understanding customers’ repeat
chatbots. Paper presented at the Proceedings of the 2018 Designing Interactive
purchase intentions in B2C e-commerce: the roles of utilitarian value, hedonic value
Systems Conference.
and perceived risk. Information Systems Journal, 24(1), 85–114.
Jasperson, J. S., Carter, P. E., & Zmud, R. W. (2005). A comprehensive conceptualization
Chiu, M.-C., & Chuang, K.-H. (2021). Applying transfer learning to achieve precision
of post-adoptive behaviors associated with information technology enabled work
marketing in an omni-channel system–a case study of a sharing kitchen platform.
systems. MIS Quarterly, 29(3), 525–557.
International Journal of Production Research, 1–16.
Jero´nimo, H. M., Henriques, P. L., de Lacerda, T. C., da Silva, F. P., & Vieira, P. R. (2020).
Chuang, S.-H. (2020). Co-creating social media agility to build strong customer-firm
Going green and sustainable: The influence of green HR practices on the
relationships. Industrial Marketing Management, 84, 202–211.
organizational rationale for sustainability. Journal of Business Research, 112,
Chung, M., Ko, E., Joung, H., & Kim, S. J. (2020). Chatbot e-service and customer
413–421.
satisfaction regarding luXury brands. Journal of Business Research, 117, 587–595.
Kappelman, L., Torres, R., McLean, E., Maurer, C., Johnson, V., & Kim, K. (2019). The
Claessen, V., Schmidt, A., & Heck, T. (2017). Virtual Assistants-A Study on the Usability
and User Perception of Customer Service Systems for E-Commerce. Paper presented 2018 SIM IT issues and trends study. MIS Quarterly Executive, 18(1), 7.
at the ISI. Karahanna, E., Xu, S. X., Xu, Y., & Zhang, N. A. (2018). The needs–affordances–features
perspective for the use of social media. MIS Quarterly, 42(3), 737–756.
1
Keil, M., Tan, B. C., Wei, K.-K., Saarinen, T., Tuunainen, V., & Wassenaar, A. (2000). Roy, R., & Naidoo, V. (2021). Enhancing chatbot effectiveness: The role of
A cross-cultural study on escalation of commitment behavior in software projects. anthropomorphic conversational styles and time orientation. Journal of Business
MIS Quarterly, 299–325. Research, 126, 23–34.
Kock, N. (2015). Common method bias in PLS-SEM: A full collinearity assessment Sambamurthy, V., Bharadwaj, A., & Grover, V. (2003). Shaping agility through digital
approach. International Journal of e-Collaboration (ijec), 11(4), 1–10. options: Reconceptualizing the role of information technology in contemporary
Lalicic, L., & Weismayer, C. (2021). Consumers’ reasons and perceived value co-creation firms. MIS Quarterly, 27(2), 237–263.
of using artificial intelligence-enabled travel service agents. Journal of Business Schuetzler, R. M., Grimes, G. M., & Scott Giboney, J. (2020). The impact of chatbot
Research, 129, 891–901. conversational skill on engagement and perceived humanness. Journal of
Lee, M. T., & Raschke, R. L. (2020). Innovative sustainability and stakeholders’ shared Management Information Systems, 37(3), 875–900.
understanding: The secret sauce to “performance with a purpose”. Journal of Business Setia, P., Setia, P., Venkatesh, V., & Joglekar, S. (2013). Leveraging digital technologies:
Research, 108, 20–28. How information quality leads to localized capabilities and customer service
Li, X., Hsieh, J. P.-A., & Rai, A. (2013). Motivational differences across post-acceptance performance. MIS Quarterly, 565–590.
information system usage behaviors: An investigation in the business intelligence Shumanov, M., & Johnson, L. (2021). Making conversations with chatbots more
systems context. Information Systems Research, 24(3), 659–682. personalized. Computers in Human Behavior, 117, Article 106627. https://doi.org/
Liang, H., Saraf, N., Hu, Q., & Xue, Y. (2007). Assimilation of Enterprise Systems: The 10.1016/j.chb.2020.106627
Effect of Institutional Pressures and the Mediating Role of Top Management. MIS Sipior, J. C. (2020). Considerations for development and use of AI in response to COVID-
Quarterly, 31(1), 59–87. 19. International Journal of Information Management, 55, Article 102170.
Lin, X., & Wang, X. (2020). EXamining gender differences in people’s information- Tallon, P. P., & Pinsonneault, A. (2011). Competing perspectives on the link between
sharing decisions on social networking sites. International Journal of Information strategic information technology alignment and organizational agility: insights from
Management, 50, 45–56. a mediation model. MIS Quarterly, 35(2), 463–486.
Lokshin, B., Belderbos, R., & Carree, M. (2008). The productivity effects of internal and Tams, S., Thatcher, J. B., & Craig, K. (2018). How and why trust matters in post-adoptive
external R&D: Evidence from a dynamic panel data model. Oxford Bulletin of usage: The mediating roles of internal and external self-efficacy. The Journal of
Economics and Statistics, 70(3), 399–413. Strategic Information Systems, 27(2), 170–190.
Lu, Y., & Ramamurthy, K. (2011). Understanding the link between information Teece, D., Peteraf, M., & Leih, S. (2016). Dynamic capabilities and organizational agility:
technology capability and organizational agility: An empirical examination. MIS Risk, uncertainty, and strategy in the innovation economy. California Management
Quarterly, 931–954. Review, 58(4), 13–35.
Mathiassen, L., & Pries-Heje, J. (2006). Business agility and diffusion of information Teo, T. S., Srivastava, S. C., & Jiang, L. (2008). Trust and electronic government success:
technology. European Journal of Information Systems, 15(2), 116–119. An empirical study. Journal of Management Information Systems, 25(3), 99–132.
McLean, G., & Osei-Frimpong, K. (2019). Chat now… EXamining the variables Trischler, J., Johnson, M., & Kristensson, P. (2020). A service ecosystem perspective on
influencing the use of online live chat. Technological Forecasting and Social Change, the diffusion of sustainability-oriented user innovations. Journal of Business Research,
146, 55–67. 116, 552–560.
Mikalef, P., & Gupta, M. (2021). Artificial intelligence capability: Conceptualization, Venkatesh, V., Brown, S. A., Maruping, L. M., & Bala, H. (2008). Predicting different
measurement calibration, and empirical study on its impact on organizational conceptualizations of system use: the competing roles of behavioral intention,
creativity and firm performance. Information & Management, 58(3), Article 103434. facilitating conditions, and behavioral expectation. MIS Quarterly, 32(3), 483–502.
Mordor Intelligence, 2021, Chatbot market - growth, trends, COVID-19 impact, and Watson, H. J. (2019). Update tutorial: Big Data analytics: Concepts. Technology, and
forecasts (2021–2026). 〈 https://www.mordorintelligence.com/industry-reports/ Applications Communications of the Association for Information Systems, 44(1),
chatbot-market〉, Accessed on September 20, 2021. 364–379.
Nambisan, S., Agarwal, R., & Tanniru, M. (1999). Organizational mechanisms for Weber, Y., & Tarba, S. Y. (2014). Strategic agility: A state of the art introduction to the
enhancing user innovation in information technology. MIS Quarterly, 23(3), special section on strategic agility. California Management Review, 56(3), 5–12.
365–395. Williams, R., 2019, Study: Chatbots to drive $112B in retail sales by 2023. 〈https://www.
Ngai, E. W., Chau, D. C., & Chan, T. (2011). Information technology, operational, and mobilemarketer.com/news/study-chatbots-to-drive-112b-in-retail-sales-by-2023/
management competencies for supply chain agility: Findings from case studies. The 554416/ 〉, Accessed on May 24, 2021.
Journal of Strategic Information Systems, 20(3), 232–249. Zhang, X., Ma, L., Xu, B., & Xu, F. (2019). How social media usage affects employees’ job
Nishant, R., Kennedy, M., & Corbett, J. (2020). Artificial intelligence for sustainability: satisfaction and turnover intention: An empirical study in China. Information &
Challenges, opportunities, and a research agenda. International Journal of Information Management, 56(6), Article 103136. https://doi.org/10.1016/j.im.2018.12.004
Management, 53, Article 102104. Zhang, Y., Qu, T., Ho, O. K., & Huang, G. Q. (2011). Agent-based smart gateway for RFID-
Nold, H., Anzengrubcr, J., Woclfle, M., & Michel, L. (2018). Organizational agility- enabled real-time wireless manufacturing. International Journal of Production
testing, validity, and reliability of a diagnostic instrument. Journal of Organizational Research, 49(5), 1337–1352.
Psychology, 18(3), 104–117.
Overby, E., Bharadwaj, A., & Sambamurthy, V. (2006). Enterprise agility and the
Xuequn (Alex) Wang is a Senior Lecturer in Edith Cowan University. He received his Ph.
enabling role of information technology. European Journal of Information Systems, 15
D. in Information Systems from Washington State University. His research interests
(2), 120–131.
include social media, privacy, e-commerce, and human-computer interaction. His
Petter, S., DeLone, W., & McLean, E. R. (2012). The past, present, and future of “IS
research has appeared in MIS Quarterly, Information Systems Journal, Information &
success”. Journal of the Association for Information Systems, 13(5), 341–362.
Management, Communications of the ACM, ACM Transactions, and Communications of
Petter, S., Straub, D., & Rai, A. (2007). Specifying formative constructs in information
the Association for Information Systems, among others.
systems research. MIS Quarterly, 623–656.
Phansalkar, S., Kamat, P., Ahirrao, S., & Pawar, A. (2019). Decentralizing AI applications
with block chain. International Journal of Scientific & Technology Research, 8(9), 362– Xiaolin Lin is currently assistant professor of management information systems in the
370. college of business at California State University, Sacramento. Prior to that, he was an
Pizzi, G., Scarpi, D., & Pantano, E. (2021). Artificial intelligence and the new forms of assistant professor of computer information systems in the Paul and Virginia Engler Col-
interaction: Who has the control when interacting with a chatbot? Journal of Business lege of Business at West Texas A&M University. He received his Ph.D. in management
Research, 129, 878–890. information systems from Washington State University. His research focuses on the im-
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common plications of IT in e-commerce and healthcare, human-AI interaction, cyber security, and
method biases in behavioral research: a critical review of the literature and gender differences in IT behavioral research. He has published papers in premier journals
recommended remedies. Journal of Applied Psychology, 88(5), 879–903. such as Journal of Business Ethics, European journal of Information Systems, Decision
Queiroz, M., Tallon, P. P., Sharma, R., & Coltman, T. (2018). The role of IT application Sciences, Information & Management, International Journal of Electronic Commerce,
orchestration capability in improving agility and performance. The Journal of Journal of the Association for Information Science and Technology, and International
Strategic Information Systems, 27(1), 4–21. Journal of Information Management, among others. He has also presented numerous pa-
Rese, A., Ganster, L., & Baier, D. (2020). Chatbots in retailers’ customer communication: pers at international and national conferences.
How to measure their acceptance? Journal of Retailing and Consumer Services, 56,
Article 102176.
Bin Shao is a professor of decision management & Terry Professor of Business in the
Roberts, N., Campbell, D. E., & Vijayasarathy, L. R. (2016). Using information systems to
Department of Computer Information and Decision Management, Paul and Virginia
sense opportunities for innovation: Integrating postadoptive use behaviors with the
Engler College of Business, West Texas A&M University. She earned her Ph.D. from
dynamic managerial capability perspective. Journal of Management Information
University of Illinois at Urbana-Champaign. Her current research interests are mainly in
Systems, 33(1), 45–69.
the interdis- ciplinary areas of information systems, decision management, and
Rosenthal, R., & Rosnow, R. L. (2008). Essentials of Behavioral Research: Methods and Data
marketing. Her works have been published in International Advances in Economic
Analysis, 3rd ed. Boston: McGraw-Hill,.
Research, International Journal of Applied Management Science, Journal of Supply Chain
and Operations Man- agement, Journal of Management Information and Decision
Sciences, and many others.

1 s2.0 S026840122200069X Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S026840122200069X Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S026840122200069X Main

Uploaded by

Copyright:

Available Formats

International Journal of Information Management 66 (2022) 102535

Contents lists available at ScienceDirect

International Journal of Information Management

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

How does artificial intelligence create business agility? Evidence

customer service. It is meaningful to study business agility by exploring

2.2. Business agility: a chatbot perspective

The term agility refers to the ability of a company or an organization

leadership, systems, and people. of IT is one way firms can sustainably

2.3. Routine versus innovative use

After chatbots have been implemented into organizations, they

capabilities (Teece et al., 2016) because they can efficiently integrate

companies further understand customer preferences and market

performance. Last, PLS requires a smaller sample size (Henseler et al.,

1 Routine use of chatbots 0.90

Fig. 2. Model Results.

4.4. Additional exploratory research

To test and substantiate the proposed research model with some

This study aimed to examine the effects of chatbot-enabled

5.1. Theoretical implications

Our study is among the very first studies to contribute to the

the routine use of chatbots. It has also further confirmed the

Customers could provide another perspective to understand the effec-

6. Conclusion CRediT authorship contribution statement

Routine Use of Chatbots (Li et al., 2013)

Appendix B. Additional analysis of discriminant validity

See AppendiX Tables B1 and B2.

You might also like