Mobile Instant Messaging for Customer Service Interaction
Preparation of a Model-based Approach Exploring Behavioral Intention
Hannes Schimmele, Stephan Schl
¨
ogl and Aleksander Groth
Interaction Lab, Department Management, Communication & IT,
MCI Management Center Innsbruck, Austria
Keywords:
Mobile Instant Messaging, Customer Service Interaction, Theory of Planned Behavior, Technology Readiness
Index.
Abstract:
Mobile Instant Messaging (IM) has become a main tool for communicating with friends and family. Recent
efforts are now exploring IM use in business-to-customer communication. To this end, particularly customer
service settings are given a high priority. A clear understanding of the relevant customer perspective is, how-
ever, still missing. Consequently, this paper reports on a study evaluating people’s behavioral intention of
adopting mobile IM for customer service interaction. We used an integrated research model based on Ajzen’s
Theory of Planned Behavior (TPB) and Parasuraman’s Technology Readiness Index. A total of 154 ques-
tionnaires were analyzed. Results indicate only small effects of technology-related personality traits on the
cognitive dimensions of the TPB. Concerning TPB-internal relations, Attitude and Subjective Norm were found
to have significant influences on Behavioral Intention. In summary, however, the study indicates that people’s
intention to adopt mobile IM for customer service interaction is only slightly positive.
1 INTRODUCTION
On February 1
st
2016 WhatsApp announced their ef-
forts in exploring alternative application areas for mo-
bile Instant Messaging (IM)
1
. In particular, they high-
lighted their goal to offer users a new way of commu-
nicating with businesses and organizations. In other
words, individuals and businesses should be able to
exchange important information in a modern, more
convenient and less complicated manner. WhatsApps
intentions may be seen as an attempt to open up a new
chapter in the field of customer-business interaction.
Since then, first prototypical implementations of us-
ing IM for customer communication have been put in
place. Yet, so far, little is know about the customers’
perspective and consequent attitude towards these ser-
vices. The goal of the work presented in this paper is
thus to shed some light on the potential use of IM in
customer service settings. Employing a model-based
research approach we aim to contribute to a better un-
derstanding of the users’ underlying motives and their
possible reluctance factors, which play an essential
role in the adoption and/or refusal of said technology
for Customer Service Interaction (CSI).
1
https://blog.whatsapp.com/ [retrieved: May 15 2016]
2 THEORETICAL BACKGROUND
With the ongoing propagation of smartphones the
availability of mobile IM is constantly increasing. In
contrast to the use of traditional SMS-based text mes-
saging, the use of mobile IM is not limited to a given
number of characters and is usually free of charge
provided the user has access to an internet con-
nection via data plan or a wireless network connec-
tion (Church and de Oliveira, 2013). Furthermore,
IM includes a number of positive characteristics, like
different types of media content such as for exam-
ple videos, photos, documents, contacts, audio files,
the user’s current location as well as his/her voice
mails (Rennecker and Godwin, 2003). In addition,
most mobile IM tools feature information about a
communication partner’s status, such as him/her be-
ing online/offline, busy, typing, idle or free, which
mimics important characteristics of a real-time physi-
cal person-to-person conversation (Deng et al., 2010).
Moreover, mobile IM applications provide numer-
ous emoticons to convey emotions (Lancaster et al.,
2007). Thus, given the variety of different transfer-
able media types combined with the ‘emotional’ ef-
fect of emoticons, studies have already shown that
users perceive (mobile) IM as more personalized than
other means of digital communication such as for ex-
Schimmele, H., Schlögl, S. and Groth, A.
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention.
DOI: 10.5220/0006490100390049
In Proceedings of the International Conference on Computer-Human Interaction Research and Applications (CHIRA 2017), pages 39-49
ISBN: 978-989-758-267-7
Copyright © 2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
39
ample email (Church and de Oliveira, 2013; Huang
et al., 2008).
However, perceived (technological) advantages
alone do usually not suffice to ensure the adoption
of technology in a given application area. Motiva-
tional, behavioral as well as social factors also play
their part. To this end, technology acceptance and/or
adoption, being “conceptualized as an outcome vari-
able in a psychological process that users go through
in making decisions about technology” (Dillon and
Morris, 1996), is perceived as a crucial determinant
influencing the success or failure of a technology
brought to market. Consequently, its underlying the-
oretical models have been subject to many previous
studies (Ghyas et al., 2012; Venkatesh et al., 2003).
2.1 Theory of Reasoned Action
The Theory of Reasoned Action (TRA) developed by
Fishbein & Ajzen (Fishbein and Ajzen, 1975) was
one of the first widely applied theoretical models to
explain human behavior. Until today, it is one of
the most applied theories and has received substantial
empirical support (Dabholkar, 1994; P
¨
uschel et al.,
2010). The model is based on the assumption that a
person’s actual behavior is influenced by his/her in-
tention to execute said behavior. Intention to perform
behavior X can thus be characterized as an indicator
“of how hard people are willing to try, of how much
of an effort they are planning to exert, in order to per-
form the behavior” (Ajzen, 1991). Consequently, a
strong behavioral intention should also increase the
likelihood of actually performing the respective be-
havior (Ajzen, 1985; Ajzen, 1991). Intention to per-
form behavior X is further mutually determined by
the individual’s Attitude towards said behavior and
his/her Subjective Norm concerning it (Fishbein and
Ajzen, 1975). These two dimensions reflect on the
one hand the personal impact and on the other hand
the social influences on the intention to perform a be-
havior (Ajzen, 1985). That is, Attitude towards be-
havior X “refers to the degree to which a person has a
favorable or unfavorable evaluation or appraisal of the
behavior in question” (Ajzen, 1991). Subjective Norm
on the other hand is defined as “the person’s per-
ception of the social pressure put on him/her to per-
form or not perform the behavior in question” (Ajzen,
1985). At a deeper level, one may thus argue that
people’s salient beliefs determine their Attitude and
Subjective Norm concerning behavior X. In summary
the TRA therefore proposes that, individuals intend
to carry out a behavior if they, for themselves, pos-
itively evaluate the behavior in question and if they
think that people of their social environment would
support them performing this behavior (Ajzen, 1985;
Ajzen, 1991).
2.2 Technology Acceptance Model
While TRA was designed to explain human behav-
ior and its antecedents in general, it also served as
the foundation for other models aiming to explain be-
havior in more specific contexts. In case of technol-
ogy, such particularly concerns the Technology Ac-
ceptance Model (TAM) developed by Davis (Davis,
1989). TAM focuses on explaining user acceptance
of technology in the workplace by replacing the TRA
element Subjective Norm with the dimensions Per-
ceived Usefulness and Perceived Ease of Use (Davis
et al., 1989; Davis, 1989). Perceived Usefulness is
defined as “the degree to which a person believes
that using a particular system would enhance his/her
job performance” (Davis, 1989). Here the term use-
ful is based on the assumption that a technology
can be used advantageously (Davis, 1989). Con-
sequently, a system which is characterized by high
Perceived Usefulness “is one for which a user be-
lieves in the existence of a positive use-performance
relationship” (Davis, 1989). As such it is directly
connected to both Attitude and Behavioral Intention
and grounded in the argument that, in a work set-
ting, the intention to use a system may also stem
directly from its assumed impact on the overall job
performance, independent of the overall attitude to-
wards using the system (Davis, 1989). Perceived Ease
of Use, on the other hand, describes “the degree to
which a person believes that using a particular system
would be free of effort” (Davis, 1989), where the term
ease implies the “freedom from difficulty or great ef-
fort” (Davis, 1989). In contrast to TRA, TAM has
been widely applied in studies on mobile services us-
age; e.g. (Basg
¨
oze, 2015; Glass and Li, 2010; Godoe
and Johansen, 2012; Gombachika and Khangamwa,
2013; Guhr et al., 2013; Jin, 2014; Li et al., 2005; Lu
et al., 2009).
2.3 Theory of Planned Behavior
Next to TAM also the Theory of Planned Behavior
(TPB) strongly builds on the basis of the TRA. Instead
of replacing the component Subjective Norm, how-
ever, it extends the model by a third dimension. Origi-
nally TRA was created to explain volitional behavior,
i.e. behavior where an individual voluntarily chooses
to perform an action (Ajzen, 1985). This has been a
major limitation to the TRA model, as it is not appli-
cable to behavior which is not under full perceived or
actual control (Ajzen, 1985; Ajzen, 1991). To over-
CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications
40
come this limitation, the TPB thus incorporates the
dimension Perceived Behavioral Control, which is de-
fined as the “perceived ease or difficulty of perform-
ing the behavior and it is assumed to reflect past ex-
perience as well as anticipated impediments and ob-
stacles” (Ajzen, 1991). Except for the amendment of
Perceived Behavioral Control and its direct impact on
the actual behavior, the TPB construct is, however,
identical to the TRA. In other words, the TPB essen-
tially assumes “the more favorable the Attitude and
Subjective Norm with respect to a behavior, and the
greater the Perceived Behavioral Control, the stronger
should be an individual’s intention to perform the be-
havior under consideration” (Ajzen, 1991).
3 ACCEPTANCE OF MOBILE
INSTANT MESSAGING
Lu et al. (Lu et al., 2009) as well as Glass & Li (Glass
and Li, 2010) aimed at generally explaining mo-
bile IM acceptance with an integrated TAM. Shih &
Fan (Shih and Fan, 2013), on the other hand, explored
IM in a more work-related context. In their case, the
TPB was extended by a construct measuring technol-
ogy readiness. The results showed that an optimistic
view of technology also fosters a positive attitude to-
wards applying IM for a professional purpose (Shih
and Fan, 2013). Building on these results, both inter-
net and mobile banking have been explored through
applying the TRA (Nor, 2015) as well as a Decom-
posed version of the Theory of Planned Behavior
(DTPB) (P
¨
uschel et al., 2010; Shih and Fang, 2004).
Whereas the DTPB studies applied the original model
almost in its pure form, Nor (Nor, 2015) integrated a
technology readiness construct into the TRA. Similar
study settings comparing the TRA and the DTPB are
found in the educational sector. For example, Ghyas
et al. (Ghyas et al., 2012) evaluated the acceptance
of e-readers using DTPT whereas Jin (Jin, 2014) in-
vestigated the technology through an integrated and
adapted TAM. E-learning, in particular, was subject
to TAM studies with technology readiness dimension
to them (Yen and Chen, 2010).
Another relevant comparison may be found in
acceptance studies focusing on mobile commerce.
Basg
¨
oze, for example investigated mobile shop-
ping using TAM and a technology readiness exten-
sion (Basg
¨
oze, 2015). Similar approaches were taken
to explore the acceptance of mobile payment (Guhr
et al., 2013; Shin and Lee, 2014). On a more general
level, the influence of gender for the acceptance of in-
formation and communication technologies was eval-
uated by Gombachika and colleagues (Gombachika
and Khangamwa, 2013). The study strongly focused
on technology readiness and again measured ICT ac-
ceptance through different TAM constructs. Finally,
we have also seen the acceptance of e-service and
self-service technologies being studied by integrating
technology readiness into the TPB (Chen and Chen,
2008; Chen and Li, 2010).
In summary, however, we see two common char-
acteristics inherent with previous studies. First, very
few studies employ the original models such as a pure
version of TRA, TPB or TAM. Instead, researchers
keep integrating other behavioral theories or seem-
ingly foreign models so as to potentially improve their
research results. Second, the extension of original
models often concerns an additional construct that
aims at measuring people’s technology readiness - an
approach which is further discussed by the following
section.
3.1 Technology Readiness
With technology increasingly becoming the main re-
quirement for services and products being transported
(not to say offered) to customers, technology readi-
ness has evolved as an important facet of behavioral
research. It allows to draw conclusions about people’s
emotional reaction to newly introduced technologies.
Mick & Fournier (Mick and Fournier, 1998) found
eight paradoxes which users are faced with when be-
ing confronted with a new technology. This shows
certain similarities to the underlying assumptions put
forward by the TPB. Dabholkar (Dabholkar, 1996),
for example, highlights that people have different
feelings about technology which influence their indi-
vidual behavioral intention to use it. Parasuraman, on
the other hand, proposes that “a combination of posi-
tive and negative feelings about technology underlies
the domain of technology readiness” (Parasuraman,
2000). Furthermore, he states that “whereas posi-
tive feelings propel people towards new technologies,
the negative feelings may hold them back” (Parasur-
aman, 2000). Building on this theory Parasuraman
developed the Technology Readiness Index (TRI) as
a construct which “refers to people’s propensity to
embrace and use new technologies for accomplish-
ing goals in home life and at work” (Parasuraman,
2000). The TRI theory in its version from 2000 in-
corporates 36 items which can be assigned to the
four categories Optimism, Innovativeness, Discom-
fort and Insecurity. They reflect the positive (Op-
timism and Innovativeness) and negative (Discom-
fort and Insecurity) perceptions of technology accord-
ing to Mick & Fournier (Mick and Fournier, 1998).
Accordingly, they can be interpreted as an individ-
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention
41
ual’s technology-related personality traits, which in-
fluence his/her technology readiness. In other words,
Optimism and Innovativeness are defined as drivers
of technology readiness whilst Discomfort and In-
security inhibit technology readiness (Parasuraman,
2000).
In 2014, Parasuraman & Colby updated the ini-
tial TRI version by introducing TRI 2.0 (Parasura-
man and Colby, 2014). Like the original model (TRI
1.0), TRI 2.0 consists of the four personality traits
Optimism, Innovativeness, Discomfort and Insecurity,
with the first two being motivators and the latter two
being inhibitors of technology readiness. Also the
terms and definitions of TRI 1.0 were transferred to
TRI 2.0. However, the items were re-worded, so as
to be less specific to one particular technology, and
condensed (Parasuraman and Colby, 2014). The fi-
nal TRI 2.0, which should be applicable to a wider
field of technology explorations, now consists of four
questions per trait, leading to a total of 16 items to
be put on a questionnaire. According to Parasuraman
& Colby (Parasuraman and Colby, 2014), the newly
integrated TRI 2.0 may be particularity used to in-
vestigate a “moderating variable in studies involving
multivariate frameworks”. Following we conclude
the necessary background analysis by discussing sev-
eral approaches that have integrated this type of tech-
nology readiness into technology acceptance studies,
leading over to the research agenda perused by our
work.
3.2 Integrating TRI with TRA, TAM
and TPB
Ever since Ajzen proposed that besides Attitude also
other personality traits have a major influence on
human behavior (Ajzen, 1991; Ajzen, 2005), re-
searchers have worked on integrating technology
readiness into behavioral models. As for an integra-
tion with technology acceptance, particularly TAM,
we have seen a number of application scenarios;
e.g. (Chen and Chen, 2008; Chen and Li, 2010; Shih
and Fan, 2013). Hence, while for TAM fundamen-
tal work such as the Technology Readiness and Ac-
ceptance Model (TRAM) by Lin et al. (Lin et al.,
2007) exists, the integration of TRI with TPB is less
researched. Chen & Chen (Chen and Chen, 2008)
were among the first to examine the effect of technol-
ogy readiness on consumer behavior and may conse-
quently be regarded as the pioneers in integrating TRI
(in version 1.0) with TPB. Doing so, they examined
the influence of personality traits on the antecedents
of Behavioral Intention, finding that personality traits
do indeed play an important role in the information
technology adoption (Chen and Chen, 2008). Another
approach to integrate TRI with TPB was undertaken
when examining the adoption of e-services (Chen and
Li, 2010). It was theorized that technology readiness
can be interpreted as a belief and would therefore pre-
cede Attitude, Subjective Norm and Perceived Behav-
ioral Control, as explained by the TPB (Ajzen, 1985;
Ajzen, 1991; Chen and Li, 2010). Finally, a third
study, set in the context of IM adoption by Taiwanese
travel agency workers, concentrated on the influence
technology-related personality traits have on the TPB
element Attitude alone (Shih and Fan, 2013).
In summary, we have seen that in the past primar-
ily two conceptions integrating technology readiness
with acceptance have emerged. On the one hand, re-
searchers have studied how technology readiness as a
whole influences the antecedents of TAM and TPB.
On the other hand, studies were aimed at a diverse set
of personality traits and how they influence technol-
ogy adoption, particularly focusing on relating single
TRI components to TAM/TPB elements; e.g. (Chen
and Chen, 2008; Godoe and Johansen, 2012; Wal-
czuch et al., 2007). Our study aims to add to this
specific research domain by investigating the accept-
ability/adoption of mobile IM in customer service set-
tings. The applied research model, which builds on
the work of Chen & Chen (Chen and Chen, 2008),
but replaces TRI 1.0 by its successor TRI 2.0, is dis-
cussed below.
4 RESEARCH MODEL AND
HYPOTHESIS
The discussion above highlights that TRA, TPB as
well as TAM are viable constructs capable of evalu-
ating technology adoption/acceptance. Also, all three
have been integrated with the TRI 1.0 measurement
scale. However, with respect to the exploration of be-
havioral intention, TPB seems to be a better fit than
TRA and TAM (and their respective derivatives). One
reason for this is because TPB provides “the possibil-
ity of making further distinctions among additional
kinds of beliefs and related dispositions” (Ajzen,
1991). In addition, personality traits do play an im-
portant role in human behavior (Ajzen, 1985; Ajzen,
2005). Most previous studies using a TRI-TPB in-
tegration are, however, grounded in the initial model
proposed by Chen & Chen (Chen and Chen, 2008).
The main reason why this initial model is preferred
over other TRI-TPB integrated approaches, such as
the one by Chen & Li (Chen and Li, 2010), lies in
its focus on each single technology readiness dimen-
sion which does not aggregate all of the items into
CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications
42
one measure. Empirical results confirmed the valid-
ity of this approach as it was found that not all TRI
components have an equal influence on the TPB el-
ements Attitude, Subjective Norm and Perceived Be-
havioral Control (Chen and Chen, 2008). Shih &
Fan’s results (Shih and Fan, 2013) support this ar-
gumentation as well, and Godoe & Johansen (Godoe
and Johansen, 2012), who employed a TRI-TAM con-
struct also recommend the decomposition of technol-
ogy readiness into its underlying technology-related
personality traits instead of relating them to the TPB
elements as one single measure.
The presented study aims to add to the body
of knowledge in this field by integrating TPB with
the updated version of TRI, i.e. TRI 2.0. As per
the TPB framework, behavior is dependent on In-
tention, and Intention in turn is explained by its an-
tecedents Attitude towards the behavior, Subjective
Norm and Perceived Behavioral Control, we may the-
orize that technology-related personality traits can di-
rectly be related to the TPB elements. Here Opti-
mism and Innovativeness are characterized as drivers,
and Insecurity and Discomfort as inhibitors of tech-
nology readiness (Parasuraman and Colby, 2014;
Parasuraman, 2000). People who hold an optimistic
and innovative position towards technology are there-
fore expected to also have a positive attitude towards
the adoption of mobile IM for Customer Service In-
teraction (CSI). In contrast, people who feel insecure
and dis-comfortable with respect to technology are
predicted to have a rather negative attitude towards
the adoption of mobile IM for CSI. Consequently one
may hypothesize that:
H1(a): Optimism is positively related to Attitude.
H1(b): Innovativeness is positively related to Atti-
tude.
H1(c): Discomfort is negatively related to Atti-
tude.
H1(d): Insecurity is negatively related to Attitude.
It is further expected that optimistic and innova-
tive character traits make people expect support from
their social environment concerning the adoption of
mobile IM for CSI. Yet, feelings of insecurity and dis-
comfort towards technology make people think that
their social environment would rather constrain such
ambitions. Therefore one may further hypothesize
that:
H2(a): Optimism is positively related to Subjective
Norm.
H2(b): Innovativeness is positively related to Sub-
jective Norm.
H2(c): Discomfort is negatively related to Subjec-
tive Norm.
H2(d): Insecurity is negatively related to Subjec-
tive Norm.
Similar arguments should account for peo-
ple’s Perceived Behavioral Control. That is, people
who have an optimistic and innovative view of tech-
nology are expected to be confident in operating mo-
bile IM for CSI. In contrast, feelings of insecurity
and discomfort towards technology should result in
a low confidence concerning these operations. Con-
sequently one may argue that:
H3(a): Optimism is positively related to Perceived
Behavioral Control.
H3(b): Innovativeness is positively related to Per-
ceived Behavioral Control.
H3(c): Discomfort is negatively related to Per-
ceived Behavioral Control.
H3(d): Insecurity is negatively related to Per-
ceived Behavioral Control.
Finally, according to the original TPB model, Atti-
tude, Subjective Norm and Perceived Behavioral Con-
trol influence Behavioral Intention (Ajzen, 1985;
Ajzen, 1991). Therefore, one may conclude that a
positive attitude towards mobile IM for CSI results
in a stronger behavioral intention to actually employ
it. Accordingly, it is also expected that people’s be-
liefs in a supporting social environment would trigger
a stronger behavioral intention. The same accounts
for people’s confidence in their ability to use mobile
IM for CSI, wherefore it may be hypothesized that:
H4(a): Attitude is positively related to Behavioral
Intention.
H4(b): Subjective Norm is positively related
to Behavioral Intention.
H4(c): Perceived Behavioral Control is positively
related to Behavioral Intention.
4.1 Questionnaire Design
Our study design tackling the above hypotheses is
strongly model-based, building on prior research con-
ducted in technology adoption of e-services. The
work by Chen & Chen (Chen and Chen, 2008) served
as the main guideline, however, we focused specifi-
cally on the use of mobile IM for CSI and further re-
placed the original TRI model with its successor TRI
2.0.
Following the example of previous surveys (Chen
and Chen, 2008; Dabholkar, 1996), we designed a
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention
43
Table 1: Questionnaire survey completed by 154 respondents.
Name Item
ATT1 Using WhatsApp for customer service interaction is a good idea.
ATT2 I like the idea of using WhatsApp for customer service interaction.
ATT3 Using WhatsApp for customer service interaction is a wise idea.
ATT4 Using WhatsApp for customer service interaction would be pleasant.
SN1 People who are important to me (e.g. friends, family) would think that I should use WhatsApp
for customer service interaction.
SN2 People who are important to me (e.g. friends, family) would think that using WhatsApp for
customer service interaction is a good idea.
SN3 People who influence me (e.g. bloggers, role models) would think that I should use WhatsApp
for customer service interaction.
SN4 People who influence me (e.g. bloggers, role models) would think that using WhatsApp for
customer service interaction is a good idea.
PBC1 I would be able to use WhatsApp for customer service interaction.
PBC2 I have the resources (e.g. Smartphone with WhatsApp installed, mobile data plan) to use What-
sApp.
PBC3 I have the knowledge to operate WhatsApp for customer service interaction.
PBC4 I have the ability how to operate WhatsApp.
BI1 I would plan to use WhatsApp for customer service interaction.
BI2 I would intend to use WhatsApp for customer service interaction within the next three months.
BI3 I would recommend WhatsApp for customer service interaction to others.
OPT1 New technologies contribute to a better quality of life.
OPT2 Technology gives me more freedom of mobility.
OPT3 Technology gives people more control over their daily lives.
OPT4 Technology makes me more productive in my personal life.
INN1 Other people come to me for advice on new technologies.
INN2 In general, I am among the first in my circle of friends to acquire new technology when it
appears.
INN3 I can usually figure out new high-tech products and services without help from others.
INN4 I keep up with the latest technological developments in my areas of interest.
DIS1 When I get technical support from a provider of a high-tech product or service, I sometimes feel
as if I am being taken advantage of by someone who knows more than I do.
DIS2 Technical support lines are not helpful because they don’t explain things in terms I understand.
DIS3 Sometimes, I think that technology systems are not designed for use by ordinary people.
DIS4 There is no such thing as a manual for a high-tech product or service that’s written in plain
language.
INS1 People are too dependent on technology to do things for them.
INS2 Too much technology distracts people to a point that is harmful.
INS3 Technology lowers the quality of relationships by reducing personal interaction.
INS4 I do not feel confident doing business with a place that can only be reached online.
GENDER What’s your gender?
AGE How old are you (in years)?
COUNTRY My home country is:
WAEXP I have used WhatsApp for:
SEXP I have owned a smartphone (or several smartphones) for:
WADAILY The time I spend every day in WhatsApp is:
CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications
44
questionnaire study which included a short introduc-
tion to the purpose of the research, a potential usage
scenario which let respondents imagine a concrete ap-
plication of mobile IM for CSI, and the actual ques-
tion constructs, covering TPB, TRI 2.0, and demo-
graphic items.
With respect to the TPB, items developed by Tay-
lor & Todd (Taylor and Todd, 1995) and validated
by several follow-up studies (Ghyas et al., 2012; Lu
et al., 2009; P
¨
uschel et al., 2010; Shih and Fang,
2004; Teo and Pok, 2003) were adjusted to the con-
text of mobile IM for CSI. The term mobile IM was
thereby substituted by the (today) more familiar term
WhatsApp. Consequently, the first part of the actual
questionnaire focusing on TPB included four Attitude
items developed according to Taylor & Todd (Taylor
and Todd, 1995). Additional four Subjective Norm
items were taken from work by Shih & Fang (Shih
and Fang, 2004) who also provide the four items
for Perceived Behavioral Control as well as two items
for Behavioral Intention. The third (and final) Behav-
ioral Intention item, which particularly incorporated
the recommendations of WhatsApp for CSI to others,
was inspired by Lu and colleagues (Lu et al., 2009).
The second part of the questionnaire incorporated
the 16 TRI 2.0 measurement items (Parasuraman and
Colby, 2014). To this end no adaptation was re-
quired as TRI measures are technology and context
independent. All variables used a five-point Likert
scale differential ranging from “strongly disagree” to
“strongly agree”, complying with recommendations
for both TPB as well as TRI 2.0 (Carifio and Perla,
2007; Parasuraman and Colby, 2014). The final part
of the questionnaire comprised six demographic vari-
ables including gender, age, WhatsApp experience,
smartphone experience and daily WhatsApp usage,
as well as one question concerning the participant’s
country of residence. The complete survey (cf. Ta-
ble 1) was distributed online using a convenience
sample of WhatsApp users in Central Europe, particu-
larly focusing on Germany, Austria and Switzerland,
and led to 154 completed responses within 29 days.
4.2 Questionnaire Reliability and
Validity
The returned data shows an equal gender distribution
(i.e. 72 male and 78 female participants) with the
respondents’ age ranging from 18 to 62 years (note:
66.7 percent were between 21 and 26 years old). The
majority (87.3%) has owned a smartphone for at least
three years. As for the WhatsApp usage, 37.3% re-
ported to use the app between 31 and 60 minutes
per day, while 38.7% would use it shorter and 24.0%
longer. Calculating Cronbach’s a the questionnaire
items investigating Attitude (ATT), Subjective Norm
(SN), Perceived Behavioral Control (PBC), Behav-
ioral Intention (BI) as well as Innovativeness (INN)
had coefficients greater than 0.7, showing high relia-
bility. The reliability of the items investigating Opti-
mism (OPT), Discomfort (DIS) and Insecurity (INS)
had coefficients ranging from 0.6 to 0.7, deeming
them slightly less reliable.
Next, similar to previous studies (Gombachika
and Khangamwa, 2013; Shih and Fan, 2013) we
used an Exploratory Factor Analysis (EFA) to en-
sure that all the TRI 2.0 and TPB items would load
to their associated construct. This process started
with a variable selection and then continued with
the consequent factor extraction. Here, the Kaiser-
Meyer-Olkin (KMO) criterion measures whether a set
of variables qualifies for an EFA (Kaiser and Rice,
1974). It is based on the anti-image covariance ma-
trix and lies between 0 and 1, where a value lower
than 0.5 indicates that a given variable set is not suit-
able. Applying the KMO criterion to the model-based
variable constructs used in the above survey (i.e. the
TRI 2.0 and the TPB questionnaire items) produced a
value of 0.848, classifying our variable set as “merito-
riously” suitable for an EFA (Kaiser and Rice, 1974).
The Measure of Sampling Adequacy (MSA) follows
an approach similar to the one pursued by the KMO
criterion but serves as an instrument for eliminating
those variables which do not qualify for the EFA.
Based on this analysis we found that he majority, i.e.
22 of our 31 model-based questionnaire items, qualify
as “meritoriously” suitable (MSA>0.8). The remain-
ing 9 values, i.e. SN1, SN2, PBC2, PBC3, PBC4,
DIS1, DIS2, DIS3 and DIS4, qualify as “mediocre”
(MSA>0.6), but still did not have to be excluded for
the EFA.
The consequent principle component analysis
pointed to an eight-factor model with each factor hav-
ing an Eigenvalue >1. As this model is able to explain
66.402% of the total variance implied by the rele-
vant data, the validity of our underlying research con-
structs composed of four TPB components and four
TRI 2.0 components, could be confirmed.
5 MULTIVARIATE STATISTICS
Previous work proposes to define the mean value of 3
as a cut-off point for measurements based on the Lik-
ert response format (Gombachika and Khangamwa,
2013). The argument derives from 3 being inter-
preted as “undecided” or “neutral”. Participants’ re-
sponses to TPB questionnaire items all exceeded this
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention
45
middle point, indicating an overall positive attitude
(ATT mean=3.46) and favorable influences from their
social environment (SN mean=3.12) towards using
mobile IM (i.e WhatsApp) for CSI. A strong agree-
ment with the Perceived Behavioral Control (PBC
mean=4.43) further shows that people belief that the
use of WhatsApp for CSI would be easy and that they
feel confident applying it. Consequently, it is not
surprising that the Behavioral Intention to use What-
sApp for CSI also surpassed an average rating (BI
mean=3.20). Finally, concerning technology-related
personality traits, the data shows that respondents
were rather Optimistic (OPT mean=3.74) as well as
Innovative (INN mean=3.43) but felt Discomfortable
(DIS mean=2.91) and Insecure (INS mean=3.35).
In order to further investigate the connection be-
tween TRI 2.0 and TPB elements, and consequently
tackle the hypotheses put forward earlier, a Multiple
Linear Regression (MRA) analysis was performed. A
step-wise method was applied, as it supports the ex-
clusion of those independent variables, which do not
meet a predefined level of significance (in our case
p<0.05). Consequently, only variables which make
a significant contribution are included into the final
model. According to our hypotheses, we tested the
influence of technology-related personality traits on
ATT, SN and PBC. Finally, one additional analysis
was necessary so as to evaluate (and potentially con-
firm) the relations within the original TPB construct.
The results of the MRA show no significant influ-
ence of INN, DIS and INS on ATT (p>0.050), lead-
ing to the rejection of hypotheses H1(b), H1(c) as
well as H1(d). OPT, however, had an influence on
ATT (p=0.000), although it was only able to explain
7.9% of its variance (R
2
=0.079). A positive connec-
tion was confirmed by the respective regression coef-
ficient (B=0.369; β=0.281). Consequently, H1(a) was
not rejected.
As for the relation between the TRI 2.0 compo-
nents and SN the data showed, similar to the previous
analysis concerning ATT, no significance with respect
to INN, DIS and INS (p>0.050). The connection be-
tween OPT and SN was, however, again significant
(p=0.002; R
2
=0.066) producing a regression coeffi-
cient that confirms the positive influence of OPT on
SN and thus supports H2(a) (B=0.240; β=0.256).
Evaluating the construct with PBC as the depen-
dent variable - i.e. focusing on hypotheses H3(a),
H3(b), H3(c), and H3(d) - identified INN as the only
influencing construct (p=0.000). OPT, INS and DIS
did not show any effects for which H3(a), H3(c) as
well as H3(d) had to be rejected. With INN we saw
that it was able to explain 8.2% of the variance ex-
hibited by PBC (R
2
=0.082). The resulting regression
coefficient led to the assumption that INN positively
influences PBC and thus supports H3(b) (B=0.201;
β=0.286).
The final analysis evaluated the relations within
the original TPB construct. Both ATT and SN were
found to have an impact on BI, whereas PBC was
excluded from the construct. With two independent
variables being significant (p<0.05) we were able to
formulate two different models. Model 1, integrat-
ing only ATT as an independent variable, was able to
explain 76.7% of its variance (R
2
=0.767; p=0.000).
Adding SN as a second independent variable we were
able to push the rate to 78.1% (R
2
=0.781). This
second model also showed positive regression coeffi-
cients for ATT (B=0.868; β=0.793; p=0.000) as well
as SN (B=0.220; β=0.144; p=0.003) consequently
supporting H4(a) and H4(b). A much higher regres-
sion coefficient of ATT, however, shows that its in-
fluence is much stronger than the influence of SN.
Figure 1 depicts the resulting research model with its
standardized path coefficients and variances of depen-
dent variables.
In summary we may argue that our data did not
support the existence of a significant relations be-
tween INN and ATT, DIS and ATT or INS and ATT.
Neither did we find a connection between INN and
SN, DIS and SN, and INS and SN, or an effect of
OPT, INS or DIS on PBC. Effects between OPT and
ATT, ATT and BI, SN and BI, as well as between OPT
and SN and between INN and PBC were, however,
significant.
6 INTERPRETATION OF
RESULTS
Looking at the above results from the perspective of
Attitude being the dependent variable, only an op-
timistic view of technology leads to a positive atti-
tude towards using mobile IM (i.e. WhatsApp) for
CSI. This complies with the findings of Chen &
Chen (Chen and Chen, 2008) and Shih & Fan (Shih
and Fan, 2013), who also found support for a positive
influence of Optimism on Attitude. Yet, while Chen &
Chen in addition found support for an INN-ATT rela-
tion, our study was unable to show that people’s Inno-
vativeness would lead to them having a more positive
Attitude towards using mobile IM for CSI. Similarly,
feelings of Discomfort and Insecurity did not show
any influence on our participants Attitude.
Concerning the influencers of Subjective Norm
we found a significant connection with Optimism,
whereas a connection to Innovativeness highlighted
by Chen and colleagues (Chen and Chen, 2008) was
CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications
46
Figure 1: Results of the computed research model.
not inherent. However, we found that being open to
new technologies makes people feel that their social
environment would support the use of mobile IM for
CSI. Again feelings of Insecurity and Discomfort had
no significant impact on people’s Subjective Norm.
Innovativeness was the only significant influence on
Perceived Behavioral Control. One may thus argue
that, people considering themselves as technology pi-
oneers makes them feel comfortable and easily able to
perform mobile IM for CSI. Surprisingly, however, an
optimistic view of technology was not found to signif-
icantly contribute to this Perceived Behavioral Con-
trol, which contradicts Chen & Chen’s results, who
also identify INN to be the most important influencer
of PBC but further highlighted a significant relation-
ship between OPT and PBC (Chen and Chen, 2008).
With respect to the actual TPB model, our study
found that a positive attitude towards the technology
is the strongest driver for people’s Behavioral Inten-
tion to use IM for CSI. Furthermore, we found that
perceived social pressure significantly influences peo-
ple’s BI, although with a much lower impact than their
positive attitude. In contrast, people’s confidence in
their ability to perform does not impact on their BI.
These results confirm findings from previous studies
which show that “intentions are often formed without
subjective norms playing a major role” (Dabholkar,
1994). The lacking influence of PBC on BI is also
supported by the findings of Teo & Pok (Teo and Pok,
2003) who furthermore found significant effects for
the relations ATT-BI as well as SN-BI.
In summary we may thus argue that our survey
results are generally in line with previous work, al-
though from a mobile IM perspective we expected
certain deviations. Social pressure, for example, was
expected to have less impact, as the actual communi-
cation would occur only between a customer and the
company, and does not include any peers. In contrast,
the perceived ability to use mobile IM was expected to
be the main requirement for BI, yet did not show any
significant effect. Consequently we may also state
that our study, which essentially explored people’s IM
use for CSI and a respective connection between the
TRI 2.0 and the TPB, confirms previous results re-
ported by Chen & Chen (Chen and Chen, 2008). In
doing so, it is particularly the TPB part which shows
high power in explaining people’s Behavioral Inten-
tion to use IM for CSI, accounting for 78.1% of the re-
spondent’s variance. To this end, previous studies fo-
cusing on other technologies were often significantly
lower, e.g. Teo & Pok (Teo and Pok, 2003) were able
to explain 17.2%, Lu and colleagues (Lu et al., 2009)
reported 61% and P
¨
uschel et al. (P
¨
uschel et al., 2010)
68.8%.
7 CONCLUSION AND FUTURE
WORK
The presented work explores the acceptance of mo-
bile IM (i.e. WhatsApp) for CSI settings, using a
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention
47
model-based questionnaire survey that integrates pre-
vious work on human behavior and technology readi-
ness. Its results may be summarized by three core
findings: (1) although we were not able to deduce a
reliable statement as to whether people would actually
like to use mobile IM for CSI, our data shows a slight
tendency towards a positive adoption; (2) the drivers
that may potentially lead to the adoption of mobile IM
for CSI are mostly found in people’s attitude towards
the technology (Note: while attitude has the strongest
influence, also people’s subjective norms play a sig-
nificant role. One may therefore argue that friends or
family members as well as other influencers such as
role models or tech bloggers have a positive impact
on people’s intention to adopt mobile IM for CSI);
(3) technology-related personality traits did not influ-
ence technology adoption, i.e. the personality traits
measured by the TRI 2.0 added little to explain the
adoption of IM for CSI.
Future studies will aim at deepening our un-
derstanding in this application area. In particular
we want to focus on other models integrating tech-
nology readiness and technology-related personality
traits. Using the model by Chen & Li (Chen and Li,
2010) may, for example, show that not the underly-
ing technology-related personality traits but technol-
ogy readiness as a whole influences the adoption of
mobile IM. Furthermore, the direct comparison of an
integrated TRI-TPB model with an integrated TRI-
TAM model might deliver valuable insights. Our
study was not able to show that technology-related
personality traits have a meaningful influence on peo-
ple’s behavioral intention to adopt mobile IM for CSI.
Such may imply that for this specific application area
technology-related personality traits are of little im-
portance. On the other hand, this could also mean
that the research model needs to be re-calibrated.
Also, an exploratory approach based on qualita-
tive research methods may be an interesting path for
further investigation, as it may add to a better discov-
ery of underlying reasons (Hyde, 2000). In particu-
lar privacy aspects could be of high relevance here.
Finally, the emerging market of mobile IM for busi-
nesses shows that artificial intelligence enabled mes-
saging bots receive a lot of attention in the tech-
community. Next to convenience tasks, also cus-
tomer service belongs to the targeted application ar-
eas, wherefore a more general exploration of messag-
ing adoption and its influencers may also be an inter-
esting future research direction.
REFERENCES
Ajzen, I. (1985). From intentions to action: A theory of
planned behavior.
Ajzen, I. (1991). The theory of planned behavior. Orgniza-
tional Behavior and Human Decision Processes,
50:179–211.
Ajzen, I. (2005). Attitudes, Personality and Behaviour.
In Manstead, T., editor, Mapping Social Psychology,
pages 1–178. McGraw-Hill Education, New York, 2
edition.
Basg
¨
oze, P. (2015). Integration of Technology Readiness
(TR) Into the Technology Acceptance Model (TAM)
for M-Shopping. International Journal of Scientific
Research and Innovative Technology, 2(3):26–35.
Carifio, J. and Perla, R. J. (2007). Ten Common Misunder-
standings, Misconceptions, Persistent Myths and Ur-
ban Legends about Likert Scales and Likert Response
Formats and their Antidotes. Journal of Social Sci-
ences, 3(3):106–116.
Chen, S.-C. and Chen, H.-H. (2008). The influence of
technology readiness on the theory of planned behav-
ior with self-service technologies. WMSCI 2008 -
The 12th World Multi-Conference on Systemics, Cy-
bernetics and Informatics, Jointly with the 14th Inter-
national Conference on Information Systems Analysis
and Synthesis, ISAS 2008 - Proc., 6(40):85–90.
Chen, S.-C. and Li, S.-H. (2010). Consumer adoption of
e-service: Integrating technology readiness with the
theory of planned behavior. African Journal of Busi-
ness Management, 4(16):3556–3563.
Church, K. and de Oliveira, R. (2013). What’s up with
whatsapp?: comparing mobile instant messaging be-
haviors with traditional sms. In Proceedings of the
15th international conference on Human-computer
interaction with mobile devices and services, pages
352–361, Munich, Germany. ACM.
Dabholkar, P. (1994). Incorporating Choice into an Attitu-
dinal Framework: Analyzing Models of Mental Com-
parison Processes. Journal of Consumer Research,
21(1):100–118.
Dabholkar, P. (1996). Consumer evaluations of new
technology-based self-service options : An investiga-
tion of alternative models of service quality. Intern. J.
of Research in Marketing, 13(1):29–51.
Davis, F. (1989). Perceived Usefulness, Perceived East of
Use, and User Acceptance of Information Technology.
MIS Quarterly, 13(3):319–340.
Davis, F., Bagozzi, R., and Warshaw, P. (1989). User
Acceptance of Computer Technology: a Comparison
of Two Theoretical Models. Management Science.
Aug1989, 35(8):982–1003.
Deng, Z., Lu, Y., Kee, K., and Zhang, J. (2010). Under-
standing customer satisfaction and loyalty: An em-
pirical study of mobile instant messages in China.
International Journal of Information Management,
30(4):289–300.
Dillon, A. and Morris, M. G. (1996). User acceptance
of new information technology: theories and models.
CHIRA 2017 - International Conference on Computer-Human Interaction Research and Applications
48
Annual Review of Information Science and Technol-
ogy Volume 31, Vol. 31:3–32.
Fishbein, M. and Ajzen, I. (1975). Belief, Attitude, Inten-
tion and Behavior: An Introduction to Theory and Re-
search. Addison-Wesley, Reading, MA, USA.
Ghyas, Q. M., Sugiura, H., and Kondo, F. N. (2012). Ex-
amining User Intention Behaviour Towards e-Readers
in Japan Using the Decomposed Theory of Planned
Behaviour. In MOBILITY 2012, The Second Interna-
tional Conference on Mobile Services, Resources, and
Users, pages 38–45, Venice, Italy. IARIA XPS Press.
Glass, R. and Li, S. (2010). Social Influence and Instant
Messaging Adoption. Journal of Computer Informa-
tion Systems, 51(2):24–30.
Godoe, P. and Johansen, T. S. (2012). Understanding adop-
tion of new technologies: Technology readiness and
technoloy acceptance as an integrated concept. Jour-
nal of European Psychology Students, 3(1):38–53.
Gombachika, H. S. and Khangamwa, G. (2013). ICT readi-
ness and acceptance among TEVT students in Univer-
sity of Malawi. Campus-Wide Information Systems,
30(1):35–43.
Guhr, N., Loi, T., Wiegard, R., and Breitner, M. H.
(2013). Technology Readiness in Customers Per-
ception and Acceptance of M(obile ) -Payment : An
Empirical Study in Finland , Germany , the USA and
Japan. International Conference on Wirtschaftsinfor-
matik, March:119–133.
Huang, A. H., Yen, D. C., and Zhang, X. (2008). Explor-
ing the potential effects of emoticons. Information &
Management, 45(7):466–473.
Hyde, K. F. (2000). Recognising deductive processes in
qualitative research. Qualitative Market Research: An
International Journal, 3(2):82–90.
Jin, C. H. (2014). Adoption of e-book among college stu-
dents: The perspective of an integrated TAM. Com-
puters in Human Behavior, 41(1):471–477.
Kaiser, H. F. and Rice, J. (1974). Little Jiffy, Mark IV. Edu-
cational and psychological measurement, 34(1):111–
117.
Lancaster, S., Yen, D. C., Huang, A. H., and Hung, S.-Y.
(2007). The selection of instant messaging or e-mail:
College students’ perspective for computer communi-
cation. Information Management & Computer Secu-
rity, 15(1):5–22.
Li, D., Chau, P. Y., and Lou, H. (2005). Understanding In-
dividual Adoption of Instant Messaging: An Emprical
Investigation. Journal of the Association for Informa-
tion Systems, 6(4):102 – 129.
Lin, C.-H., Shih, H.-Y., and Sher, P. J. (2007). Integrat-
ing Technology Readiness into Technology Accep-
tance: The TRAM Model. Psychology & Marketing,
24(7):641–657.
Lu, Y., Zhou, T., and Wang, B. (2009). Exploring Chinese
users’ acceptance of instant messaging using the the-
ory of planned behavior, the technology acceptance
model, and the flow theory. Computers in Human Be-
havior, 25(1):29–39.
Mick, D. G. and Fournier, S. (1998). Paradoxes of
Technology: Consumer Cognizance, Emotions, and
Coping Strategies. Journal of Consumer Research,
25(2):123–143.
Nor, A. S. A.-A. K. M. (2015). Challenges of adoption
of internet banking service in Yemen. International
Journal of Bank Marketing, 33(2):178–194.
Parasuraman, A. (2000). Technology Readiness Index
(TRI) - A Multiple-Item Scale to Embrace New Tech-
nologies. Journal of Service Research, 2(4):307–320.
Parasuraman, A. and Colby, C. L. (2014). An updated
and streamlined technology readiness index: TRI 2.0.
Journal of service research, 18(1):1–16.
P
¨
uschel, J., Mazzon, J. A., and Hernandez, J. M. C. (2010).
Mobile banking: proposition of an integrated adoption
intention framework. International Journal of Bank
Marketing, 28:389–409.
Rennecker, J. and Godwin, L. (2003). Theorizing the Un-
intended Consequences of Instant Messaging (IM) for
Worker Productivity. Sprouts: Working Papers on In-
formation Environments, Systems and Organizations,
3(3):137–168.
Shih, Y.-C. and Fan, S.-t. (2013). Adoption of Instant
Messaging By Travel Agency Workers in Taiwan: In-
tegrating Technology Readiness with the Theory of
Planned Behavior. International Journal of Business
and Information, 8(1):120–136.
Shih, Y.-Y. and Fang, K. (2004). The use of a decomposed
theory of planned behavior to study Internet banking
in Taiwan. Internet Research, 14(3):213–223.
Shin, S. and Lee, W. J. (2014). The effects of technology
readiness and technology acceptance on NFC mobile
payment services in Korea. Journal of Applied Busi-
ness Research, 30(6):1615–1626.
Taylor, S. and Todd, P. A. (1995). Understanding informa-
tion technology usage: A test of competing models.
Teo, T. S. H. and Pok, S. H. (2003). Adoption of WAP-
enabled mobile phones among Internet users. Omega,
31(6):483–498.
Venkatesh, V., Morris, M. G., Davis, G. B., and Davis, F. D.
(2003). User Acceptance of Information Technology:
Toward a Unified View. MIS Quarterly, 27(3):pp. 425
– 478.
Walczuch, R., Lemmink, J., and Streukens, S. (2007). The
effect of service employees’ technology readiness on
technology acceptance. Information and Manage-
ment, 44(2):206–215.
Yen, J.-y. and Chen, S.-c. (2010). Perspectives from the
TRAM Model on Adopting e-Learning: An Analysis
of the Chain and Franchise Industry in Taiwan.
Mobile Instant Messaging for Customer Service Interaction - Preparation of a Model-based Approach Exploring Behavioral Intention
49