Role of Shopping App Attributes in Creating Urges for Impulse Buying: An Empirical Investigation Using SEM and Neural Network Technique

Abstract

With high speed internet, the retailers are continually engaged in upgrading mobile apps that facilitate shoppers in shopping anywhere-anytime and arousing their sudden urges to buy impulsively. The present study endeavors to decipher the antecedents of mobile app-based impulsive buying behavior and determining their relative significance in triggering impulsive urges. Using structural equation modeling, causal analysis was undertaken to identify the role of effort expectancy, price and discounts, atmosphere and layout of app, and user experience and satisfaction in creating impulsive buying intentions. It was observed that price and discounts and user experience didn't have any influence in stirring the consumer for impulsive buying. To determine the relative significance of remaining four, artificial neural network modeling was undertaken. Effort expectancy was noted to have highest influence in creating impulsive urges, followed by atmosphere and layout of an app. User satisfaction had minimum impact. The paper concludes with practical implications for m-commerce players.

Full text

1
Role of Shopping App Attributes in Creating Urges for Impulse
Buying: An Empirical Investigation Using SEM and Neural
Network Technique
Priyanka Gupta (Indian Institute of Management, Raipur, India), Sanjeev Prashar (Indian Institute of Management,
Raipur, India), Chandan Parsad (Indian Institute of Management (IIM), Bodh Gaya, India) and T. Sai Vijay (Indian
Institute of Management, Ranchi, India)
Abstract
Estimated to touch 2.56 billion in 2018, smartphone users are among the largest segments
being targeted by retailers. Backed with high speed internet, these retailers are
continually engaged in upgrading mobile apps that facilitate shoppers in shopping
anywhere-anytime and arousing their sudden urges to buy impulsively. The present study
has been endeavored to decipher the antecedents of mobile app based impulsive buying
behavior and determining their relative significance in triggering impulsive urges. Using
structural equation modelling, causal analysis was undertaken, which identified the role
of effort expectancy, atmosphere, layout and user satisfaction in creating impulsive
buying intentions. The result of structural equation modelling was used as an input for
artificial neural network modeling to determine the relative significance of these four
factors. The result shows that effort expectancy and atmosphere are the strongest
antecedents, while price and discounts, and user experience were noted to have no effect.
The paper concludes with practical implications for m- commerce players.
Key Words: Mobile applications (apps); Impulse buying intention; Impulse buying;
Effort expectancy; User Experience; User Satisfaction; App layout.
1. Introduction
Over the period, phones have transformed from being a traditional voice-based function
to the smartphones that facilitate multimedia exchange, financial transaction, social
media and mobile games. This has powered the development of mobile applications
(apps) that satisfy customers’ multiple expectations (Techcrunch, 2014). With the
increasing usage of smartphones, mobile apps have become popular communication tools
that connects businesses with their customers. Defined as “software downloadable to a
mobile device” (Bellman et al., 2011), mobile apps offer a large gamut of benefits like
time and location independence, easier administration and ubiquity, and context
awareness (Nikou & Economides, 2017). Displaying brand identity throughout the user
experience (Bellman et al., 2011), branded apps serve users with a variety of content and
extend services any time through smartphones. While using mobile apps, shoppers have
control over their decisions according to their requirement and brand preference. It is
pertinent that an app stands on customers’ expectations and have essential features such
as simplicity, social media integration, easy payment and customization (Rose, 2017).
The worldwide mobile commerce revenues are expected to rise from US$ 96.34 billion in
2015 to US$ 693 billion in 2019 (Statista, 2015). Half of all e-commerce site traffic

2
emanated from mobile users, with smartphones and tablets contributing forty and ten
percentages respectively (Dynamic Web, 2015). While Android has launched 2.8 million
apps globally under twenty-nine categories, its competitor Apple has 2.2 million apps
across twenty-one categories in 2017 (Statista, 2017). Owing to the rapid growth of
smartphones, India has emerged as the fastest growing (71%) country in the mobile
application market in terms of downloading (268 million) and usage of apps in 2017
(Saifi, 2017). As against the global average of thirty-eight percent, Statista (2016)
reported that forty-nine percent of respondents in India used mobile devices for
purchasing goods or services.
It has been reported that mobile ecommerce triggers impulsive buying owing to the
availability of multiple pathways through which shoppers are exposed to cues both from
surroundings and within the actual devise (Dynamic Web, 2015). Besides being
simplistic in browsing and buying, the ease of use of mobile devises catalyze urges to buy
product instantaneously.
Upon the scanning of existing literature, it has been observed that studies on mobile apps
largely pertain to factors influencing the installation of mobile app (Harris, Brookshire &
Goyal, 2016), app adoption (Bellman et al., 2011; Kim et al., 2013), usage of mobile app
(Hew et al., 2015; Kim et al., 2016), factors influencing continuous attachment to app
(Furner et al., 2014; Kim et al., 2016; Kim et al., 2015) and usage commendation (Xu et
al., 2015; Yan & Chen, 2011), and relationship with personality traits (Xu et al., 2016). A
study by Yang and Lin (2014) examined satisfaction with and purchase intention on apps.
The study on online impulse buying has grown steadily over the period, along with the
growth of e- commerce industry (Chan et. al., 2017). Mostly such research pertains to
determining the influence of website cues on online impulse buying (Santini et. al., 2018)
or differentiating the factors influencing the offline impulse buying vide online media.
The recent studies on impulse buying have focused on studying the impact of online
reviews on impulsive buying behavior (Zhang et. al., 2017), impact of demonetization on
such buying (Pandya and Pandya, 2018), motivation behind impulse buying (Sundström
et. al., 2019), and impact of emotions on impulse buying (Yi and Jai, 2019). However, no
studies has attempted to decipher the influence of various antecedents of shopping using
mobile app on impulsive buying behavior.
Using structural equation modeling, this paper addresses the gap in the literature to
evolve a model that identifies significant antecedents triggering impulsive urges to buy in
the context of mobile apps. Based upon these identified determinants, artificial neural
networks technique has been used to predict shoppers’ impulsive behavior. The
advantage of using neural networks technique pertains to its capability of modelling non-
linear relationships as against the linear relationships modelled by regression based
methods. In their study, West et al. (1997) observed that neural network technique
performs better than other binary techniques like discriminant analysis and logistic
regression. The paper also examines the relative significance of various determinants that
were identified using structural equation modelling.
Thus, this study is a pioneer attempt at predicting impulsiveness among buyers using

3
mobile shopping apps. Using the techniques - structural equation modelling and artificial
neural network, the paper identifies the specific features of mobile shopping apps, which
marketers must consider while targeting shoppers and triggering their impulsivity. The
paper flow is as follows; The subsequent section details the review of literature and
formulation of hypotheses. This is followed by description of research methodology,
analysis and presentation of results. Discussion, implications and conclusions have
presented in the later part followed by limitations and the scope for future research.
2. Literature Review
The existing literature on mobile app-based consumer behavior, and its antecedents
influencing users’ decision to adopt mobile apps reflects the utilization of well-known
traditional theories on technology adoption like technology acceptance model, diffusion
of innovation and the unified theory of acceptance and use of technology. The present
study examines the influence of various features of mobile apps - effort expectancy,
layout, price and discount, atmosphere, user experience and satisfaction, on shoppers’
impulsive buying. This section details about these six aspects of mobile app and relevant
hypotheses.
2.1. Effort Expectancy
The study by Venkatesh (2012) defined the effort expectancy (EE) as “the degree of ease
associated with consumers’ use of technology” also called as ‘perceived ease of use’
under TAM model (Pynoo et al., 2011). The shopper is likely to achieve greater
satisfaction if the web portal is easy-going (Liu et al, 2013). Studies by Chang et al.
(2012) and Islam and Mazumder (2010) have observed that the mobile device with
touchscreen ability makes the browsing much easier than web portals. While the touch
screen keeps uninterrupted control, which is easy for shoppers to use (Brasel and Gips,
2014 and Leong et al., 2013), Liu et al. (2013) observed that website’s ease of use is an
essential element of visual beauty of the portal. The study by Shaikh et. al. (2018) found
the relationship with adoption intention with the mediating effect of attitude in the
context of mobile banking. Another study by Tandon and Kiran (2018) identified the
effort expectancy is driver leading of behavioral intention for online shopping. The
significance of effort efficiency in prompting shoppers’ behavioral intention has been
demonstrated by Venkatesh (2012) and Hew et al. (2015). The perceived ease of use
triggers both positive and negative emotions of shoppers (Ethier et al., 2006). With
positive influence on positive feelings and negative on negative emotions (Verhagen &
Dolen, 2011), this ease of use further impacts their impulsive buying behavior (Ruiter et
al., 2001). Accordingly, it is proposed:
H1: Mobile shopping app’s effort expectancy will have positive influence on
impulse buying intention.
2.2. Price and Discount
Along with the other important factors that affect shoppers’ buying behavior, the study by
Stern (1962) observed that low price stimulates a feeling among consumers that they are

4
‘spending less than the original’ amount. With the price, discounting is the way to add
value which is temporary and tangible monetary incentive for customers to get benefit for
an specific product for a particular time period (Sheehan et. al., 2019). This perception of
‘unexpected saving’ through low-price influences shoppers’ mood and increases their
intention to buy more (Heilman et al., 2002). As per Arnold and Reynolds (2003), the
consumers’ value shopping includes discounts and price-promotions. Providing a cost
saving experience and economic control, these monetary benefits force consumers to use
the internet (Flanagin & Metzger, 2001; Wolin & Korgaonkar, 2003). The price of a
product and discount offered stimulate both utilitarian and hedonic motives (Verhagen &
Dolen, 2011), which trigger impulse buying behavior (Hardesty & Bearden, 2003). Babin
et al. (1994) posited that shoppers have affinity for bargains and seek price deals, which
satisfy their hedonic motives. Hence, the price and attached discounts are significant
influencers in exciting shoppers for web browsing. The study by Lee and Chen-Yu
(2019) identified the relationship between the price discount and purchase intention via
price discount effect and perceived value using price-quality-value model. As per
Mazaheri et al. (2010), web retailers must consider emotional and hedonic elements while
evolving pricing strategies. A study by Xu and Huang (2014) observed that for hedonic
products, discounts offered have high trigger value on impulsive buying intention.
Therefore, the hypothesis is
H2: Price and discount available on mobile shopping apps will have positive
influence on impulse buying intention
2.3. Layout and Atmosphere of App
In the online context, the layout pertains to consumers’ navigation within the store. As
per Baker et al. (1994), the design factors, in online format are more towards the visibility
of the platform instead of ambient factor. The association among website design and
shoppers’ impulse buying behavior have demonstrated by existing studies (Verhagen &
Dolen, 2011). Krasonikolakis et al. (2018) stated that the layout is practical factor for
consumer decision making process and it is major component of the for consumer to
connect. The study by Vrechopoulos et al. (2004) observed that the freeform layout is
significantly more useful in organizing consumers’ shopping list in the online store. The
study also noted that the grid layout is easier than the racetrack and the freeform layout
formats, though the freeform layout is more entertaining. A proper layout could excite
utilitarian buyers too for purchasing products that have not been planned (Sherman et al.,
1997). Wu et al. (2013) confirmed that the layout affects purchase intention of consumers
positively through attitude towards the website and positively effects their impulse
purchase intention. This study has also perceived the influence of atmosphere on impulse
purchase intention.
As per Kotler (1973, P.50), atmosphere as “the conscious designing of space to create
certain buyer effects, specifically, the designing of buying environment to produce
specific emotional effects in the buyer that enhance purchase probability.” Though in the
offline context, the store atmosphere relates to visual, aural, olfactory and tactile
channels. only two channels - aural and visual deliver the environmental cues in the
online settings. Sherman et al. (1997) observed that the atmosphere factors of the store

5
have a direct positive impact on behavioral intention, these factors arouse desire (urge) to
buying impulsively (Eroglu & Machleit, 1993). Eroglu et al. (2001) also observed that
this in-store retail atmosphere provides pleasure to shoppers.
Inferring the fact that a good atmosphere and an optimal layout extends ease in gathering
product and related information, and facilitates buyers in decision-making, it is
hypothesized that:
H3: Layout of the mobile shopping app will have positive influence on
impulse purchase intention.
H4: Atmosphere of the mobile shopping app will have positive influence on
impulse purchase intention.
2.4. User Experience and Satisfaction of Shopping App
Kang et. al. (2017) stated in the social networking site context, that a satisfied consumer,
experience overall flow of the online system and which leads to enhance consumer
intention to use. Hassenzahl (2008) defines user experience as “a momentary, primarily
evaluative feeling (good-bad) while interacting with a product or service.” An optimal
user experience is cumulatively manifested through the product’s ability to fulfill the
shoppers’ hedonic needs for self- sufficiency, fitness and stimulation. The users interact
through the surroundings and its environment for enhancing online experience (Li et al.,
2002). Keng et al. (2011) categories it into two group - direct and indirect product
experience. While the direct experience is acquired directly from the website and
influences shoppers’ future purchases (Chiou et al., 2008), the indirect experience
includes activities like looking at graphics and spending time for reading text, basically
related to interaction or willingness to move (Huang, 2003). The emotional experience of
consumers is conceptualized as flow - enjoyment and intense concentration (Trevino &
Webster, 1992). There is encouraging relationship among user experience and impulsive
buying through mood and emotional experience (Kahn and Isen, 1993).
Olaru et al. (2008), defined the user satisfaction “customers evaluate future purchase
intention based on the value obtained from episode/contacts, with relationship benefits
being a proxy for expectations of future benefits.” For converting internet users into
online shoppers and providing them with sustained satisfaction, O’cass, and Fenech
(2003) observed that it is pertinent for retailers to focus on “web shopping compatibility,
internet self-efficacy evaluations and portal’s security dimensions.” Later, Khalifa and
Liu (2007) observed and confirmed the association between the user experience and
satisfaction in an online shopping setting that leads to purchase intention The study by
Alnawas and Aburub (2016) in the context of branded mobile apps said that consumers
up-to-date, smarter, improve status among colleagues, refreshes their mood, connect with
the community and promote value due to the introduction of apps, which impact future
purchase intention. Thus, user satisfaction is the main variable, which stimuluses
consumer buying behaviour (Wu et. al., 1970) and on the other side, in the context of
hotel booking, the user satisfaction is positively related to consumer intention to book
hotel (Hwang et. al., 2018). The study in online context by Bressolles et al. (2007)

6
posited the direct impact of consumer satisfaction on impulse buying behavior.
Based on the above arguments, the following hypotheses are stated:
H5: User experience with mobile shopping apps will have positive impact on
shoppers’ impulse purchase intention.
H6: User satisfaction with mobile shopping apps will have positive impact on
impulse purchase intention.
2.5. Impulse Buying Intention
As against the cognitive processes, impulse buying involves extemporaneity in exhibiting
conative behavior for the product (Vohn & Faber, 2007). Rook and Fisher (1995) posited
that impulsivity traits determine consumers’ intention to buy impulsively. Shoppers with
high impulsivity are usually casual, follow affection and are prominently emotional
(Wood, 1998). Such shoppers with excessive impulsive traits have greater impulse
buying intentions (Dholakai, 2000; Puri, 1996). As compared to low impulsive buyers,
the high impulsive shoppers connect internal emotion to the external information, and
show aggressive impulse buying intentions to purchase. As per Lee and Kacen (2008),
the influence of store environmental stimuli triggers immediate reactions and negates
planned buying. It has also been noted that experiencing of urge to buy impulsively leads
to impulsive behaviour, but it is not necessary that high urge to buy always leads to actual
impulse buying behaviour (Beatty & Ferrell, 1998). Drossos et al. (2014) examined the
impact of mobile-based advertisements on consumers impulsiveness and observed that
the shoppers exhibit more impulsiveness for low involvement products.
Thus, it is hypothesized that
H7: The impulse buying intention will have positive impact on impulse
buying behaviour.

7
Figure 1: Research Model
3. Research Methodology
3.1. Step 1: Research Design

Complete assessment of previous literature on mobile apps adoption and impulse buying
was assumed to increase perceptions of mobile app based consumer behavior. A list of
six variables that have influence on impulse buying behavior through mobile apps was
prepared. The existing validated scales were used which was taken form past studies. To
measure effort expectancy and price and discount, two scales with three items each were
adapted from Venkatesh (2012). Two scales of four-item each from Wu et al. (2013)
were adopted to measure layout and atmospherics of mobile apps. Prashar et al. (2016a)
study was referred and a four-item scale was adopted to measure the construct ‘user
experience.’ The scale for construct ‘user satisfaction’ was assessed using four items
scale from Alnawas and Aburub (2016). Finally, for measuring impulse buying intention
and impulse buying, two scales with seven and four items, were adapted from the study
of Chen and Wang (2016) and Beatty and Ferrell (1998) respectively. The total of thirty-
three items, measuring eight variables, were framed into statements. The seven-point
Likert’s scale was used to measure the respondents’ agreement/ disagreement with the
statements.
This list of six independent variables impacting mobile app based impulse buying, along
with impulse buying intention and impulse buying, was presented to the expert panel.
This panel, comprising of four scholars pursuing their doctoral degree in marketing, one

8
industry expert and three members from academia, examined the instrument with respect
to its language and usability. With minor modifications, the instrument was taken for
final data collection. To ensure the pertinence of sampling unit, the instrument opened
with the question – Have you used Amazon mobile app for shopping? The respondents
who confirmed the usage of this specific app were eligible for the study.
3.2. Step 2: Participants and Data Collection

Since majority of app consumers are under the age of thirty years (e-marketer, 2015), it
was decided to undertake the study among college students (Lester et. al., 2005). The
questionnaire was loaded in Qualtrics, an online survey tool. The convenience sampling
technique were used for the survey. The link was mailed to the students pursuing their
post-graduation. The data collected between February – April 2017 period. The
respondents were requested to specify shopping behavior using Amazon app only. In all,
responses from 357 respondents were collected. After removing partially filled
instruments, 297 data were used for further analysis. Thus sample can be considered
sufficient as suggested by Manhotra and Birks (1998); Westland (2010).
3.3. Step 3: Data Analysis
The analysis was carried out in three steps. Using SmartPLS 2.0, reliability, convergent
and discriminant validities of the hypothesized research model were checked. This
followed by testing the proposed hypotheses using structural equation modeling (SEM),
lastly, artificial neural network technique has been applied to get the importance of each
variable in buying process (Chen and Chong, 2012). Since, the research model in this
study is simple and does not have complex cause and effect relationship among the latent
variables, PLS based regression approach shall provide generalized results through the
combined features of component analysis and multiple regressions (Hair et al., 2011).
4. Result and Analysis
4.1. Structural Equation Modelling
The analysis had been done using multiple analytical approach. In the first phase of the
study structural equation modeling (SEM) has been applied to demonstrate the reliability
and validity of the measured relationship, while the other technique, Artificial neural
network (ANN) will provide the importance of each variable on impulse buying. The
SEM analysis divided into two parts: first measurement model, in which the latent
variables are defined by observed variables and identify reliability and validity of the of
construct. Second, structural model, which run regression between dependent variables
and independent variables and overall, measure the model validity (Farag et al., 2007).
4.1.1. Reliability and Validity Analyses
To assess the robustness of the proposed model, reliability and validity of the factors
were examined. From Table 1, it is observed that Cronbach’s alpha value for each
construct is more than 0.7, values of average variance extracted (AVE) are more than 0.5
and composite reliability (CR) are more than 0.8. Since these values are above the
threshold values as suggested by Hair et al. (2010), this establishes the reliability and

9
validity of the constructs. Also, the diagonal values representing square-root of AVE
were observed to be greater than all the correlation values in the corresponding columns.
This confirms the discriminant validity.
Table I: AVE, CR and Inter-Correlation Matrix
Construct

CR
AVE
EE
PD
LYT
ATM
UE
US
IBI
Effort Expectancy
0.851
0.910
0.771
0.878
Price and Discount
0.894
0.934
0.825
0.756
0.908
Layout
0.849
0.897
0.685
0.875
0.752
0.828
Atmosphere
0.814
0.877
0.641
0.842
0.805
0.802
0.801
User Experience
0.82
0.88
0.648
0.809
0.792
0.792
0.800
0.868
User Satisfaction
0.89
0.924
0.754
0.783
0.782
0.782
0.766
0.713
0.805
Impulse Buying Intention
0.859
0.892
0.542
0.446
0.524
0.588
0.620
0.446
0.589
0.736
4.1.2. Hypothesis Testing
The model comprised of six exogenous latent variables, and two endogenous latent
constructs - impulse buying intention and impulse buying. Table 2 reflects structural
parameter estimates. Out of the six proposed hypotheses (H1 to H6), four were found to
be significant. The effort expectancy (=0.555, p<0.05), which is the most used
antecedent in technology adoption, was observed to significantly affect app users’
impulse buying intention. Thus, the hypothesis, H1 is accepted. The ambient factors -
layout (= 0.383, p<0.05) and atmosphere (=0.564, p<0.05) were also noted to
significantly affect shoppers’ impulse buying intention in the context of mobile apps.
Hence, both the hypotheses H3 and H4 are accepted. Similarly, with the values of =
0.387 and p<0.05, user satisfaction was also found to impact app shoppers’ impulse
buying intention, indicating the acceptance of hypothesis H6.
On the other hand, two of the independent variables - price and discount (= 0.031,
p>0.05) and user experience (= -0.168, p>0.05) were not supported for their influence
on impulse buying intention. Accordingly, hypotheses H2 and H5 are not accepted.
Finally, the influence of impulse buying intention on impulsive buying was noted to be
significant with values - = 0.430 and p<0.05. Thus, hypothesis H7 is accepted.
Table II: Results of Hypothesis Testing
Hypothesis

P-Value
Result
H1: Effort Expectancy: Impulse Buying Intention
0.555
0.014*
Accepted
H2: Price and Discount: Impulse Buying Intention
0.031
0.827
Not Accepted
H3: Layout: Impulse Buying Intention
0.383
0.046*
Accepted
H4: Atmosphere: Impulse Buying Intention
0.564
0.004*
Accepted
H5: User Experience: Impulse Buying Intention
-0.168
0.350
Not Accepted
H6: User Satisfaction: Impulse Buying Intention
0.387
0.019*
Accepted
H7: Impulse Buying Intention: Impulse Buying
Behaviour
0.430
0.00*
Accepted

10
4.2. Neural Network Analysis
Predictive modeling is applied to define rules for forecasting the value(s)data set using
the value(s) of element(s) from the input data set. Various predictive models have
evolved under the categories of statistical techniques, pattern recognition, and machine
learning. The present study has uniqueness of applying a binary technique of artificial
intelligence on the results obtained from SEM based impulse buying. The linear
regression techniques like multiple regression and structural equation modeling
determine linear relationships between the variables and lead to simplification of the
complex decision process. However, these are limited to linear relationships only. The
application of binary predictive techniques like artificial neural network, help in
determining the non-linear relationships between the variables. The advantage of neural
network technique is, to solve the complex linear and non-linear connection between the
technology adoption antecedents and its adoption and also, the technique can solve the
non-compensatory IT adoption decisions (Chan and Chong, 2012). As an information
processing system, a neural network “uses a number of simple processors that are linked
to learn the relationships between sets of variables.” Artificial neural network operates by
continually altering the values of available interlinks among neural units (Zhang et al.,
1998). This classifier technique has been used in marketing decision making for
estimating impulsive buyers, online consumer buyers and price and promotional
sensitivities etc. (Prashar et al., 2015; 2016b). Comparing the forecasting ability of select
binary classifiers, Agrawal and Schorling (1996) observed that artificial neural network
presents better than other binary techniques.
In the present context, this technique shall facilitate in deciphering the complex linear
and/ or non- linear relationships among the predictor variables and impulse buying.
However, since artificial neural network technique is not suitable for testing the causal
relationships, hence the current study has used two techniques sequentially. In the first
stage, structural equation modeling was applied to confirm the relationships between six
antecedent variables, impulse buying intention and impulse buying (dependent variable).
The variables - price and discount, and user experience, were found to be insignificant in
SEM analysis, and were removed from further analysis.
This was followed by application of neural network technique endeavored to determine
the relative importance of the predictor variables. The data set was randomly distributed
in two sets - a training set and a holdout (test) set. With 68% of the respondents, the
elements of training group were used to determine possible predictive associations and to
develop a predictive model. The test group (32% of the data set) was used to validate the
model.
4.2.1. Sensitivity Analysis
The prime objective of performing sensitivity analysis is to identify relevant factors that
have influence on the dependent variable (impulse buying in the context). In the process,
variables with less influence on impulse buying shall be removed (Engelbrecht & Cloete,
1996). For the identification of relative importance of five factors (four predictor
variables and impulse buying intention), sensitivity analysis using neural network

11
technique was carried out. Table 3 reflects the values arrived from the analysis. The
maximum and minimum absolute sensitivity values are 0.611 and 0.011 respectively. As
none of the variables had value equal or near equal to zero in all three hidden layers, there
was no need to remove any variable(s) from the analysis (Chiang et al., 2006). Finally,
data from test set was used to analyze the predictive accuracy.
Table III: Sensitivity Analysis
Variable
Predicted
Hidden layers 1
H (1:1)
H (1:2)
H (1:3)
This shopping app is highly attractive
.129
-.346
-.349
This shopping app is visibly light
-.309
.242
-.224
This shopping app is highly stimulating
-.315
.027
-.346
This shopping app is more lively
.074
.321
-.197
Learning how to use this shopping app is easy for me.
-.404
.249
-.449
My interaction with this app is clear and understandable.
.153
-.152
-.093
In my daily life this app saves time and effort
-.317
-.310
.169
This shopping app has well organized layout
.611
-.345
-.341
This shopping app has good display
.083
-.128
.027
This shopping app has large collection
.162
-.471
-.168
This shopping app has helpful signage
.124
.083
.015
I am satisfied with my overall experience from shopping app
-.068
.303
-.047
I am satisfied with the pre-purchase experience from using this
shopping app
-.256
.410
-.268
I am satisfied with the purchase experience from using this shopping
app
.090
.375
-.014
I am satisfied with the post-purchase experience from using the
shopping app
.011
.425
-.230
I am prompted to buy more because of the discount offer. 
-.149
-.506
-.161
If I see something that really interests me; I buy it without considering
the consequences.
-.125
-.339
-.219
I buy things even though they were not on the shopping list.
-.370
-.347
-.511
I am prompted to buy more because of the faster navigation.
-.270
.406
-.021
I am prompted to buy more because of attractiveness.
-.113
-.447
-.199
I like to buy because of well-organized layout
-.105
-.090
-.476
I am prompted to buy more because of variety of product/choices
-.471
-.450
-.313
4.2.2. Classification Matrix
The accuracy with which a model predicts various outcomes reflects its forecasting
ability (Allenby et al., 2002). This is measured by “the fraction of correct predictions
made.” The forecasting correctness is represented through a classification matrix. The
consequences of the model using training and testing data sets are exhibited in Table 4.
From the table, it is observed that 175 out of 177 shopping app users (based on training
set) who buy impulsively were identified accurately by the model. However, the model
wrongly classified fourteen out of total twenty-five non-impulsive shoppers. Hence,
overall accuracy of the model’s prediction is 92.1%. With the data from the test group,
eighty-eight out of ninety-two users of shopping apps were predicted accurately with
respect to their impulsive buying behavior on their mobile apps. Thus, this neural
network model has accuracy of 95.7%. Accordingly, more than ninety-five mobile app
users out of every hundred were correctly identified by the model either as impulsive or

12
non-impulsive buyers.
Precision or positive predictive value is another indicator that is used to assess the
goodness of a model, is measured by calculating the ratio of forecasted true cases to total
true cases. With ninety- seven percent of precision, the derived model may be considered
to be highly precise. Thus, out of every hundred predictions made from this model,
ninety-seven predictions shall turn out to be correct. Sensitivity (or recall) indicates the
model’s ability to correctly identify the true positive cases (those were also predicted to
be impulsive buyers) out of the overall impulsive buyers (Powers 2011). The model’s
sensitivity value is 97.67%, which indicates its good ability in identifying mobile app
based impulse-buying behavior.
Table IV: Classification Matrix for Training and Test Set
Sample
Observed
Predicted
No
Yes
Percent correct
Training
No
11(TN)
14(FP)
44%
Yes
2(FN)
175(TP)
98.9%
Overall Percentage
6.4%
93.6%
92.1%
Testing
No
4(TN)
2(FP)
66.7%
Yes
2(FN)
84(TP)
97.7%
Overall Percentage
6.5%
93.5%
95.7%
*TN (True Negative), FP (False Positive), FN (False Negative), TP (True Positive)
To determine the model’s ability in forecasting true negative cases, specificity index is
used. Specificity, characterizes as the condition, in which a model correctly identifies the
non-impulsive buyer (“no") among all the negative outcomes (non-impulsive). It reflects
the model’s ability in preventing false alarms. The simulation's specificity value of
66.66% demonstrates that it was able to recognize more than 66% of non-impulsive
buyers that were truly a negative case.
Additionally, to check the goodness of a model, F1 score is calculated. The score equal to
or approaching ‘1’ indicates the model to be a good fit, while the score approaching
towards the zero is termed as bad fit model. The F1 score from the model is 97.67%,
(0.976), indicating the model to be a good fit. All four values - accuracy, precision,
sensitivity and F1 score show that the analytical model developed using artificial neural
network is a good fit.
Table V: Calculated Indicators using Test Set
Indicators
Calculation
Percentage
Accuracy
(TP+TN)/(TP+FP+FN+TN)
(84+4)/(84+4+2+2)
95.65
Positive Predictive
Value
TP/(TP + FP)
84/(84+2)
97.67
Negative Predictive
Value
TN/(TN + FN)
4/(4+2)
66.66
Recall or Sensitivity
TP/(TP + FN)
84/(84+2)
97.67
Specificity
TN/(TN + FP)
4/(4+2)
66.66
F1 score
2 * (PPV * Recall)/(PPV +
Recall)
2*{(97.67*97.67)/(97.67 +
97.67)}
97.67

13
4.2.3. Importance and Significance of Select Variables
The predictor variable’s importance and significance is the amount of variation in the
model’s prediction value for the changing values of the variable. The relative importance
of the four factors comprising of fifteen elements that influence mobile app users’
impulse buying is presented in Table 6.
Table VI: Relative Significance of Select Variables
Variable
Import
ance
Factor
Overall
Importan
ce
Aggregate
Importance
Independent Variables
This shopping app is highly attractive
.027
Atmosphere
2.7%
16.1%
This shopping app is visibly light
.049
Atmosphere
4.9%
This shopping app is highly stimulating
.072
Atmosphere
6.2%
This shopping app is more lively
.023
Atmosphere
2.3%
Learning how to use this shopping app is easy for me.
.082
Effort Expectancy
8.2%
17.3%
My interaction with this app is clear and understandable.
.019
Effort Expectancy
1.9%
In my daily life this app saves time and effort
.072
Effort Expectancy
7.2%
This shopping app has well organized layout
.060
Layout
6%
12.5%
This shopping app has good display
.011
Layout
1.1%
This shopping app has large collection
.029
Layout
2.9%
This shopping app has helpful signage
.025
Layout
2.5%
I am satisfied with my overall experience from shopping
app
.013
User Satisfaction
1.3%
11.3%
I am satisfied with the pre-purchase experience from using
this shopping app
.043
User Satisfaction
4.3%
I am satisfied with the purchase experience from using this
shopping app
.039
User Satisfaction
3.9%
I am satisfied with the post-purchase experience from
using the shopping app
.018
User Satisfaction
1.8%
I am prompted to buy more because of the discount offer.
.061
Impulse Buying Intention
6.1%
42%
If I see something that really interests me I buy it without
considering the consequences.
.044
Impulse Buying Intention
4.4%
I buy things even though they were not on the shopping
list.
.088
Impulse Buying Intention
8.8%
I am prompted to buy more because of the faster
navigation.
.027
Impulse Buying Intention
2.7%
I am prompted to buy more because of attractiveness.
.050
Impulse Buying Intention
5%
I like to buy because of well-organized layout
.037
Impulse Buying Intention
3.7%
I am prompted to buy more because of variety of
product/choices
.113
Impulse Buying Intention
11.3%
Furthermore, impulse buying intention, which is the most important factor leading to
impulse buying, four other factors related to mobile app were found to influence impulse
buying, albeit with varying significance levels. On the aggregate level, effort expectancy
(with 17.3%) was noted to be the most important factor among the four. This was

14
followed by atmosphere (16.1%) and layout (12.5%) of the app. User satisfaction
(11.3%) was observed to have the least important significance. Accordingly, effort
expectancy, being the most important factor, should capture maximum effort and
investment of app developers (Alshare et al., 2015).
While considering the importance of individual elements, the top four important elements
influencing impulse buying were – ‘learning how to use this shopping app is easy for
me,’ ‘in my daily life this app saves time and effort,’ ‘this shopping app is highly
stimulating’ and this ‘shopping app has well organized layout.’ On the other hand, the
four least influential elements were – “my interaction with the app is clear and
understandable”, “I am satisfied with the post-purchase experience from using the
shopping app”, “I am satisfied with my overall experience from shopping app” and “this
shopping app has good display”.
While considering the importance of impulse buying intention, which has 42% of the
total influence, it can be observed that the element ‘I am prompted to buy more because
of variety of product/choices’ contributes eleven percent of the total influence. This is an
important observation for retailers trying to target the consumers through mobile apps.
They should focus heavily on the variety of the products displayed in the app.
5. Discussion
5.1. Theoritical Implication
The exponential growth in the number of smartphones across the world has resulted in
large number of people undertaking their shopping through mobile applications (apps).
By facilitating the comparison of competitors’ pricing and various promotional offers,
and extending consumers’ reviews and rating to the potential buyers, these mobile phones
have bridged the gap between the product research and impulse buying. The ubiquitous
nature of mobile apps and ease of access from any place have impacted shoppers’
purchasing behaviour, especially impulse buying behaviour. Thus, understanding
shoppers’ impulse buying behaviour in the context of mobile shopping apps is a key
research area of consumer purchasing behaviour.
Considering the global smartphone penetration of 7.7 billion in 2019 that is estimated to
reach 3.3 billion in 2018 (Statista, 2019), this research will contribute towards
understanding shopping app users’ impulse buying behavior. This is the first study that
has endeavored to determine the antecedents of shoppers’ impulse buying intention,
followed by the deciphering of the relative significance of each of the factors influencing
such urges to buy in the context of mobile shopping apps. The model delivers a predictive
tool for marketers using mobile apps in identifying the specific features that attract
shoppers to buy product impulsively.
The present paper is pioneer in examining impulse buying behaviour as displayed by
mobile shoppers, with application of two techniques - causal analysis (SEM) and
predictive analysis (artificial neural network). The main objective of the study was to
recognize and examine the best significant predictors of impulse buying behaviour in the
context of mobile shopping apps. The paper presents and investigates an innovative

15
model that comprises six probable predictor variables that influence impulse buying
intention. The results of causal analysis establish that only four (effort expectancy,
layout, atmosphere and user satisfaction) of the six predictors were found to have
significant influence on impulse buying intention. The influence of effort expectancy
with impulse buying intention confirm the study of Verhagen and Dolen (2011). Layout
and atmosphere positively influence impulse buying intention confirm the study of Wu
et. al. (2013); Akram et. al. (2016) in the context of normal purchase. Lastly, the
significance of user satisfaction with impulse buying intention confirm the study of
Alnawas and Aburub (2016). Price & discount and user experience were noted to be
insignificant. The four noteworthy variables along with impulse buying intention were
used in the predictive analysis to examine their relative significance and predictive
abilities. This innovative combination of two different methodologies is the novelty being
offered by this study.
This paper confirms the findings of Wu et al. (2013), they identified the influence of
retail environment in enhancing the probability of impulsive purchasing. Drawing
similarity, it can be stated that shoppers give a lot of importance to the in-store shopping
environment and so expect the same in online shopping (e-commerce/m-commerce/app
based) context. If the feature of the mobile app matches with that of consumer
expectation, they can easily connect with the app and its offerings.
The mobile app providers need to draft their strategic decisions according to consumer
segment(s) and target the shoppers’ segment as per their need and expectation. App
providers must position the product and its element as urge creators for attracting
shoppers to purchase impulsively. The operational plan and other inventory related
management could be taken from this model and its predictive values analyzed through
neural network technique.
5.2. Practical Implication
The mobile app providers must focus on all the antecedents (effort expectancy,
atmosphere, user satisfaction and layout) of the app. The study will help app providers to
understand the basic requirements of the consumers and focus on those for increasing
sales. Building a respectable brand image through effective management of the mobile
app will lure the shoppers to spend more time on app, and to buy without any plan.
Testing the impact of six variables on mobile app users’ impulse buying intention, the
findings of the study emphasize on the importance of effort expectancy, layout,
atmosphere, and user satisfaction on impulse buying intention. The insignificance of
price & discount variable on impulse buying intention was unexpected and in
contradiction to the study by Xu and Xuang (2014), which had observed that promotional
discount is one of the strongest factors for triggering impulse buying intention. On the
other hand, Xu (2015) noted that discounts trigger impulse buying among the shoppers
but only for hedonic products and/or inexpensive products. The present study also
observed an insignificant effect of user experience on impulse buying intention. User
experience pertains to the mixture of different emotions along with affective and
cognitive responses, which basically measure the after-purchase process, whereas

16
impulse buying is a frequent and sudden process that may not take experience into
consideration (Herabadi et al., 2009).
These significant variables should be taken into consideration by app providers for
developing and maintaining long-term relationship with the users of mobile shopping
apps. It primarily depends on atmosphere of the app, which is related to the effort
expectancy, ambient factors, color and background design of the app, followed by, layout
and user satisfaction. In the context of effort expectancy, the apps provider must design
the app interface in a way that the most relevant buttons are located within the convenient
reach of the shoppers and require single hand usage. Further, it is advised that highly
functional and frequently used buttons are grouped together. Simplified language for
instructions adds to enhanced effort expectancy.
Related to the product portfolio’s depth and breadth, user satisfaction forms another
essential requisite for making app impulsive oriented. It must be endeavored that
information in the app is perceived to be customized and personalized by the shoppers.
Also, it must extend appealing and gratifying features that arouse amusing and emotional
experience (Alnawas and Aburub, 2016).
The study has also demonstrated that atmosphere of the app has maximum impact on
shoppers’ impulse buying behavior. Hence, while focusing on improving the app’s
atmosphere, it is imperative that app providers understand that consumer satisfaction can
be generated by undertaking proactive steps to create a sustainable environment that
would take care of all the characteristics of the mobile apps. For example, regarding the
atmospherics of the app, the endeavor should be to improve app’s design and color, the
way the product is shown in the app, and the pattern of product arrangement. This must
be aimed at raising shoppers’ pleasure and arousal. Some of the interventions that can be
taken include using consistent design theme, and graphics and animation, as against
textual information, which has relatively less noticeability. Bright and lively colored apps
are expected to generate emotional arousal (Wu et al., 2013). Floh and Madlberger (2013)
recommended the usage of moderate resolution that enables the viewing of mobile screen
at a glance. Interactive mobile app shall contribute to arousing users for enhanced
impulsiveness.
6. Limitation and Future Research
The present study has examined the impact of significant features of the mobile apps on
shoppers’ impulsive buying behaviour and has also identified the relative significance of
the select variables. However, the study has few limitations that can be focused in future
research work. This paper is an outcome of cross-sectional research and hence, it might
have missed the dynamics of Indian m- commerce retail market. Future studies must
consider this dynamism by undertaking longitudinal studies. Another set of studies can
examine shoppers’ perceived difference of buying from offline stores, online stores and
mobile apps. Studies may also explore the impact of important product features like price
and category, and shoppers’ characteristics including perceived brand distinctions,
gender, etc., in moderating the relationships between impulse buying intention and
continuous impulse buying intention. Further studies can assess the impact of social

17
influences on individuals’ impulsive triggers. Cross-cultural research, if taken, may
facilitate in generalizing the research findings and to examine the differences, if any, in
behaviour of shoppers across different cultures. This study pertains to shopping apps only
for examining users’ impulsive behaviour. The future studies must use other categories of
apps like personalized app, gaming app, entertainment app, etc. The study has not taken
any promotional offer which is very common in the online shopping like instant discount
and cashback, the future research can be done using promotional offer as moderating
variables.
References
Agrawal, D. and Schorling, C., 1996. Market share forecasting: An empirical
comparison of artificial neural networks and multinomial logit model. Journal of
Retailing, 72(4), pp.383-407.
Alnawas, I. and Aburub, F., 2016. The effect of benefits generated from interacting
with branded mobile apps on consumer satisfaction and purchase intentions. Journal
of Retailing and Consumer Services, 31, pp.313-322.
Alshare, K.A., El-Masri, M. and Lane, P.L., 2015. The determinants of student effort
at learning ERP: A cultural perspective. Journal of Information Systems
Education, 26(2), p.117.
Allenby, G., Arora, N., Diener, C., Kim, J., Lotti, M. and Markowitz, P., 2002.
Distinguishing likelihoods, loss functions, and heterogeneity in the evaluation of
marketing models. Canadian Journal of Marketing Research, 20(1), pp.44-59.
Arnold, M.J. and Reynolds, K.E., 2003. Hedonic shopping motivations. Journal of
retailing, 79(2), pp.77-95.
Babin, B.J., Darden, W.R. and Griffin, M., 1994. Work and/or fun: measuring
hedonic and utilitarian shopping value. Journal of consumer research, 20(4), pp.644-
656.
Baker, J., Grewal, D. and Parasuraman, A., 1994. The influence of store environment
on quality inferences and store image. Journal of the academy of marketing
science, 22(4), pp.328-339.
Beatty, S.E. and Ferrell, M.E., 1998. Impulse buying: Modeling its
precursors. Journal of retailing, 74(2), pp.169-191.
Bellman, S., Potter, R.F., Treleaven-Hassard, S., Robinson, J.A. and Varan, D., 2011.
The effectiveness of branded mobile phone apps. Journal of interactive
Marketing, 25(4), pp.191-200.
Brasel, S.A. and Gips, J., 2014. Tablets, touchscreens, and touchpads: How varying
touch interfaces trigger psychological ownership and endowment. Journal of
Consumer Psychology, 24(2), pp.226-233.

18
Brasel, S.A. and Gips, J., 2014. Tablets, touchscreens, and touchpads: How varying
touch interfaces trigger psychological ownership and endowment. Journal of
Consumer Psychology, 24(2), pp.226-233.
Chang, T.Y., Tsai, C.J. and Lin, J.H., 2012. A graphical-based password keystroke
dynamic authentication system for touch screen handheld mobile devices. Journal of
Systems and Software, 85(5), pp.1157-1165.
Chen, Y.F. and Wang, R.Y., 2016. Are humans rational? Exploring factors
influencing impulse buying intention and continuous impulse buying
intention. Journal of Consumer Behaviour, 15(2), pp.186-197.
Chan, T.K., Cheung, C.M. and Lee, Z.W., 2017. The state of online impulse-buying
research: A literature analysis. Information & Management, 54(2), pp.204-217.
Chiang, W.Y.K., Zhang, D. and Zhou, L., 2006. Predicting and explaining patronage
behavior toward web and traditional stores using neural networks: a comparative
analysis with logistic regression. Decision Support Systems, 41(2), pp.514-531.
Chiou, W.B., Wan, C.S. and Lee, H.Y., 2008. Virtual experience vs. brochures in the
advertisement of scenic spots: How cognitive preferences and order effects influence
advertising effects on consumers. Tourism Management, 29(1), pp.146-150.
Dholakia, U.M., 2000. Temptation and resistance: An integrated model of
consumption impulse formation and enactment. Psychology & Marketing, 17(11),
pp.955-982.
Drossos, D.A., Kokkinaki, F., Giaglis, G.M. and Fouskas, K.G., 2014. The effects of
product involvement and impulse buying on purchase intentions in mobile text
advertising. Electronic Commerce Research and Applications, 13(6), pp.423-430.
Dynamic Web, (2015). The Impulse for Ecommerce, pt. III: Harnessing Mobile
Impulse Buys. https://dynamicweb.com/company/blog/the-impulse-for-ecommerce-
pt-iii-harnessing-mobile-impulse-buys. (Accessed on 17 Jan 2018).
E-marketer. (2014). Smartphone Users Worldwide Will Total 1.75 Billion in 2014.
https://www.emarketer.com/Article/Smartphone-Users-Worldwide-Will-Total-175-
Billion-2014/1010536 (Accessed on 18 Jan 2018)
E-marketer. Urban India's Smartphone Population Ages.
https://www.emarketer.com/Article/Urban-Indias-Smartphone-Population-
Ages/1012743. (2015) (Accessed on 23 Feb 2018)
Engelbrecht, A.P. and Cloete, I., 1996, June. A sensitivity analysis algorithm for
pruning feedforward neural networks. In Neural Networks, 1996., IEEE International
Conference on(Vol. 2, pp. 1274-1278). IEEE.
Eroglu, S.A. and Machleit, K.A., 1993. Atmospheric factors in the retail environment:

19
sights, sounds and smells. ACR North American Advances.
Éthier, J., Hadaya, P., Talbot, J. and Cadieux, J., 2006. B2C web site quality and
emotions during online shopping episodes: An empirical study. Information &
Management, 43(5), pp.627-639.
Eroglu, S.A., Machleit, K.A. and Davis, L.M., 2001. Atmospheric qualities of online
retailing: A conceptual model and implications. Journal of Business research, 54(2),
pp.177-184.
Even Rose, (2017). Business.com: Tap for success; 5 must have features of a
corporate mobile app. https://www.business.com/articles/5-must-have-features-of-a-
corporate-mobile-app/. Assessed on 17 May 2017
Flanagin, A.J. and Metzger, M.J., 2001. Internet use in the contemporary media
environment. Human communication research, 27(1), pp.153-181.
Floh, A. and Madlberger, M., 2013. The role of atmospheric cues in online impulse-
buying behavior. Electronic Commerce Research and Applications, 12(6), pp.425-
439.
Furner, C.P., Racherla, P. and Babb, J.S., 2014. Mobile app stickiness (MASS) and
mobile interactivity: a conceptual model. The Marketing Review, 14(2), pp.163-188.
Hair, J.F., Ringle, C.M. and Sarstedt, M., 2011. PLS-SEM: Indeed a silver
bullet. Journal of Marketing theory and Practice, 19(2), pp.139-152.
Hardesty, D.M. and Bearden, W.O., 2003. Consumer evaluations of different
promotion types and price presentations: the moderating role of promotional benefit
level. Journal of retailing, 79(1), pp.17-25.
Harris, M.A., Brookshire, R. and Chin, A.G., 2016. Identifying factors influencing
consumers’ intent to install mobile applications. International Journal of Information
Management, 36(3), pp.441-450.
Hassenzahl, M., 2008, September. User experience (UX): towards an experiential
perspective on product quality. In Proceedings of the 20th Conference on
l'Interaction Homme-Machine (pp. 11-15). ACM.
Heilman, C.M., Nakamoto, K. and Rao, A.G., 2002. Pleasant surprises: Consumer
response to unexpected in-store coupons. Journal of Marketing Research, 39(2),
pp.242-252.
Herabadi, A.G., Verplanken, B. and Van Knippenberg, A., 2009. Consumption
experience of impulse buying in Indonesia: Emotional arousal and hedonistic
considerations. Asian Journal of Social Psychology, 12(1), pp.20-31.
Hew, J.J., Lee, V.H., Ooi, K.B. and Wei, J., 2015. What catalyses mobile apps usage

20
intention: an empirical analysis. Industrial Management & Data Systems, 115(7),
pp.1269-1291.
Huang, M.H., 2003. Modeling virtual exploratory and shopping dynamics: an
environmental psychology approach. Information & Management, 41(1), pp.39-47.
Islam, R., Islam, R. and Mazumder, T., 2010. Mobile application and its global
impact. International Journal of Engineering & Technology (IJEST), 10(6), pp.72-78.
Kahn, B.E. and Isen, A.M., 1993. The influence of positive affect on variety seeking
among safe, enjoyable products. Journal of Consumer Research, 20(2), pp.257-270.
Keng, C.J., Ting, H.Y. and Chen, Y.T., 2011. Effects of virtual-experience
combinations on consumer-related “sense of virtual community”. Internet
Research, 21(4), pp.408-434.
Khalifa, M. and Liu, V., 2007. Online consumer retention: contingent effects of
online shopping habit and online shopping experience. European Journal of
Information Systems, 16(6), pp.780-792.
Kotler, P., 1973. Atmospherics as a marketing tool. Journal of retailing, 49(4), pp.48-
64.
Kim, E., Lin, J.S. and Sung, Y., 2013. To app or not to app: Engaging consumers via
branded mobile apps. Journal of Interactive Advertising, 13(1), pp.53-65.
Kim, S.C., Yoon, D. and Han, E.K., 2016. Antecedents of mobile app usage among
smartphone users. Journal of marketing communications, 22(6), pp.653-670.
Kim, S.J., Wang, R.J.H. and Malthouse, E.C., 2015. The effects of adopting and using
a brand's mobile application on customers' subsequent purchase behavior. Journal of
Interactive Marketing, 31, pp.28-41.
Lee, J.A. and Kacen, J.J., 2008. Cultural influences on consumer satisfaction with
impulse and planned purchase decisions. Journal of Business Research, 61(3),
pp.265-272.
Leong, L.Y., Hew, T.S., Tan, G.W.H. and Ooi, K.B., 2013. Predicting the
determinants of the NFC-enabled mobile credit card acceptance: A neural networks
approach. Expert Systems with Applications, 40(14), pp.5604-5620.
Lester, D.H., Forman, A.M. and Loyd, D., 2006. Internet shopping and buying
behavior of college students. Services Marketing Quarterly, 27(2), pp.123-138.
Li, H., Daugherty, T. and Biocca, F., 2002. Impact of 3-D advertising on product
knowledge, brand attitude, and purchase intention: The mediating role of
presence. Journal of advertising, 31(3), pp.43-57.

21
Liu, Y., Li, H. and Hu, F., 2013. Website attributes in urging online impulse
purchase: An empirical investigation on consumer perceptions. Decision Support
Systems, 55(3), pp.829-837.
Malhotra, N. and Birks, D.F., 1998. Marketing research: An applied approach.
Prentice Hall/Financial Times.
Mazaheri, E., Richard, M.O. and Laroche, M., 2010. Investigating the moderating
impact of hedonism on online consumer behavior. Journal of Global Academy of
Marketing Science, 20(2), pp.123-134.
Nikou, S.A. and Economides, A.A., 2017. Mobile-Based Assessment: Integrating
acceptance and motivational factors into a combined model of Self-Determination
Theory and Technology Acceptance. Computers in Human Behavior, 68, pp.83-95.
O’Cass, A. and Fenech, T., 2003. Web retailing adoption: exploring the nature of
internet users Web retailing behaviour. Journal of Retailing and Consumer
services, 10(2), pp.81-94.
Olaru, D., Purchase, S. and Peterson, N., 2008. From customer value to repurchase
intentions and recommendations. Journal of Business & Industrial Marketing, 23(8),
pp.554-565.
Powers, D.M., 2011. Evaluation: from precision, recall and F-measure to ROC,
informedness, markedness and correlation.
Prashar, S., Parsad, C. and Vijay, T.S., 2015a. Factors prompting impulse buying
behaviour-study among shoppers in India. International Journal of Indian Culture
and Business Management, 11(2), pp.219-244.
Prashar, S., Parsad, C. and Vijay, T.S., 2015b. Application of neural networks
technique in predicting impulse buying among shoppers in India. Decision, 42(4),
pp.403-417.
Prashar, S., Vijay, T.S. and Parsad, C., 2016a. SEGMENTING ONLINE
SHOPPERS: A STUDY OF CONSUMERS'WEB PORTAL SELECTION
MOTIVATIONS FOR E-SHOPPING. Asian Academy of Management
Journal, 21(1), p.27.
Prashar, S., Vijay, T.S. and Parsad, C., 2016b. Predicting Online Buying Behavior
Among Indian Shoppers Using a Neural Network Technique. International Journal of
Business and Information, 11(2), p.175.
Parsad, C., Prashar, S. and Sahay, V., 2017. Impact of Impulsive Personality Traits
and Store Environment on Impulse Buying Behavior. Journal of Business &
Management, 23.
Puri, R., 1996. Measuring and Modifying Consumer Impulsiveness: A Cost‐Benefit

22
Accessibility Framework. Journal of consumer Psychology, 5(2), pp.87-113.
Pynoo, B., Devolder, P., Tondeur, J., Van Braak, J., Duyck, W. and Duyck, P., 2011.
Predicting secondary school teachers’ acceptance and use of a digital learning
environment: A cross-sectional study. Computers in Human Behavior, 27(1), pp.568-
575.
Rahis Saifi, B2C (Jan 6, 2017). The 2017 Mobile App Market: Statistics, Trends, and
Analysis. https://www.business2community.com/mobile-apps/2017-mobile-app-
market-statistics-trends-analysis-01750346. (Accessed on 17 Jan, 2018)
Rook, D.W. and Fisher, R.J., 1995. Normative influences on impulsive buying
behavior. Journal of consumer research, 22(3), pp.305-313.
Ruiter, R.A., Kok, G., Verplanken, B. and Brug, J., 2001. Evoked fear and effects of
appeals on attitudes to performing breast self-examination: An information-
processing perspective. Health Education Research, 16(3), pp.307-319.
Sherman, E., Mathur, A. and Smith, R.B., 1997. Store environment and consumer
purchase behavior: mediating role of consumer emotions. Psychology and
Marketing, 14(4), pp.361-378.
Statista, (2015). Transaction value of global m-commerce sales from 2014 to 2019 (in
billion U.S. dollars). https://www.statista.com/statistics/557951/mobile-commerce-
transaction-value-worldwide/. (Assessed on 17 Jan 2018)
Statista, (2016). Countries that exceed global average for purchasing a product or
service via mobile device as of March 2016.
https://www.statista.com/statistics/278083/mobile-commerce-penetration-in-selected-
countries/. (Assessed on 17 Jan 2018)
Statista, (2017). Number of smartphone users in India from 2015 to 2021 (in
millions). [online] https://www.statista.com/statistics/203734/global-smartphone-
penetration-per-capita-since-2005/. (Assessed on 12 December 2019)
Statista, (2019). Global smartphone penetration rate as share of population from 2016
to 2020. [online] https://www.statista.com/statistics/467163/forecast-of-smartphone-
users-in-india/. (Assessed on 16 May 2017)
Stern, H., 1962. The significance of impulse buying today. The Journal of Marketing,
pp.59-62.
TechCrunch (2014). Mobile App Usage Increases In 2014, As Mobile Web Surfing
Declines. https://techcrunch.com/2014/04/01/mobile-app-usage-increases-in-2014-as-
mobile-web-surfing-declines/. Accessed on 17 May 2017

23
Trevino, L.K. and Webster, J., 1992. Flow in computer-mediated communication:
Electronic mail and voice mail evaluation and impacts. Communication
research, 19(5), pp.539-573.
Venkatesh, V., Thong, J.Y. and Xu, X., 2012. Consumer acceptance and use of
information technology: extending the unified theory of acceptance and use of
technology. MIS quarterly, pp.157-178.
Verhagen, T. and van Dolen, W., 2011. The influence of online store beliefs on
consumer online impulse buying: A model and empirical application. Information &
Management, 48(8), pp.320-327.
Vohs, K.D. and Faber, R.J., 2007. Spent resources: Self-regulatory resource
availability affects impulse buying. Journal of consumer research, 33(4), pp.537-547.
Vrechopoulos, A.P., O’keefe, R.M., Doukidis, G.I. and Siomkos, G.J., 2004. Virtual
store layout: an experimental comparison in the context of grocery retail. Journal of
Retailing, 80(1), pp.13-22.
West, G.B., Brown, J.H. and Enquist, B.J., 1997. A general model for the origin of
allometric scaling laws in biology. Science, 276(5309), pp.122-126.
Westland, J.C., 2010. Lower bounds on sample size in structural equation
modeling. Electronic Commerce Research and Applications, 9(6), pp.476-487.
Wolin, L.D. and Korgaonkar, P., 2003. Web advertising: gender differences in
beliefs, attitudes and behavior. Internet Research, 13(5), pp.375-385.
Wood, M., 1998. Socio-economic status, delay of gratification, and impulse
buying. Journal of economic psychology, 19(3), pp.295-320.
Wu, L., KANG, M. and Yang, S.B., 1970. What makes users buy paid smartphone
applications? Examining app, personal, and social influences. The Journal of Internet
Banking and Commerce, 20(1), pp.1-22.
Wu, W.Y., Lee, C.L., Fu, C.S. and Wang, H.C., 2013. How can online store layout
design and atmosphere influence consumer shopping intention on a
website?. International Journal of Retail & Distribution Management, 42(1), pp.4-24.
Xu, C., Peak, D. and Prybutok, V., 2015. A customer value, satisfaction, and loyalty
perspective of mobile application recommendations. Decision Support Systems, 79,
pp.171-183.
Xu, R., Frey, R.M., Fleisch, E. and Ilic, A., 2016. Understanding the impact of
personality traits on mobile app adoption–Insights from a large-scale field
study. Computers in Human Behavior, 62, pp.244-256.
Xu, Y. and Huang, J.S., 2014. Effects of price discounts and bonus packs on online

24
impulse buying. Social Behavior and Personality: an international journal, 42(8),
pp.1293-1302.
Yang, H.L. and Lin, C.L., 2014. Why do people stick to Facebook web site? A value
theory-based view. Information Technology & People, 27(1), pp.21-37.
Yan, B. and Chen, G., 2011, June. AppJoy: personalized mobile application
discovery. In Proceedings of the 9th international conference on Mobile systems,
applications, and services (pp. 113-126). ACM.
Zhang, G., Patuwo, B.E. and Hu, M.Y., 1998. Forecasting with artificial neural
networks:: The state of the art. International journal of forecasting, 14(1), pp.35-62.
Appendix I
Effort Expectancy
Learning how to use this shopping app is easy for me.
My interaction with this app is clear and understandable.
In my daily life this app saves time and effort
Price and Discount
The shopping app offers discount and promotions
The shopping app offers me more discount as compared to website
The shopping app offers me more deal
Layout
This shopping app has well-organized layout
This shopping app has good display
This shopping app has large collection
This shopping app has helpful signage
Atmosphere
This shopping app is highly attractive
This shopping app is visibly light
This shopping app is highly stimulating
This shopping app is more lively
User Experience
The shopping app is easy to use
The shopping app has faster navigation
The shopping app provides personalized features
The shopping app offers privacy of personal information
User Satisfaction
I am satisfied with my overall experience from shopping app
I am satisfied with the pre-purchase experience from using this shopping app
I am satisfied with the purchase experience from using this shopping app
I am satisfied with the post-purchase experience from using the shopping app
Impulse Buying Intention
I am prompted to buy more because of the discount offer.
If I see something that really interests me I buy it without considering the consequences.
I buy things even though they were not on the shopping list.
I am prompted to buy more because of the faster navigation.
I am prompted to buy more because of attractiveness.
I like to buy because of well-organized layout

25
I am prompted to buy more because of variety of product/choices
Impulse buying
I experienced a number of sudden urges to buy things, which I had not planned to
purchase using this shopping app.
On this shopping app I saw a number of things I wanted to buy even though they were
not on my shopping list.
I experienced no strong urges to make unplanned purchases from the app (r).
While using this shopping app, I felt a sudden urge to buy something
Appendix II
Table I: Demographic Information
Demographic
Category
Percentage
Gender
Male
67.80%
Female
32.20%
Age
Under 21
5.08%
21 to 30
83.47%
31 to 40
9.75%
40 or older
1.69%
Education
Less than Senior high
school
0.85%
Graduation
25.00%
Post- Graduation
61.44%
Professional qualification
12.71%
Occupation
Business
2.54%
Government employee
5.51%
Private employee
31.36%
Student
58.90%
Monthly
family
Income (in
INR)
Less than 50000
71.19%
50001 to 10,0000
15.25%
10,0001 to 150,000
4.66%
More than 150,000
8.47%