- Email: [email protected]
Factors influencing the adoption of mobile commerce applications in Cameroon
Telematics and Informatics xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Telematics and Informatics journal homepage: www.elsevier.com/locate/tele
Factors influencing the adoption of mobile commerce applications in Cameroon Silas Formunyuy Verkijika Department of Computer Science & Informatics, University of the Free State, 205 Nelson Mandela Drive, Bloemfontein, South Africa
A R T IC LE I N F O
ABS TRA CT
Keywords: M-commerce UTAUT2 Cameroon Perceived risk Perceived trust
As smartphone penetration continues to double in Sub-Saharan Africa, many businesses are looking into this channel for conducting their business activities. In Cameroon, all the top ecommerce giants have deployed smartphone applications to facilitate m-commerce activities. However, little is known about the factors that influence m-commerce adoption in the country. As such, this study had as objective to determine the key factors that influence consumer’s adoption of m-commerce applications in Cameroon. Using data from 372 respondents, a modified version of the extended unified theory of acceptance and use of technology (UTAUT2) was validated in the Cameroon context. The findings showed that social influence, facilitating conditions, hedonic motivations, perceived risk and perceived trust were significant predictors of the behavioural intention to adopt m-commerce applications. Also, the results showed that consumers who had a high intention to adopt m-commerce were more likely to recommend the technology to others. For researchers, the study depicts the relevance of extending existing technology acceptance models like the UTUAT2 with appropriate factors in different technological and geographical context. For practitioners, the study identifies customer-specific and environmental factors that m-commerce providers in Cameroon and other regions with similar characteristics could consider when designing and implementing strategies for attracting consumers to use their m-commerce applications.
1. Introduction Over the past two decades, the world has seen significant advances in the area of mobile and wireless communication systems, which has opened a whole new world of possibilities for ubiquitous solutions for improving different areas of our daily activities. One booming area of such solutions has been in the domain of commercial activities. As electronic commerce activities progressed over the years, businesses became increasingly interested in delivering similar services over mobile devices in a bid to reach a wider customer base. This follows from the growing adoption and use of mobile phones around the world with a penetration rate of about 96.4% (International Telecommunications Union, 2014). Many users now have smartphones which can be used for far more than simple voice communications as users can conduct complex activities like electronic payments, shopping, and mobile marketing services (Ko et al., 2009; Zhang et al., 2012). The phenomenon of using mobile devices for business activities is generally termed as mobile commerce or simple m-commerce. M-commerce can be broadly defined as a business model that enables consumers to complete business transactions on a mobile device (Chong, 2013; Zhang et al., 2012). Given the growing nature and scope of m-commerce applications, numerous definitions of m-commerce have emerged to suit the specific context of m-commerce applications. As such, the context of this study will focus
E-mail address: [email protected]. https://doi.org/10.1016/j.tele.2018.04.012 Received 29 September 2017; Received in revised form 21 April 2018; Accepted 21 April 2018 0736-5853/ © 2018 Elsevier Ltd. All rights reserved.
Please cite this article as: Silas Formunyuy, V., Telematics and Informatics (2018), https://doi.org/10.1016/j.tele.2018.04.012
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
mainly on m-commerce applications that facilitate the buying and selling of goods via a mobile device in line with some prior studies (Chong, 2013; Liebana-Cabanillas et al., 2017; Zhang et al., 2012; Sohn, 2017). Albeit m-commerce has been widely seen as an extension of electronic commerce, m-commerce definitely has numerous advantages over electronic commerce especially in its ability to provide localized services, ubiquity, instantaneous services and personalized services (Liebana-Cabanillas et al., 2017; Zhang et al., 2012). Researchers today consider m-commerce as one of the fastest growing business models (Liebana-Cabanillas et al., 2017), however, just a few years ago m-commerce was considered to be in its early stages with little success in many different regions (Alhinai et al., 2010; Zhang et al., 2012). Also, the mobile commerce adoption and development varies significantly across different countries and has been particularly low in developing countries (Zhang et al., 2012). In Cameroon, E-commerce penetration is very low, barely at about 2%. However, mobile phone penetration in the country reached 80% at the end of 2015 according to the country’s Telecommunications Regulatory Board. Internet penetration in the country currently stands at about 21%. Since most people in sub-Saharan Africa access the internet via mobile phones, m-commerce seems to be a logical direction to be taken by e-commerce businesses in the region. As such, it is not surprising to see that almost all ecommerce giants in Cameroon such as Jumia, and Sellam Quick have deployed smartphone apps for carrying out m-commerce activities. However, the download rates of these apps from Google Play store are quite low, suggesting that much still needs to be done to improve user adoption of m-commerce applications in Cameroon. Consequently, identifying factors that influence m-commerce adoption by consumers in Cameroon can be very useful in helping m-commerce service providers in the country to develop better strategies for increasing the uptake of their services by consumers. Recently, researchers have shown that the extended unified theory of acceptance and use of technology (UTAUT2) by Venkatesh et al. (2012) is a valuable model for understanding consumer adoption of mobile applications, especially mobile payment solutions (Alalwan et al., 2017; Oliveira et al., 2016). Since mobile-payments are extremely valuable as a means of paying for goods purchased via m-commerce apps, there is a high likelihood that the same UTAUT2 factors that influence the adoption of m-payment solutions could have a similar influence on m-commerce apps in general. Consequently, this study will use the UTAUT2 as the theoretical framework for evaluating the factors influencing consumers’ adoption of m-commerce in Cameroon. The rest of the paper is structured as follows: Section 2 discusses the proposed model and development of the hypothesis. Section 3 outlines the methodology used in the study. Section 4 presents an analysis of the data and finding thereof. Section 5 presents a discussion on the outcome of the hypothesis, as well as the implications of the study and its apparent limitations. Lastly, section 6 presents the conclusion of the study. 2. Proposed model and development of hypothesis In examining the determinants of technology adoption, researchers have often considered behavioural intention as an important part of understanding actual use behaviour. This follows from the growing empirical evidence that has shown that behavioural intention is one of the best predictors of actual use behaviour (Zhang et al., 2012; Venkatesh et al., 2012). Behavioural intention in the context of m-commerce can be defined as a consumer’s subjective probability of using an m-commerce application, such as an application for buying and selling of goods via a mobile device (Chong, 2013; Liebana-Cabanillas et al., 2017; Sohn, 2017). Most mcommerce studies (Al-Louzi and Iss, 2011; Chong, 2013; Liebana-Cabanillas et al., 2017) have thus focused only on behavioural intention to adopt without examining the post-adoption behaviour, as m-commerce is considered to be in its early stages. Additionally, those that have gone further to evaluate the post-adoption behaviour of m-commerce users have primarily focused on use behaviour (Zhang et al., 2012) without considering other post-adoption behaviours. One of such post-adoption behaviours that have been widely ignored is the intention to recommend m-commerce applications. The intention to recommend a technology is a valuable post-adoption behaviour often ignored by researchers who place emphasis on use behaviour (Miltgen et al., 2013; Oliveira et al., 2016). Recommending a technology such as an m-commerce app to others can have significant commercial benefits, as consumers often adopt technologies proposed to them by their social and work associates (Oliveira et al., 2016). Thus, consumer recommendations can be valuable in increasing the penetration of m-commerce in Cameroon through peer-to-peer recommendations. In addition to the intention to recommend, this study also proposes extending the UTAUT2 with perceived risk and perceived trust as these two factors have shown to be valuable in the context of m-commerce and technology adoption in developing countries (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). The proposed model is presented in Fig. 1. 2.1. UTAUT2 variables 2.1.1. Performance expectancy (PE) Venkatesh et al. (2012) conceptualised performance expectancy as “degree to which using a technology will provide benefits to consumers in performing certain activities.” Performance expectancy, therefore, constitutes the different attributes of information systems that can offer benefits to users. This is quite similar to the perceived usefulness dimension of the Technology Acceptance Model (TAM). The general consensus from prior literature is that individuals will be more inclined to adopt and use a new technology if they believe that the technology will be useful to them (Alalwan et al., 2017; Venkatesh et al., 2012). In the context of m-commerce, performance expectancy will entail the extent to which a consumer perceives that using an m-commerce application can be beneficial in completing their business transactions. However, such a view has received mixed findings with some studies supporting the positive influence of perceived usefulness on intention to adopt m-commerce (Chong, 2013; Faqih and Jaradat, 2015; LiébanaCabanillas et al., 2017), while others found no significant association (Zhang et al., 2012). Nonetheless, since performance expectancy goes beyond simply perceived usefulness to include aspects of relative advantage and extrinsic motivation (Huang and Kao, 2
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Performance Expectancy
Effort Expectancy Social Influence
Perceived Risk
Behavioural Intention to Adopt
Behavioural Intention to Recommend
Facilitating Conditions Hedonic Motivation Perceived Trust
Price Value Fig. 1. Proposed modification of the UTAUT2.
2015), it is possible that the outcome might be different from that of perceived usefulness. For example, Jaradat and Rababaa (2013) showed that performance expectancy was a significant predictor of behavioural intention to adopt m-commerce in Jordan. Also, current evidence shows that performance expectancy plays a significant role in the adoption of mobile payments (Alalwan et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). As such, this study hypothesises that: H1. Performance expectancy positively influences the behavioural intention to adopt m-commerce
2.1.2. Effort expectancy (EE) Effort expectancy is defined as “degree of ease associated with consumers’ use of technology” (Venkatesh et al., 2012). This is similar to the perceived ease of use construct in the TAM, which is the degree to which people believe that using a given system would be used free of effort. In the context of m-commerce, effort expectancy can be described as the ability to complete an m-commerce transaction with minimal effort. In such a situation, consumers will feel very comfortable carrying out m-commerce transactions. In examining m-commerce adoption, several researchers (Zhang et al., 2012; Faqih and Jaradat, 2015) have shown that perceived ease of use (equivalence of EE) is a significant predictor of the behavioural intention to adopt m-commerce solutions with those that are easy to use more likely to be adopted. Similarly, Jaradat and Rababaa (2013) showed that effort expectancy had a significant positive association with the intention to use m-commerce in Jordan. However, recent studies on the adoption of other mobile technologies have failed to find any significant direct association between effort expectancy and behavioural intention (Morosan & DeFranco, 2016; Oliveira et al., 2016). Instead, effort expectancy is seen to have an indirect effect on behavioural intention through its positive influence on performance expectancy (Alalwan et al., 2017; Herero et al., 2017; Oliveira et al., 2016). Nevertheless, since these studies did not focus on m-commerce but other types of mobile technologies and since perceived ease of use (equivalence of EE) have shown to be influential in m-commerce adoption, this study hypothesises that: H2. Effort expectancy positively influences the behavioural intention to adopt m-commerce H3. Effort expectancy has a positive influence the performance expectancy.
2.1.3. Social influence (SI) Social influence refers to the extent to which consumers perceive that significant others (e.g. family and friends) believe they should use a novel technological system (Venkatesh et al., 2012; Oliveira et al., 2016). Social influence suggests that the decisions made by individuals are influenced by their social networks as individuals often take into consideration the opinions of others when deciding on whether or not use a given technology. When such opinions are positive, they will encourage user adoption while the reverse is true for non-adoption. Prior studies (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013) have shown that social influence has a significant positive influence on consumer acceptance of m-commerce solutions. As such, this study hypothesises that: H4. Social influence positively influences the behavioural intention to adopt m-commerce
3
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
2.1.4. Facilitating conditions (FC) Facilitation conditions refer to “the consumers’ perceptions of the resources and support available to perform behaviour” (Venkatesh et al., 2012). In other words, facilitation conditions can be seen as the perceptions of consumers regarding environmental barriers or available resources that ease the use of m-commerce solutions. For example, the cost or availability of mobile internet services in a given region can facilitate or hinder consumer’s use of mobile phone apps for shopping. The initial conceptualisation of the UTAUT considered facilitation conditions as a predictor of only use behaviour; however, Venkatesh et al. (2012) later showed in the UTAUT2 that facilitation conditions also affects behavioural intention to adopt a given technology. However, the universality of this association is questionable as some studies have shown significant outcomes (Lallmahomed et al., 2017; Morosan and DeFranco, 2016) while others have failed to find any significant association of facilitation conditions with behavioural intention (Herero et al., 2017; Oliveira et al., 2016). Despite these mixed findings, his study goes with view of Venkatesh et al. (2012) that facilitation conditions significantly influences behavioural intention. The facilitation conditions that influenced the use of mobile internet in the study of Venkatesh et al. (2012) can very much influence m-commerce as mobile internet is needed for m-commerce transactions. Thus this study posits that: H5. Facilitating conditions positively influences the behavioural intention to adopt m-commerce 2.1.5. Hedonic motivation (HM) Hedonic motivation refers to the pleasure that a consumer derives from using a given technology (Venkatesh et al., 2012). The inclusion of hedonic motivation in the UTAUT2 followed from the notion that consumer adoption of technology was not only based on internal beliefs and the associated utility of the technology, as consumers have been increasingly concern about the overall user experience of the technology. As such, the enjoyment or pleasure a consumer can derive from using a given technology is a very important determinant of the consumer’s acceptance of use of the technology (Morosan & DeFranco, 2016; Venkatesh et al., 2012). Even though Oliveira et al. (2016) did not find support for the association between hedonic motivation and behavioural intention, many other studies have shown support for the significant role of hedonic motivation in predicting intentions to adopt various technologies (Alalwan et al., 2017; Herero et al., 2017; Morosan and DeFranco, 2016; Venkatesh et al., 2012). Additionally, Davis (2010) argued that consumers often use m-commerce services to hedonically experience fun. As such, this study posits that: H6. Hedonic motivation positively influences the behavioural intention to adopt m-commerce 2.1.6. Price value (PV) Venkatesh et al. (2012) define price value as “consumers’ cognitive trade-off between the perceived benefits of the applications and the monetary cost for using them” This is particularly important in the adoption of consumer technology as consumers often bear the cost of using the technology. Generally, price value is expected to influence behavioural intention when the perceived benefits of using the technology is higher the monetary cost of using it (Venkatesh et al., 2012). In the context of m-commerce, this can be seen as the benefits of using m-commerce applications being higher than the monetary cost of conducting similar transactions in a face-toface setting or other means. For example, the mobile internet cost required to run m-commerce shopping apps can be high for consumers as such apps load many images of products on sale thus using more data than the average smartphone app. Consequently, consumers might be forced to reconsider whether such cost is justifiable given the potential benefits of using the m-commerce shopping app. Since the introduction of price value in the UTAUT2, some researchers (Alalwan et al., 2017; Arenas-Gaitán et al., 2015; Lallmahomed et al., 2017; Venkatesh et al., 2012) have shown a significant association between PV and behavioural intention. However, others (Macedo, 2017; Oliveira et al., 2016) found that price value was not significant in predicting the behavioural intention. Despite these mixed findings, this study sides with the view that price value will significantly influence behavioural intention as prior studies (Wei et al., 2009; Zhang et al., 2012) on m-commerce have shown that the perceived cost of using mcommerce has a significant negative impact on consumer’s adoption of the technology. This suggests that when costs are perceived to be high, consumers are less likely to adopt m-commerce. Consequently, it can be suggested if the benefits outweigh the perceived cost of using the m-commerce application, consumers are more likely to adopt the m-commerce solution. Therefore, this study posits that: H7. Price value positively influences the behavioural intention to adopt m-commerce 2.2. Extending the UTUAT2 2.2.1. Perceived risk (PR) Perceived risk refers to an individual’s perceptions regarding the risks liked with using a given technology. Such risk can include financial, psychological, social, physical or time risk (Zhang et al., 2012). Existing research suggests that the risks perceived by consumers regarding using internet technologies contribute significantly to their restraint in adopting electronic systems (Dwivedi et al., 2017). Over 80% of internet users worry about having their personal details on the web (Rana et al., 2015). This is not surprising given the increasing trends of consumer information being stolen company information systems and leaked online or sold in the black market. As such, in the context of m-commerce, it is expected that if consumers perceived the risks of using an mcommerce application to be high, the likelihood of adopting such a technology will be low. Zhang et al. (2012) showed that perceived risk had a significant negative influence on behavioural intention to adopt m-commerce. This is quite expected, as consumers are less likely to pay for goods and services via mobile devices if they perceive the risks to be high (Slade et al., 2015). Therefore, it was 4
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
hypothesized that: H8. Perceived risk has a negative and significant influence on the intention to adopt m-commerce
2.2.2. Perceived trust (PT) Zhang et al. (2012) define perceived trust in the context of m-commerce as “the extent to which an individual believes that using m-commerce is secure and has no privacy threats.” Given that m-commerce is not well understood by many customers, the issues of trust are often raised by consumers (Liebana-Cabanillas et al., 2017). Prior literature has shown that trust has a significant influence on consumer’s intention to adopt m-commerce solutions (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). When perceived trust is high, consumers are more likely to adopt an m-commerce solution and they are convinced that their personal identifiable information and other sensitive data (e.g. payment information) can be trust with the m-commerce service provider. Therefore, it was hypothesized that: H9. Perceived trust has a positive and significant influence on the intention to adopt m-commerce
2.3. Users’ intention to recommend m-commerce applications The intention to recommend a given m-commerce solution depicts the willingness of a user or potential user to put forward with approval the suitability of a given m-commerce application for use by others. Prior research has increasingly shown that word-ofmouth plays a vital role in swaying consumers towards a given product (Miltgen et al., 2013). Recommending a solution is an important post-adoption behaviour that can be beneficial in increasing the diffusion of a new technology. With today’s growth in social network communications, the views expressed by users about a given product can be valuable in assisting consumers in making adoption decisions. For example, when downloading an m-commerce app from the Google Play Store, a potential user has the opportunity to evaluate existing user comments and appraisal of the app which can influence their decision on whether or not to download and test the app. Prior literature has shown that those who have a high intention to use a given product are often more likely to recommend it to others (Miltgen et al., 2013; Oliveira et al., 2016).Therefore, it was hypothesized that: H10. Behavioural intention to adopt m-commerce positively influences the intention to recommend m-commerce applications to others
3. Methodology Albeit the UTAUT2 is a widely validated model, its constructs have shown mixed findings across different context (Alalwan et al., 2017; Oliveira et al., 2016). As such, it is increasingly important to modify and validate the UTAUT2 to increase the generalizability of its findings. This is particularly important in the context of m-commerce where different studies have shown mixed outcomes for the same variables (Chong, 2013; Faqih & Jaradat, 2015; Liébana-Cabanillas et al., 2017; Zhang et al., 2012). As such, there is a need to increasingly validate the hypothesized relationships associated with the proposed model in a manner that allows for the generalizability of the findings. In this regard, a quantitative research methodology was adopted as the most suitable approach for the present study. Although researchers (Chan and Ngai, 2007; Vogelsang et al., 2013; Wu, 2012) have increasingly called for a need to include qualitative data collection and analysis in the study of technology adoption, a key weakness of qualitative research still remains the issues associated with the generalizability of the findings (Houser, 2015, Wu, 2012). Consequently, the need for generalizability and the hypothetico-deductive logic approach adopted in the study limited the suitability of qualitative analysis in the present study. The measurement scales used for capturing data on the variables in the proposed model were adapted from prior literature. All items were measured on a five-point Likert scale anchored by 1(strongly disagree) to 5 (strongly agree). Measurements for the core UTAUT2 variables (PE, EE, SI, FC, HM and PV) were adapted from Venkatesh et al. (2012) and properly worded to suit the technology used in the study (i.e. m-commerce adoption). The scale used for measuring perceived risk was adapted from Slade et al. (2015) while that used for measuring perceived trust was adapted from Liebana-Cabanillas et al. (2017). The questionnaire also included background information such as age (measured in years), gender (measured as a dummy variable with 0 for female and 1 otherwise), and formal education (measured as the highest attained formal qualification of the respondent). The questionnaire was pilot tested with 36 subjects. The outcome of the pilot test indicated that the questions were reliable and valid. As such, no revisions were made on the questionnaire items. The data from the pilot test was not used in the main study to minimize any possibility of skewing the data. In order to obtain the empirical data to validate the model and test the hypothesis, quantitative questionnaires were issued to 600 individuals above 18 years of age in four towns in Cameroon (Bamenda, Buea, Douala and Yaoundé). Out of the 600 issued questionnaires, 372 fully completed valid responses were obtained (67.6% valid response rate). The sample comprised of civil servants (30.1%), private sector employees (35.4%), small business owners (15.9) and students (18.6%). All the participants were owners of a smartphone device and thus are potential users of m-commerce applications. The demographic information of the respondents is presented in Table 1 below. Most of the respondents were female (51.1%). Also, the majority were in the age group of 21 to 30 (38.7%) and 31–40 years (31.2%) respectively. Lastly, over 70.4% of the respondents had at least an undergraduate degree. 5
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 1 Descriptive statistics. Demographic information Gender Male Female Age Up to 20 years 21–30 years 31–40 years Above 40 years Education High school diploma or below Higher education diploma Undergraduate degree Postgraduate (above degree)
# 182 190 # 69 144 116 43 # 30 80 160 102
% 48.9 51.1 % 18.5 38.7 31.2 11.6 % 8.1 21.5 43.0 27.4
Note: # is the frequency, while % is the percentage.
4. Data analysis and results 4.1. Measurement model The model was tested using Smart PLS 3.0 (Ringle et al., 2015). Prior to testing the hypothesis, several quality criteria were used to evaluate the reliability (construct and indicator) and validity (convergent and discriminant) of the proposed model. Table 2 indicates the quality criteria used to evaluate the constructs which include Cronbach’s alpha, average variance extracted (AVE), composite reliability (CR) and factor loadings. Indicator reliability was assessed using factor loadings based on the popular view that all item should have loadings above 0.7 and that loadings below 0.4 should be eliminated (Hair et al., 2010; Henseler et al., 2009). The range of the loadings clearly shows that all items loaded above 0.7, with the lowest loading being 0.756. As such, no item was dropped and the instrument was considered to demonstrate appropriate levels of indicator reliability. Construct reliability was tested using Cronbach’s alpha and composite reliability. For construct reliability to be confirmed, Cronbach’s alpha values should be above 0.7, while composite reliability values should be preferably above 0.8 although values above 0.6 are acceptable (Henseler et al., 2009; Liebana-Cabanillas et al., 2017). Looking at Table 1it is observed that Cronbach’s alpha values ranged from 0.743 to 979 while those for composite reliability ranged from 0.847 to 0.983. As such both criteria for Cronbach’s alpha and composite reliability were met, therefore, confirming the reliability of the constructs. Convergent validity was assessed using the AVE, based on the criteria that AVE values should be above 0.5 so that the latent variable is able to explain more than 50% of the variance of its indicators (Hair et al., 2010; Oliveira et al., 2016). The data in Table 2 indicates that AVE values ranged from 0.649 to 0.949, thus satisfying the criteria for convergent validity. Discriminant validity was evaluated using the Fornell-larcker criterion presented in Table 3. The Fornell-Larcker criterion suggests that a construct has discriminant validity if the square-root of the AVE is greater than the paired inter-correlation between the latent constructs (Fornell and Larcker, 1981; Henseler et al., 2009). In Table 3, the diagonal values in bold are the square roots of the AVE while the off-diagonal values are the inter-correlations between the latent constructs. The results show that all the diagonal values are greater than the corresponding off-diagonal values, thus satisfying the FornellLarcker criterion. This, therefore, confirms the discriminant validity of the scales used. 4.2. Structural model Fig. 2 shows the outcome of the PLS-SEM estimations. SEM model fitness in Smart PLS is evaluated using Standard Root Mean Table 2 Quality criterion (Alpha, AVE, and CR) and loadings. Construct
Cronbach’s Alpha
AVE
CR
Loadings (Range)
BI EE FC HM IR PE PR PT SI
0.784 0.958 0.939 0.973 0.743 0.880 0.979 0.947 0.823
0.698 0.889 0.846 0.949 0.649 0.806 0.904 0.862 0.712
0.874 0.970 0.957 0.983 0.847 0.925 0.983 0.961 0.881
0.777 0.908 0.890 0.972 0.756 0.834 0.933 0.905 0.797
6
to to to to to to to to to
0.826 0.965 0.953 0.985 0.893 0.938 0.965 0.950 0.901
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 3 Fornell-Larcker Criterion: Correlation matrix of constructs and square root of AVE (in bold).
BI EE FC HM IR PE PR PT PV SI
BI
EE
FC
HM
IR
PE
PR
PT
PV
SI
0.806 0.145 0.247 −0.114 0.487 0.169 −0.175 0.408 0.196 0.270
0.943 0.662 −0.047 0.180 0.401 −0.074 0.145 0.389 0.279
0.920 −0.052 0.272 0.448 −0.064 0.224 0.417 0.357
0.974 −0.216 −0.030 0.053 −0.044 0.067 −0.105
0.836 0.148 −0.029 0.195 0.134 0.175
0.898 −0.118 0.154 0.323 0.350
0.951 0.176 −0.023 0.076
0.928 0.163 0.204
0.988 0.401
0.844
Fig. 2. Structural model results.
Square Residual (SMRM) and Normed Fit Index (NFI). Generally, an SMRM of 0 indicates perfectly fit model while a value less than 0.05 indicates a good fit (Byrne, 2008). However, researchers have also shown that the cut-off for the SMRM should be 0.08 as correctly specified models have shown to have SMRM values above 0.06 (Henseler et al., 2016). However, in the case of the tested model, the SMRM value was 0.047 which is less than 0.05, thus depicting a good fitness of the model. With regards to the NFI, a model with good fitness needs to have an NFI value of at least 0.9 (Alalwan et al., 2017). This was also supported in this study as the NFI value for the model was 0.961. The model was therefore concluded to have indicated a good level of fitness. The model explains 43.2% of the consumer’s intention to adopt m-commerce applications. Out of the 8 variables proposed as predictors of m-commerce adoption 5 (social influence, facilitating conditions, hedonic motivation, perceived risk and perceived rust) were significant while 3 (performance expectancy, effort expectancy, price value) were not significant. For the significant variables, PT (β = 0.329; p < 0.001) had the highest influence of BI followed by SI (β = 0.130; p < 0.05), FC (β = 0.127; p < 0.05), PR (β = −0.111; p < 0.05), and HM (β = 0.103; p < 0.05) respectively. The model also explains 35.4% of the behavioural intention to recommend m-commerce apps, with a significant path showing that the intention to adopt m-commerce is highly associated with the intention to recommend it to others (β = 0.487; p < 0.001). Lastly, it was observed that effort expectancy explains 24.6% of the variance in performance expectancy with a significant association between the two factors (β = 0.401; p < 0.001). The associated hypotheses developed from the proposed model are presented below (Table 4). 5. Discussion 5.1. Outcome of hypotheses The hypothesized relationships were evaluated using the structural model presented in Fig. 2. A total of 10 hypotheses were 7
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 4 Outcome of hypothesis. No
Hypothesis
Path
T-Value
Significance
Supported
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10
PE → BI EE → BI EE → PE SI → BI FC → BI HM → BI PV → BI PR → BI PT → BI BI → IR
0.005 −0.063 0.401 0.130 0.127 0.103 0.068 −0.111 0.329 0.487
0.087 0.848 8.297 2.376 2.263 2.068 1.266 −2.116 6.437 7.851
0.931ns 0.397ns 0.000** 0.018* 0.024* 0.039* 0.206ns 0.035* 0.000** 0.000**
No No Yes Yes Yes Yes No Yes Yes Yes
Note: ns = not significant;
**
p < 0.001; *p < 0.05.
developed in this study. Based on the findings, it 7 hypothesis (H1, H2 and H7) were supported while 3 (H3, H4, H5, H6, H8, H9, H9) were not supported. Each of these hypotheses is discussed below. Hypothesis H1 suggested a positive and significant association between performance expectancy and behavioural intention to adopt m-commerce. However, this study failed to find support for this association. The findings are contrary to other studies that have shown that performance expectancy is an important factor in m-commerce adoption (Jaradat and Rababaa, 2013) and the adoption of other mobile technologies (Alalwan et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). Nevertheless, given the overlap between performance expectancy and the perceived usefulness dimension of the TAM, it is not surprising to find that the universality of the association between performance expectancy and behavioural intention to adopt m-commerce might be uncertain, as findings for the impact of perceived usefulness on the adoption of m-commerce have been mixed (Chong, 2013; Faqih and Jaradat, 2015; Liébana-Cabanillas et al., 2017; Zhang et al., 2012). Similarly to hypothesis H1, the findings also failed to support H2 as effort expectancy did not have a significant association with the behavioural intention to adopt m-commerce. The findings are contrary to that of Jaradat and Rababaa (2013) who showed that effort expectancy was significant in predicting the intention to adopt m-commerce in Jordan. Nonetheless, the non-significant outcome of effort expectancy in this study is not surprising as many recent studies have shown that effort expectancy is not an important prediction of intention to adopt many different kinds of technologies (Herero et al., 2017; Lallmahomed et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). This study, however, found support for the association between effort expectancy and performance expectancy thus supporting hypothesis H3, which is in line with prior studies (Alalwan et al., 2017; Herero et al., 2017; Oliveira et al., 2016). Also, the study supported hypotheses H4, H5 and H6 indicated that social influence, facilitating conditions and hedonic motivation positively influence the behavioural intention to adopt m-commerce. The outcome of social influence was expected as it is strongly in line with prior studies on m-commerce (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013). With respect to hedonic motivation, albeit there are limited or no evidence of its association with m-commerce, its association with the intention to adopt other technologies is mixed (Morosan & DeFranco, 2016; Oliveira et al., 2016). As such, this study sides with the previous evidence supporting the important role of hedonic motivation on technology adoption (Alalwan et al., 2017; Herero et al., 2017; Venkatesh et al., 2012). Moreover, the findings support the assertion by Davis (2010) that m-commerce users like to use the technology to hedonically experience fun. Similar to hedonic motivation, prior research has shown mixed findings for the impact of FC in technology adoption (Herero et al., 2017; Lallmahomed et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). However, this study now finds support for the importance of facilitating conditions in the context of m-commerce. This study found no support for the association of price value and behavioural intention thus hypothesis H7 was not supported. The association between price value and behavioural intention has received mixed findings (Macedo, 2017; Oliveira et al., 2016). Although in the context of m-commerce, perceived cost has been shown to influence behavioural intention, this study indicates that the same is not true for price values despite the closeness between perceived cost and price value. The addition of perceived risk and perceived trust were seen to play a vital role in influencing the behavioural intention to adopt m-commerce thus supporting hypothesis H8 and H9. Perceived trust was found to have the highest influence on behavioural intention, thus supporting the need for including trust as an added factor in the UTAUT2. The outcome of hypothesis H8 and H9 is congruent with prior studies that have also shown the significant role played by perceived risk and perceived trust in influencing the intentions to adopt m-commerce (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). Lastly, the study showed that the behavioural intention to adopt m-commerce was significantly associated with the intention to recommend m-commerce thus supporting hypothesis H10. This finding is in line with prior studies that have also shown that the users with an intention to use a given technology are more likely to recommend the technology to others (Miltgen et al., 2013; Oliveira et al., 2016). 5.2. Practical implications This study provided several implications for practice. As seen from the results of this study and the growing literature on mcommerce, it is increasingly evident that different factors influence m-commerce adoption in different settings and cultures and thus 8
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
necessitates the need for country-specific studies on m-commerce adoption (Liebana-Cabanillas et al., 2017; Zhang et al., 2012). Many m-commerce providers operate across different countries and it is important for country operations to adopt appropriate strategies for promoting m-commerce solutions. In the case of m-commerce in Cameroon, this study identified two key categories that m-commerce provides must focus on. These categories are customer-specific factors and environmental factors. In the customer-specific domain, the key factors for m-commerce providers to focus on are perceived trust, perceived risk and hedonic motivation. Researchers have increasingly shown that because customers know little about m-commerce, they tend to have many questions about trust (Liebana-Cabanillas et al., 2017). It is imperative for m-commerce providers in Cameroon to develop commercial campaigns that centre on gaining the trust of the customers. M-commerce firms should have open and truthful communications with existing and potential customers to create an atmosphere of transparency that can develop customer trust in their services. It was recently observed in Cameroon that when ECOBANK closed in bank branches in favour of focusing its efforts on digital banking, the population at large started spreading the message that the bank was closing due to possible liquidation until the bank had to openly communicate its plans more effectively through an official press release. Also, as suggested by Liebana-Cabanillas et al. (2017), m-commerce firms can use adequate refund policies and money back guarantees to gain customer trust. Closely tied with perceived trust is the perceived risk that existing and potential customers of m-commerce in Cameroon have. Some risk associated with m-commerce can include fraud, identity theft, quality of products and unjustifiable delays in product delivery (Zhang et al., 2012). M-commerce firms should assure the market about their commitment to quality and protection of their sensitive information as most consumers worry about their sensitive information on the web (Rana et al., 2015). It was also observed that the enjoyment customers get from using m-commerce solutions (hedonic motivation) increases their intention to adopt m-commerce. Mcommerce firms, therefore, need to advertise the fun features of their solutions. Also, a focus should be placed on creating mcommerce applications with user-friendly interfaces as this can increase users overall hedonic experience with the applications. In the environmental domain, the important factors for m-commerce acceptance in Cameroon are social influence and facilitating conditions. With respect to social influence, m-commerce firms in Cameroon need to develop strategies that take advantage of the social milieu of their consumers. For example, m-commerce providers can exploit the power of social media to promote the use of their services while advising their existing customers to share their advertisements to their networks. Also, by capitalising on convincing existing m-commerce customer to recommend the technology to others, it can be possible for m-commerce firms to easily reach a critical mass. Also, m-commerce providers should encourage customer reviews of their services on public platforms. This can be achieved by especially targeting extroverted customers as they can be easily convinced to share and promote positive information and recommendations of the m-commerce solution (Morosan and DeFranco, 2016). With respect to facilitating conditions, Mcommerce suppliers should focus on aspects they can easily control such as improving customers knowledge on how to use their mcommerce applications while making them easy enough for them to use. Additionally, it is vital to create a 24/7 customer support helpline to guide potential and existing customers during their use of the m-commerce applications. 5.3. Theoretical implications The UTAUT2 as a theoretical model has been widely argued to be an effective model for understanding the factors that influence technology adoption from a consumer’s perspective. However, extant research has shown that the universality of the association between UTAUT2 variables and behavioural intention is questionable as many studies have shown mixed findings with different types of technologies and different implementation context (Alalwan et al., 2017; Herero et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016; Venkatesh et al., 2012). As a result, researchers have increasingly shown that the effectiveness of the UTAUT2 model can be more effective when other factors are introduced (Alalwan et al., 2017; Lallmahomed et al., 2017; Oliveira et al., 2016). In the context of m-commerce, the UTAUT2 model has not been expanded with perceived risk and perceived trust even though these factors have gained prominence in prior m-commerce literature. This study makes a theoretical contribution by showing that adding perceived risk and perceived trust to the current UTAUT2 model improves its effectiveness in predicting the behavioural intention to use m-commerce both factors were significantly associated with behavioural intention to adopt m-commerce when used to extend the UTAUT2. The study also contributed to the limited literature the advocates the need for considering the intention to recommend as a vital post-adoption behaviour that can benefit businesses. 5.4. Limitations of the study Even though this study presents a noteworthy attempt an understanding the phenomenon of m-commerce adoption in Cameroon, the study also has some limitations that need to be acknowledged. Firstly, the study did not include habit in the UTAUT2 model which is a dimension of the UATUT2 that has been known to influence adoption behaviour (Herero et al., 2017; Morosan and DeFranco, 2016; Venkatesh et al., 2012). However, researchers (Alalwan et al., 2017; Oliveira et al., 2016) have cautioned that when respondents have not had enough experience with a given technology, it can be difficult to effectively evaluate the influence of habit. As such, since m-commerce adoption is still low in Cameroon and most respondents had little experience with the technology, it was deemed necessary not to include habit. However, as many people continue to adopt m-commerce. Future studies can include habit to evaluate its influence on adoption and also the intention recommend. Secondly, the Venkatesh et al. (2012) argued that the UTAUT2 factors were moderated by Age, gender and experience. However, this study did not evaluate the moderating effects of these factors. This is another avenue for future studies in Cameroon to evaluate the moderating effects. Also, since this study used cross-sectional analysis with data gathered at a single point in time, it is important to note that consumer attitudes can change over time. As such, a longitudinal study should be considered in future studies as it can provide better insights. Lastly, the study did not make use of 9
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
qualitative data collection and analysis. Several researchers have emphasized the need to include qualitative analysis when studying technology adoption in order to provide further insights on the phenomenon (Vogelsang et al., 2013; Wu, 2012). In fact, the combination of qualitative and quantitative approaches in a mixed method design can significantly enhance human understanding of technology adoption behaviour (Wu, 2012). As such, future studies can use mixed methods to further investigate the factors that influence mobile commerce adoption in different settings. 6. Conclusion M-commerce adoption is a growing research domain that has gained the interest of many researchers across different countries (Chong, 2013; Liebana-Cabanillas et al., 2017; Zhang et al., 2012). However, the factors influencing the behavioural intention to adopt m-commerce solutions have not been comprehensively understood, as prior studies have shown mixed findings across different countries (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013; Zhang et al., 2012). In the context of Cameroon, little is known about the antecedents of m-commerce adoption. In order to fill this gap, this study had as objective to evaluate the factors influencing the behavioural intention to adopt m-commerce in Cameroon as well as the intention to recommend its use to others. The UTAUT2 model was selected as the appropriate theoretical framework for the study following its wide usage in studying technology adoption from a consumer perspective. The model was further extended with perceived risk and perceived trust to ensure that key factors that could influence the intention to use m-commerce were not left out. The study found that out of that social influence, facilitating conditions, hedonic motivations, perceived risk and perceived trust were significant predictors of behavioural intention while performance expectancy and effort expectancy were not. The results also showed a significant association between the behavioural intention to adopt m-commerce and the intention to recommend it to others. This study contributes to the growing literature on m-commerce and the extensions of the UTAUT2 model while also providing practical guidance for m-commerce providers in Cameroon on how to improve user adoption of their systems. References Alalwan, A.A., Dwivedi, Y.K., Rana, N.P., 2017. Factors influencing adoption of mobile banking by Jordanian bank customers: extending UTAUT2 with trust. Int. J. Inf. Manage. 37, 99–110. Alhinai, Y.S., Kurnia, S., Smith, S.P., 2010. The adoption of mobile commerce services by individuals: a current state of the literature. Retrieved from:. Pacific Asia Conference on Information Systems Proceedings. Al-Louzi, B., Iss, B., 2011. Factors influencing customer acceptance of m-commerce services in Jordan. J. Commun. Comput. 9, 1424–1436. Arenas-Gaitán, J., Peral-Peral, B., Ramón-Jerónimo, M.A., 2015. Elderly and Internet Banking: an application of UTAUT2. J. Internet Banking Commerce 20 (1), 1–23. Byrne, B.M., 2008. Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming. Psychology Press, New York, NY. Chan, S.C., Ngai, E.W., 2007. A qualitative study of information technology adoption: how ten organizations adopted Web-based training? Inf. Syst. J. 17 (3), 289–315. Chong, A.Y., 2013. Predicting m-commerce adoption determinants: A neural network approach. Expert Syst. Appl. 40, 523–530. Davis, R., 2010. Conceptualising fun in mobile commerce environments. Int. J. Mobile Commun. 8 (1), 21–40. Dwivedi, Y.K., Rana, N.P., Janssen, M., Lal, B., Williams, M., Clement, M., 2017. An empirical validation of a unified model of electronic government adoption (UMEGA). Gov. Inform. Q. 34 (2), 211–230. Faqih, K.M., Jaradat, M.R., 2015. Assessing the moderating effect of gender differences and individualism-collectivism at individual-level on the adoption of mobile commerce technology: TAM3 perspective. J. Retailing Consum. Serv. 22, 37–52. Fornell, C., Larcker, D., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 18 (1), 39–50. Hair, J.F., Black, W.C., Barbin, B.J., Anderson, R.E., 2010. Multivariate Data Analysis, 6th ed. Prentice Hall, New Jersey. Henseler, J., Hubona, G., Ray, P., 2016. Using PLS path modelling in new technology research: updated guidelines. Ind. Manage. Data Syst. 116 (1), 2–20. Henseler, J., Ringle, C.M., Sinkovics, R.R., 2009. The use of partial least squares path modelling in international marketing. Adv. Int. Marketing 20, 277–319. Herero, A., Martin, H.S., Salmones, M.G., 2017. Explaining the adoption of social networks sites for sharing user-generated content: a revision of the UTAUT2. Comput. Hum. Behav. 71, 209–217. Houser, R.A., 2015. Counselling and Educational Research: Evaluation and Application. SAGE Publications Inc, Thousand Oaks, CA. Huang, C., Kao, Y., 2015. UTAUT2 Based Predictions of Factors Influencing the Technology Acceptance of Phablets by DNP. Math. Problems Eng Article ID 603747. International Telecommunications Union. (2014). Measuring the information society report. Retrieved from: https://www.itu.int/en/ITU-D/Statistics/Documents/publications// MIS2014_without_Annex_4.pdf. Jaradat, M., Rababaa, M., 2013. Assessing key factor that influence on the acceptance of mobile commerce based on modified UTAUT. Int. J. Bus. Manag. 8 (23), 102–112. Ko, E., Kim, E.Y., Lee, E.K., 2009. Modelling consumer adoption if mobile shopping for fashion products in Korea. Psychol. Marketing 26, 669–687. Lallmahomed, M.Z., Lallmahomed, N., Lallmahomed, G.M., 2017. Factors influencing the adoption of e-Government services in Mauritius. Telematics Inf. 34 (4), 57–72. Liebana-Cabanillas, F., Marinkovic, V., Kalinic, Z., 2017. A SEM-neural network approach for predicting antecedents of m-commerce acceptance. Int. J. Inf. Manage. 37, 14–24. Macedo, I.M., 2017. Predicting the acceptance and use of information and communication technology by older adults: an empirical examination of the revised UTAUT2. Comput. Hum. Behav. 75, 935–948. Miltgen, C.L., Popovič, A., Oliveira, T., 2013. Determinants of end-user acceptance of biometrics: integrating the “Big 3” of technology acceptance with privacy context. Decis. Support Syst. 56, 103–114. Morosan, C., DeFranco, A., 2016. It’s about time: Revisiting UTAUT2 to examine consumers’ intentions to use NFC mobile payments in hotels. Int. J. Hospitality Manag. 53, 17–29. Oliveira, T., Thomas, M., Baptista, G., Campos, F., 2016. Mobile payment: Understanding the determinants of customer adoption and intention to recommend the technology. Comput. Hum. Behav. 61, 404–414. Rana, N.P., Dwivedi, Y.K., Williams, M.D., Weerakkody, V., 2015. Investigating success of an e-government initiative: Validation of an integrated IS success model. Inf. Syst. Front. 17 (1), 127–142. Ringle, C.M., Wende, S., Becker, J.M. 2015. SmartPLS 3. Boenningstedt: SmartPLS GmbH. Slade, E., Williams, M., Dwivedi, Y., Piercy, N., 2015. Exploring customer adoption of proximity mobile payments. J. Strategic Marketing 23 (3), 209–223. Sohn, S., 2017. A contextual perspective on consumers’ perceived usefulness: the case of mobile online shopping. J. Retailing Consum. Serv. 38, 22–33. Venkatesh, V., Thong, J., Xu, X., 2012. Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology. MIS Quarterly 36 (1), 157–178. Vogelsang, K., Steinhüser, M., Hoppe, U. 2013. A Qualitative Approach to Examine Technology Acceptance. Retrieved from: https://aisel.aisnet.org/cgi/viewcontent.cgi? article=1183&context=icis2013. Wei, T.T., Marthandan, G., Chong, L.Y., Ooi, K., Arumugam, S., 2009. What drives Malaysian m-commerce adoption? an empirical analysis. Ind. Manag. Data Syst. 109 (3), 370–388. Wu, P.F., 2012. A mixed methods approach to technology acceptance research. J. Assoc. Inf. Syst. 13 (3), 172–187. Zhang, L., Zhu, J., Liu, Q., 2012. A meta-analysis of mobile commerce adoption and the moderating effect of culture. Comput. Human Behav. 28, 1902–1911.
10
Contents lists available at ScienceDirect
Telematics and Informatics journal homepage: www.elsevier.com/locate/tele
Factors influencing the adoption of mobile commerce applications in Cameroon Silas Formunyuy Verkijika Department of Computer Science & Informatics, University of the Free State, 205 Nelson Mandela Drive, Bloemfontein, South Africa
A R T IC LE I N F O
ABS TRA CT
Keywords: M-commerce UTAUT2 Cameroon Perceived risk Perceived trust
As smartphone penetration continues to double in Sub-Saharan Africa, many businesses are looking into this channel for conducting their business activities. In Cameroon, all the top ecommerce giants have deployed smartphone applications to facilitate m-commerce activities. However, little is known about the factors that influence m-commerce adoption in the country. As such, this study had as objective to determine the key factors that influence consumer’s adoption of m-commerce applications in Cameroon. Using data from 372 respondents, a modified version of the extended unified theory of acceptance and use of technology (UTAUT2) was validated in the Cameroon context. The findings showed that social influence, facilitating conditions, hedonic motivations, perceived risk and perceived trust were significant predictors of the behavioural intention to adopt m-commerce applications. Also, the results showed that consumers who had a high intention to adopt m-commerce were more likely to recommend the technology to others. For researchers, the study depicts the relevance of extending existing technology acceptance models like the UTUAT2 with appropriate factors in different technological and geographical context. For practitioners, the study identifies customer-specific and environmental factors that m-commerce providers in Cameroon and other regions with similar characteristics could consider when designing and implementing strategies for attracting consumers to use their m-commerce applications.
1. Introduction Over the past two decades, the world has seen significant advances in the area of mobile and wireless communication systems, which has opened a whole new world of possibilities for ubiquitous solutions for improving different areas of our daily activities. One booming area of such solutions has been in the domain of commercial activities. As electronic commerce activities progressed over the years, businesses became increasingly interested in delivering similar services over mobile devices in a bid to reach a wider customer base. This follows from the growing adoption and use of mobile phones around the world with a penetration rate of about 96.4% (International Telecommunications Union, 2014). Many users now have smartphones which can be used for far more than simple voice communications as users can conduct complex activities like electronic payments, shopping, and mobile marketing services (Ko et al., 2009; Zhang et al., 2012). The phenomenon of using mobile devices for business activities is generally termed as mobile commerce or simple m-commerce. M-commerce can be broadly defined as a business model that enables consumers to complete business transactions on a mobile device (Chong, 2013; Zhang et al., 2012). Given the growing nature and scope of m-commerce applications, numerous definitions of m-commerce have emerged to suit the specific context of m-commerce applications. As such, the context of this study will focus
E-mail address: [email protected]. https://doi.org/10.1016/j.tele.2018.04.012 Received 29 September 2017; Received in revised form 21 April 2018; Accepted 21 April 2018 0736-5853/ © 2018 Elsevier Ltd. All rights reserved.
Please cite this article as: Silas Formunyuy, V., Telematics and Informatics (2018), https://doi.org/10.1016/j.tele.2018.04.012
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
mainly on m-commerce applications that facilitate the buying and selling of goods via a mobile device in line with some prior studies (Chong, 2013; Liebana-Cabanillas et al., 2017; Zhang et al., 2012; Sohn, 2017). Albeit m-commerce has been widely seen as an extension of electronic commerce, m-commerce definitely has numerous advantages over electronic commerce especially in its ability to provide localized services, ubiquity, instantaneous services and personalized services (Liebana-Cabanillas et al., 2017; Zhang et al., 2012). Researchers today consider m-commerce as one of the fastest growing business models (Liebana-Cabanillas et al., 2017), however, just a few years ago m-commerce was considered to be in its early stages with little success in many different regions (Alhinai et al., 2010; Zhang et al., 2012). Also, the mobile commerce adoption and development varies significantly across different countries and has been particularly low in developing countries (Zhang et al., 2012). In Cameroon, E-commerce penetration is very low, barely at about 2%. However, mobile phone penetration in the country reached 80% at the end of 2015 according to the country’s Telecommunications Regulatory Board. Internet penetration in the country currently stands at about 21%. Since most people in sub-Saharan Africa access the internet via mobile phones, m-commerce seems to be a logical direction to be taken by e-commerce businesses in the region. As such, it is not surprising to see that almost all ecommerce giants in Cameroon such as Jumia, and Sellam Quick have deployed smartphone apps for carrying out m-commerce activities. However, the download rates of these apps from Google Play store are quite low, suggesting that much still needs to be done to improve user adoption of m-commerce applications in Cameroon. Consequently, identifying factors that influence m-commerce adoption by consumers in Cameroon can be very useful in helping m-commerce service providers in the country to develop better strategies for increasing the uptake of their services by consumers. Recently, researchers have shown that the extended unified theory of acceptance and use of technology (UTAUT2) by Venkatesh et al. (2012) is a valuable model for understanding consumer adoption of mobile applications, especially mobile payment solutions (Alalwan et al., 2017; Oliveira et al., 2016). Since mobile-payments are extremely valuable as a means of paying for goods purchased via m-commerce apps, there is a high likelihood that the same UTAUT2 factors that influence the adoption of m-payment solutions could have a similar influence on m-commerce apps in general. Consequently, this study will use the UTAUT2 as the theoretical framework for evaluating the factors influencing consumers’ adoption of m-commerce in Cameroon. The rest of the paper is structured as follows: Section 2 discusses the proposed model and development of the hypothesis. Section 3 outlines the methodology used in the study. Section 4 presents an analysis of the data and finding thereof. Section 5 presents a discussion on the outcome of the hypothesis, as well as the implications of the study and its apparent limitations. Lastly, section 6 presents the conclusion of the study. 2. Proposed model and development of hypothesis In examining the determinants of technology adoption, researchers have often considered behavioural intention as an important part of understanding actual use behaviour. This follows from the growing empirical evidence that has shown that behavioural intention is one of the best predictors of actual use behaviour (Zhang et al., 2012; Venkatesh et al., 2012). Behavioural intention in the context of m-commerce can be defined as a consumer’s subjective probability of using an m-commerce application, such as an application for buying and selling of goods via a mobile device (Chong, 2013; Liebana-Cabanillas et al., 2017; Sohn, 2017). Most mcommerce studies (Al-Louzi and Iss, 2011; Chong, 2013; Liebana-Cabanillas et al., 2017) have thus focused only on behavioural intention to adopt without examining the post-adoption behaviour, as m-commerce is considered to be in its early stages. Additionally, those that have gone further to evaluate the post-adoption behaviour of m-commerce users have primarily focused on use behaviour (Zhang et al., 2012) without considering other post-adoption behaviours. One of such post-adoption behaviours that have been widely ignored is the intention to recommend m-commerce applications. The intention to recommend a technology is a valuable post-adoption behaviour often ignored by researchers who place emphasis on use behaviour (Miltgen et al., 2013; Oliveira et al., 2016). Recommending a technology such as an m-commerce app to others can have significant commercial benefits, as consumers often adopt technologies proposed to them by their social and work associates (Oliveira et al., 2016). Thus, consumer recommendations can be valuable in increasing the penetration of m-commerce in Cameroon through peer-to-peer recommendations. In addition to the intention to recommend, this study also proposes extending the UTAUT2 with perceived risk and perceived trust as these two factors have shown to be valuable in the context of m-commerce and technology adoption in developing countries (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). The proposed model is presented in Fig. 1. 2.1. UTAUT2 variables 2.1.1. Performance expectancy (PE) Venkatesh et al. (2012) conceptualised performance expectancy as “degree to which using a technology will provide benefits to consumers in performing certain activities.” Performance expectancy, therefore, constitutes the different attributes of information systems that can offer benefits to users. This is quite similar to the perceived usefulness dimension of the Technology Acceptance Model (TAM). The general consensus from prior literature is that individuals will be more inclined to adopt and use a new technology if they believe that the technology will be useful to them (Alalwan et al., 2017; Venkatesh et al., 2012). In the context of m-commerce, performance expectancy will entail the extent to which a consumer perceives that using an m-commerce application can be beneficial in completing their business transactions. However, such a view has received mixed findings with some studies supporting the positive influence of perceived usefulness on intention to adopt m-commerce (Chong, 2013; Faqih and Jaradat, 2015; LiébanaCabanillas et al., 2017), while others found no significant association (Zhang et al., 2012). Nonetheless, since performance expectancy goes beyond simply perceived usefulness to include aspects of relative advantage and extrinsic motivation (Huang and Kao, 2
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Performance Expectancy
Effort Expectancy Social Influence
Perceived Risk
Behavioural Intention to Adopt
Behavioural Intention to Recommend
Facilitating Conditions Hedonic Motivation Perceived Trust
Price Value Fig. 1. Proposed modification of the UTAUT2.
2015), it is possible that the outcome might be different from that of perceived usefulness. For example, Jaradat and Rababaa (2013) showed that performance expectancy was a significant predictor of behavioural intention to adopt m-commerce in Jordan. Also, current evidence shows that performance expectancy plays a significant role in the adoption of mobile payments (Alalwan et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). As such, this study hypothesises that: H1. Performance expectancy positively influences the behavioural intention to adopt m-commerce
2.1.2. Effort expectancy (EE) Effort expectancy is defined as “degree of ease associated with consumers’ use of technology” (Venkatesh et al., 2012). This is similar to the perceived ease of use construct in the TAM, which is the degree to which people believe that using a given system would be used free of effort. In the context of m-commerce, effort expectancy can be described as the ability to complete an m-commerce transaction with minimal effort. In such a situation, consumers will feel very comfortable carrying out m-commerce transactions. In examining m-commerce adoption, several researchers (Zhang et al., 2012; Faqih and Jaradat, 2015) have shown that perceived ease of use (equivalence of EE) is a significant predictor of the behavioural intention to adopt m-commerce solutions with those that are easy to use more likely to be adopted. Similarly, Jaradat and Rababaa (2013) showed that effort expectancy had a significant positive association with the intention to use m-commerce in Jordan. However, recent studies on the adoption of other mobile technologies have failed to find any significant direct association between effort expectancy and behavioural intention (Morosan & DeFranco, 2016; Oliveira et al., 2016). Instead, effort expectancy is seen to have an indirect effect on behavioural intention through its positive influence on performance expectancy (Alalwan et al., 2017; Herero et al., 2017; Oliveira et al., 2016). Nevertheless, since these studies did not focus on m-commerce but other types of mobile technologies and since perceived ease of use (equivalence of EE) have shown to be influential in m-commerce adoption, this study hypothesises that: H2. Effort expectancy positively influences the behavioural intention to adopt m-commerce H3. Effort expectancy has a positive influence the performance expectancy.
2.1.3. Social influence (SI) Social influence refers to the extent to which consumers perceive that significant others (e.g. family and friends) believe they should use a novel technological system (Venkatesh et al., 2012; Oliveira et al., 2016). Social influence suggests that the decisions made by individuals are influenced by their social networks as individuals often take into consideration the opinions of others when deciding on whether or not use a given technology. When such opinions are positive, they will encourage user adoption while the reverse is true for non-adoption. Prior studies (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013) have shown that social influence has a significant positive influence on consumer acceptance of m-commerce solutions. As such, this study hypothesises that: H4. Social influence positively influences the behavioural intention to adopt m-commerce
3
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
2.1.4. Facilitating conditions (FC) Facilitation conditions refer to “the consumers’ perceptions of the resources and support available to perform behaviour” (Venkatesh et al., 2012). In other words, facilitation conditions can be seen as the perceptions of consumers regarding environmental barriers or available resources that ease the use of m-commerce solutions. For example, the cost or availability of mobile internet services in a given region can facilitate or hinder consumer’s use of mobile phone apps for shopping. The initial conceptualisation of the UTAUT considered facilitation conditions as a predictor of only use behaviour; however, Venkatesh et al. (2012) later showed in the UTAUT2 that facilitation conditions also affects behavioural intention to adopt a given technology. However, the universality of this association is questionable as some studies have shown significant outcomes (Lallmahomed et al., 2017; Morosan and DeFranco, 2016) while others have failed to find any significant association of facilitation conditions with behavioural intention (Herero et al., 2017; Oliveira et al., 2016). Despite these mixed findings, his study goes with view of Venkatesh et al. (2012) that facilitation conditions significantly influences behavioural intention. The facilitation conditions that influenced the use of mobile internet in the study of Venkatesh et al. (2012) can very much influence m-commerce as mobile internet is needed for m-commerce transactions. Thus this study posits that: H5. Facilitating conditions positively influences the behavioural intention to adopt m-commerce 2.1.5. Hedonic motivation (HM) Hedonic motivation refers to the pleasure that a consumer derives from using a given technology (Venkatesh et al., 2012). The inclusion of hedonic motivation in the UTAUT2 followed from the notion that consumer adoption of technology was not only based on internal beliefs and the associated utility of the technology, as consumers have been increasingly concern about the overall user experience of the technology. As such, the enjoyment or pleasure a consumer can derive from using a given technology is a very important determinant of the consumer’s acceptance of use of the technology (Morosan & DeFranco, 2016; Venkatesh et al., 2012). Even though Oliveira et al. (2016) did not find support for the association between hedonic motivation and behavioural intention, many other studies have shown support for the significant role of hedonic motivation in predicting intentions to adopt various technologies (Alalwan et al., 2017; Herero et al., 2017; Morosan and DeFranco, 2016; Venkatesh et al., 2012). Additionally, Davis (2010) argued that consumers often use m-commerce services to hedonically experience fun. As such, this study posits that: H6. Hedonic motivation positively influences the behavioural intention to adopt m-commerce 2.1.6. Price value (PV) Venkatesh et al. (2012) define price value as “consumers’ cognitive trade-off between the perceived benefits of the applications and the monetary cost for using them” This is particularly important in the adoption of consumer technology as consumers often bear the cost of using the technology. Generally, price value is expected to influence behavioural intention when the perceived benefits of using the technology is higher the monetary cost of using it (Venkatesh et al., 2012). In the context of m-commerce, this can be seen as the benefits of using m-commerce applications being higher than the monetary cost of conducting similar transactions in a face-toface setting or other means. For example, the mobile internet cost required to run m-commerce shopping apps can be high for consumers as such apps load many images of products on sale thus using more data than the average smartphone app. Consequently, consumers might be forced to reconsider whether such cost is justifiable given the potential benefits of using the m-commerce shopping app. Since the introduction of price value in the UTAUT2, some researchers (Alalwan et al., 2017; Arenas-Gaitán et al., 2015; Lallmahomed et al., 2017; Venkatesh et al., 2012) have shown a significant association between PV and behavioural intention. However, others (Macedo, 2017; Oliveira et al., 2016) found that price value was not significant in predicting the behavioural intention. Despite these mixed findings, this study sides with the view that price value will significantly influence behavioural intention as prior studies (Wei et al., 2009; Zhang et al., 2012) on m-commerce have shown that the perceived cost of using mcommerce has a significant negative impact on consumer’s adoption of the technology. This suggests that when costs are perceived to be high, consumers are less likely to adopt m-commerce. Consequently, it can be suggested if the benefits outweigh the perceived cost of using the m-commerce application, consumers are more likely to adopt the m-commerce solution. Therefore, this study posits that: H7. Price value positively influences the behavioural intention to adopt m-commerce 2.2. Extending the UTUAT2 2.2.1. Perceived risk (PR) Perceived risk refers to an individual’s perceptions regarding the risks liked with using a given technology. Such risk can include financial, psychological, social, physical or time risk (Zhang et al., 2012). Existing research suggests that the risks perceived by consumers regarding using internet technologies contribute significantly to their restraint in adopting electronic systems (Dwivedi et al., 2017). Over 80% of internet users worry about having their personal details on the web (Rana et al., 2015). This is not surprising given the increasing trends of consumer information being stolen company information systems and leaked online or sold in the black market. As such, in the context of m-commerce, it is expected that if consumers perceived the risks of using an mcommerce application to be high, the likelihood of adopting such a technology will be low. Zhang et al. (2012) showed that perceived risk had a significant negative influence on behavioural intention to adopt m-commerce. This is quite expected, as consumers are less likely to pay for goods and services via mobile devices if they perceive the risks to be high (Slade et al., 2015). Therefore, it was 4
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
hypothesized that: H8. Perceived risk has a negative and significant influence on the intention to adopt m-commerce
2.2.2. Perceived trust (PT) Zhang et al. (2012) define perceived trust in the context of m-commerce as “the extent to which an individual believes that using m-commerce is secure and has no privacy threats.” Given that m-commerce is not well understood by many customers, the issues of trust are often raised by consumers (Liebana-Cabanillas et al., 2017). Prior literature has shown that trust has a significant influence on consumer’s intention to adopt m-commerce solutions (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). When perceived trust is high, consumers are more likely to adopt an m-commerce solution and they are convinced that their personal identifiable information and other sensitive data (e.g. payment information) can be trust with the m-commerce service provider. Therefore, it was hypothesized that: H9. Perceived trust has a positive and significant influence on the intention to adopt m-commerce
2.3. Users’ intention to recommend m-commerce applications The intention to recommend a given m-commerce solution depicts the willingness of a user or potential user to put forward with approval the suitability of a given m-commerce application for use by others. Prior research has increasingly shown that word-ofmouth plays a vital role in swaying consumers towards a given product (Miltgen et al., 2013). Recommending a solution is an important post-adoption behaviour that can be beneficial in increasing the diffusion of a new technology. With today’s growth in social network communications, the views expressed by users about a given product can be valuable in assisting consumers in making adoption decisions. For example, when downloading an m-commerce app from the Google Play Store, a potential user has the opportunity to evaluate existing user comments and appraisal of the app which can influence their decision on whether or not to download and test the app. Prior literature has shown that those who have a high intention to use a given product are often more likely to recommend it to others (Miltgen et al., 2013; Oliveira et al., 2016).Therefore, it was hypothesized that: H10. Behavioural intention to adopt m-commerce positively influences the intention to recommend m-commerce applications to others
3. Methodology Albeit the UTAUT2 is a widely validated model, its constructs have shown mixed findings across different context (Alalwan et al., 2017; Oliveira et al., 2016). As such, it is increasingly important to modify and validate the UTAUT2 to increase the generalizability of its findings. This is particularly important in the context of m-commerce where different studies have shown mixed outcomes for the same variables (Chong, 2013; Faqih & Jaradat, 2015; Liébana-Cabanillas et al., 2017; Zhang et al., 2012). As such, there is a need to increasingly validate the hypothesized relationships associated with the proposed model in a manner that allows for the generalizability of the findings. In this regard, a quantitative research methodology was adopted as the most suitable approach for the present study. Although researchers (Chan and Ngai, 2007; Vogelsang et al., 2013; Wu, 2012) have increasingly called for a need to include qualitative data collection and analysis in the study of technology adoption, a key weakness of qualitative research still remains the issues associated with the generalizability of the findings (Houser, 2015, Wu, 2012). Consequently, the need for generalizability and the hypothetico-deductive logic approach adopted in the study limited the suitability of qualitative analysis in the present study. The measurement scales used for capturing data on the variables in the proposed model were adapted from prior literature. All items were measured on a five-point Likert scale anchored by 1(strongly disagree) to 5 (strongly agree). Measurements for the core UTAUT2 variables (PE, EE, SI, FC, HM and PV) were adapted from Venkatesh et al. (2012) and properly worded to suit the technology used in the study (i.e. m-commerce adoption). The scale used for measuring perceived risk was adapted from Slade et al. (2015) while that used for measuring perceived trust was adapted from Liebana-Cabanillas et al. (2017). The questionnaire also included background information such as age (measured in years), gender (measured as a dummy variable with 0 for female and 1 otherwise), and formal education (measured as the highest attained formal qualification of the respondent). The questionnaire was pilot tested with 36 subjects. The outcome of the pilot test indicated that the questions were reliable and valid. As such, no revisions were made on the questionnaire items. The data from the pilot test was not used in the main study to minimize any possibility of skewing the data. In order to obtain the empirical data to validate the model and test the hypothesis, quantitative questionnaires were issued to 600 individuals above 18 years of age in four towns in Cameroon (Bamenda, Buea, Douala and Yaoundé). Out of the 600 issued questionnaires, 372 fully completed valid responses were obtained (67.6% valid response rate). The sample comprised of civil servants (30.1%), private sector employees (35.4%), small business owners (15.9) and students (18.6%). All the participants were owners of a smartphone device and thus are potential users of m-commerce applications. The demographic information of the respondents is presented in Table 1 below. Most of the respondents were female (51.1%). Also, the majority were in the age group of 21 to 30 (38.7%) and 31–40 years (31.2%) respectively. Lastly, over 70.4% of the respondents had at least an undergraduate degree. 5
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 1 Descriptive statistics. Demographic information Gender Male Female Age Up to 20 years 21–30 years 31–40 years Above 40 years Education High school diploma or below Higher education diploma Undergraduate degree Postgraduate (above degree)
# 182 190 # 69 144 116 43 # 30 80 160 102
% 48.9 51.1 % 18.5 38.7 31.2 11.6 % 8.1 21.5 43.0 27.4
Note: # is the frequency, while % is the percentage.
4. Data analysis and results 4.1. Measurement model The model was tested using Smart PLS 3.0 (Ringle et al., 2015). Prior to testing the hypothesis, several quality criteria were used to evaluate the reliability (construct and indicator) and validity (convergent and discriminant) of the proposed model. Table 2 indicates the quality criteria used to evaluate the constructs which include Cronbach’s alpha, average variance extracted (AVE), composite reliability (CR) and factor loadings. Indicator reliability was assessed using factor loadings based on the popular view that all item should have loadings above 0.7 and that loadings below 0.4 should be eliminated (Hair et al., 2010; Henseler et al., 2009). The range of the loadings clearly shows that all items loaded above 0.7, with the lowest loading being 0.756. As such, no item was dropped and the instrument was considered to demonstrate appropriate levels of indicator reliability. Construct reliability was tested using Cronbach’s alpha and composite reliability. For construct reliability to be confirmed, Cronbach’s alpha values should be above 0.7, while composite reliability values should be preferably above 0.8 although values above 0.6 are acceptable (Henseler et al., 2009; Liebana-Cabanillas et al., 2017). Looking at Table 1it is observed that Cronbach’s alpha values ranged from 0.743 to 979 while those for composite reliability ranged from 0.847 to 0.983. As such both criteria for Cronbach’s alpha and composite reliability were met, therefore, confirming the reliability of the constructs. Convergent validity was assessed using the AVE, based on the criteria that AVE values should be above 0.5 so that the latent variable is able to explain more than 50% of the variance of its indicators (Hair et al., 2010; Oliveira et al., 2016). The data in Table 2 indicates that AVE values ranged from 0.649 to 0.949, thus satisfying the criteria for convergent validity. Discriminant validity was evaluated using the Fornell-larcker criterion presented in Table 3. The Fornell-Larcker criterion suggests that a construct has discriminant validity if the square-root of the AVE is greater than the paired inter-correlation between the latent constructs (Fornell and Larcker, 1981; Henseler et al., 2009). In Table 3, the diagonal values in bold are the square roots of the AVE while the off-diagonal values are the inter-correlations between the latent constructs. The results show that all the diagonal values are greater than the corresponding off-diagonal values, thus satisfying the FornellLarcker criterion. This, therefore, confirms the discriminant validity of the scales used. 4.2. Structural model Fig. 2 shows the outcome of the PLS-SEM estimations. SEM model fitness in Smart PLS is evaluated using Standard Root Mean Table 2 Quality criterion (Alpha, AVE, and CR) and loadings. Construct
Cronbach’s Alpha
AVE
CR
Loadings (Range)
BI EE FC HM IR PE PR PT SI
0.784 0.958 0.939 0.973 0.743 0.880 0.979 0.947 0.823
0.698 0.889 0.846 0.949 0.649 0.806 0.904 0.862 0.712
0.874 0.970 0.957 0.983 0.847 0.925 0.983 0.961 0.881
0.777 0.908 0.890 0.972 0.756 0.834 0.933 0.905 0.797
6
to to to to to to to to to
0.826 0.965 0.953 0.985 0.893 0.938 0.965 0.950 0.901
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 3 Fornell-Larcker Criterion: Correlation matrix of constructs and square root of AVE (in bold).
BI EE FC HM IR PE PR PT PV SI
BI
EE
FC
HM
IR
PE
PR
PT
PV
SI
0.806 0.145 0.247 −0.114 0.487 0.169 −0.175 0.408 0.196 0.270
0.943 0.662 −0.047 0.180 0.401 −0.074 0.145 0.389 0.279
0.920 −0.052 0.272 0.448 −0.064 0.224 0.417 0.357
0.974 −0.216 −0.030 0.053 −0.044 0.067 −0.105
0.836 0.148 −0.029 0.195 0.134 0.175
0.898 −0.118 0.154 0.323 0.350
0.951 0.176 −0.023 0.076
0.928 0.163 0.204
0.988 0.401
0.844
Fig. 2. Structural model results.
Square Residual (SMRM) and Normed Fit Index (NFI). Generally, an SMRM of 0 indicates perfectly fit model while a value less than 0.05 indicates a good fit (Byrne, 2008). However, researchers have also shown that the cut-off for the SMRM should be 0.08 as correctly specified models have shown to have SMRM values above 0.06 (Henseler et al., 2016). However, in the case of the tested model, the SMRM value was 0.047 which is less than 0.05, thus depicting a good fitness of the model. With regards to the NFI, a model with good fitness needs to have an NFI value of at least 0.9 (Alalwan et al., 2017). This was also supported in this study as the NFI value for the model was 0.961. The model was therefore concluded to have indicated a good level of fitness. The model explains 43.2% of the consumer’s intention to adopt m-commerce applications. Out of the 8 variables proposed as predictors of m-commerce adoption 5 (social influence, facilitating conditions, hedonic motivation, perceived risk and perceived rust) were significant while 3 (performance expectancy, effort expectancy, price value) were not significant. For the significant variables, PT (β = 0.329; p < 0.001) had the highest influence of BI followed by SI (β = 0.130; p < 0.05), FC (β = 0.127; p < 0.05), PR (β = −0.111; p < 0.05), and HM (β = 0.103; p < 0.05) respectively. The model also explains 35.4% of the behavioural intention to recommend m-commerce apps, with a significant path showing that the intention to adopt m-commerce is highly associated with the intention to recommend it to others (β = 0.487; p < 0.001). Lastly, it was observed that effort expectancy explains 24.6% of the variance in performance expectancy with a significant association between the two factors (β = 0.401; p < 0.001). The associated hypotheses developed from the proposed model are presented below (Table 4). 5. Discussion 5.1. Outcome of hypotheses The hypothesized relationships were evaluated using the structural model presented in Fig. 2. A total of 10 hypotheses were 7
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
Table 4 Outcome of hypothesis. No
Hypothesis
Path
T-Value
Significance
Supported
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10
PE → BI EE → BI EE → PE SI → BI FC → BI HM → BI PV → BI PR → BI PT → BI BI → IR
0.005 −0.063 0.401 0.130 0.127 0.103 0.068 −0.111 0.329 0.487
0.087 0.848 8.297 2.376 2.263 2.068 1.266 −2.116 6.437 7.851
0.931ns 0.397ns 0.000** 0.018* 0.024* 0.039* 0.206ns 0.035* 0.000** 0.000**
No No Yes Yes Yes Yes No Yes Yes Yes
Note: ns = not significant;
**
p < 0.001; *p < 0.05.
developed in this study. Based on the findings, it 7 hypothesis (H1, H2 and H7) were supported while 3 (H3, H4, H5, H6, H8, H9, H9) were not supported. Each of these hypotheses is discussed below. Hypothesis H1 suggested a positive and significant association between performance expectancy and behavioural intention to adopt m-commerce. However, this study failed to find support for this association. The findings are contrary to other studies that have shown that performance expectancy is an important factor in m-commerce adoption (Jaradat and Rababaa, 2013) and the adoption of other mobile technologies (Alalwan et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). Nevertheless, given the overlap between performance expectancy and the perceived usefulness dimension of the TAM, it is not surprising to find that the universality of the association between performance expectancy and behavioural intention to adopt m-commerce might be uncertain, as findings for the impact of perceived usefulness on the adoption of m-commerce have been mixed (Chong, 2013; Faqih and Jaradat, 2015; Liébana-Cabanillas et al., 2017; Zhang et al., 2012). Similarly to hypothesis H1, the findings also failed to support H2 as effort expectancy did not have a significant association with the behavioural intention to adopt m-commerce. The findings are contrary to that of Jaradat and Rababaa (2013) who showed that effort expectancy was significant in predicting the intention to adopt m-commerce in Jordan. Nonetheless, the non-significant outcome of effort expectancy in this study is not surprising as many recent studies have shown that effort expectancy is not an important prediction of intention to adopt many different kinds of technologies (Herero et al., 2017; Lallmahomed et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). This study, however, found support for the association between effort expectancy and performance expectancy thus supporting hypothesis H3, which is in line with prior studies (Alalwan et al., 2017; Herero et al., 2017; Oliveira et al., 2016). Also, the study supported hypotheses H4, H5 and H6 indicated that social influence, facilitating conditions and hedonic motivation positively influence the behavioural intention to adopt m-commerce. The outcome of social influence was expected as it is strongly in line with prior studies on m-commerce (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013). With respect to hedonic motivation, albeit there are limited or no evidence of its association with m-commerce, its association with the intention to adopt other technologies is mixed (Morosan & DeFranco, 2016; Oliveira et al., 2016). As such, this study sides with the previous evidence supporting the important role of hedonic motivation on technology adoption (Alalwan et al., 2017; Herero et al., 2017; Venkatesh et al., 2012). Moreover, the findings support the assertion by Davis (2010) that m-commerce users like to use the technology to hedonically experience fun. Similar to hedonic motivation, prior research has shown mixed findings for the impact of FC in technology adoption (Herero et al., 2017; Lallmahomed et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016). However, this study now finds support for the importance of facilitating conditions in the context of m-commerce. This study found no support for the association of price value and behavioural intention thus hypothesis H7 was not supported. The association between price value and behavioural intention has received mixed findings (Macedo, 2017; Oliveira et al., 2016). Although in the context of m-commerce, perceived cost has been shown to influence behavioural intention, this study indicates that the same is not true for price values despite the closeness between perceived cost and price value. The addition of perceived risk and perceived trust were seen to play a vital role in influencing the behavioural intention to adopt m-commerce thus supporting hypothesis H8 and H9. Perceived trust was found to have the highest influence on behavioural intention, thus supporting the need for including trust as an added factor in the UTAUT2. The outcome of hypothesis H8 and H9 is congruent with prior studies that have also shown the significant role played by perceived risk and perceived trust in influencing the intentions to adopt m-commerce (Liebana-Cabanillas et al., 2017; Wei et al., 2009; Zhang et al., 2012). Lastly, the study showed that the behavioural intention to adopt m-commerce was significantly associated with the intention to recommend m-commerce thus supporting hypothesis H10. This finding is in line with prior studies that have also shown that the users with an intention to use a given technology are more likely to recommend the technology to others (Miltgen et al., 2013; Oliveira et al., 2016). 5.2. Practical implications This study provided several implications for practice. As seen from the results of this study and the growing literature on mcommerce, it is increasingly evident that different factors influence m-commerce adoption in different settings and cultures and thus 8
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
necessitates the need for country-specific studies on m-commerce adoption (Liebana-Cabanillas et al., 2017; Zhang et al., 2012). Many m-commerce providers operate across different countries and it is important for country operations to adopt appropriate strategies for promoting m-commerce solutions. In the case of m-commerce in Cameroon, this study identified two key categories that m-commerce provides must focus on. These categories are customer-specific factors and environmental factors. In the customer-specific domain, the key factors for m-commerce providers to focus on are perceived trust, perceived risk and hedonic motivation. Researchers have increasingly shown that because customers know little about m-commerce, they tend to have many questions about trust (Liebana-Cabanillas et al., 2017). It is imperative for m-commerce providers in Cameroon to develop commercial campaigns that centre on gaining the trust of the customers. M-commerce firms should have open and truthful communications with existing and potential customers to create an atmosphere of transparency that can develop customer trust in their services. It was recently observed in Cameroon that when ECOBANK closed in bank branches in favour of focusing its efforts on digital banking, the population at large started spreading the message that the bank was closing due to possible liquidation until the bank had to openly communicate its plans more effectively through an official press release. Also, as suggested by Liebana-Cabanillas et al. (2017), m-commerce firms can use adequate refund policies and money back guarantees to gain customer trust. Closely tied with perceived trust is the perceived risk that existing and potential customers of m-commerce in Cameroon have. Some risk associated with m-commerce can include fraud, identity theft, quality of products and unjustifiable delays in product delivery (Zhang et al., 2012). M-commerce firms should assure the market about their commitment to quality and protection of their sensitive information as most consumers worry about their sensitive information on the web (Rana et al., 2015). It was also observed that the enjoyment customers get from using m-commerce solutions (hedonic motivation) increases their intention to adopt m-commerce. Mcommerce firms, therefore, need to advertise the fun features of their solutions. Also, a focus should be placed on creating mcommerce applications with user-friendly interfaces as this can increase users overall hedonic experience with the applications. In the environmental domain, the important factors for m-commerce acceptance in Cameroon are social influence and facilitating conditions. With respect to social influence, m-commerce firms in Cameroon need to develop strategies that take advantage of the social milieu of their consumers. For example, m-commerce providers can exploit the power of social media to promote the use of their services while advising their existing customers to share their advertisements to their networks. Also, by capitalising on convincing existing m-commerce customer to recommend the technology to others, it can be possible for m-commerce firms to easily reach a critical mass. Also, m-commerce providers should encourage customer reviews of their services on public platforms. This can be achieved by especially targeting extroverted customers as they can be easily convinced to share and promote positive information and recommendations of the m-commerce solution (Morosan and DeFranco, 2016). With respect to facilitating conditions, Mcommerce suppliers should focus on aspects they can easily control such as improving customers knowledge on how to use their mcommerce applications while making them easy enough for them to use. Additionally, it is vital to create a 24/7 customer support helpline to guide potential and existing customers during their use of the m-commerce applications. 5.3. Theoretical implications The UTAUT2 as a theoretical model has been widely argued to be an effective model for understanding the factors that influence technology adoption from a consumer’s perspective. However, extant research has shown that the universality of the association between UTAUT2 variables and behavioural intention is questionable as many studies have shown mixed findings with different types of technologies and different implementation context (Alalwan et al., 2017; Herero et al., 2017; Morosan and DeFranco, 2016; Oliveira et al., 2016; Venkatesh et al., 2012). As a result, researchers have increasingly shown that the effectiveness of the UTAUT2 model can be more effective when other factors are introduced (Alalwan et al., 2017; Lallmahomed et al., 2017; Oliveira et al., 2016). In the context of m-commerce, the UTAUT2 model has not been expanded with perceived risk and perceived trust even though these factors have gained prominence in prior m-commerce literature. This study makes a theoretical contribution by showing that adding perceived risk and perceived trust to the current UTAUT2 model improves its effectiveness in predicting the behavioural intention to use m-commerce both factors were significantly associated with behavioural intention to adopt m-commerce when used to extend the UTAUT2. The study also contributed to the limited literature the advocates the need for considering the intention to recommend as a vital post-adoption behaviour that can benefit businesses. 5.4. Limitations of the study Even though this study presents a noteworthy attempt an understanding the phenomenon of m-commerce adoption in Cameroon, the study also has some limitations that need to be acknowledged. Firstly, the study did not include habit in the UTAUT2 model which is a dimension of the UATUT2 that has been known to influence adoption behaviour (Herero et al., 2017; Morosan and DeFranco, 2016; Venkatesh et al., 2012). However, researchers (Alalwan et al., 2017; Oliveira et al., 2016) have cautioned that when respondents have not had enough experience with a given technology, it can be difficult to effectively evaluate the influence of habit. As such, since m-commerce adoption is still low in Cameroon and most respondents had little experience with the technology, it was deemed necessary not to include habit. However, as many people continue to adopt m-commerce. Future studies can include habit to evaluate its influence on adoption and also the intention recommend. Secondly, the Venkatesh et al. (2012) argued that the UTAUT2 factors were moderated by Age, gender and experience. However, this study did not evaluate the moderating effects of these factors. This is another avenue for future studies in Cameroon to evaluate the moderating effects. Also, since this study used cross-sectional analysis with data gathered at a single point in time, it is important to note that consumer attitudes can change over time. As such, a longitudinal study should be considered in future studies as it can provide better insights. Lastly, the study did not make use of 9
Telematics and Informatics xxx (xxxx) xxx–xxx
S.F. Verkijika
qualitative data collection and analysis. Several researchers have emphasized the need to include qualitative analysis when studying technology adoption in order to provide further insights on the phenomenon (Vogelsang et al., 2013; Wu, 2012). In fact, the combination of qualitative and quantitative approaches in a mixed method design can significantly enhance human understanding of technology adoption behaviour (Wu, 2012). As such, future studies can use mixed methods to further investigate the factors that influence mobile commerce adoption in different settings. 6. Conclusion M-commerce adoption is a growing research domain that has gained the interest of many researchers across different countries (Chong, 2013; Liebana-Cabanillas et al., 2017; Zhang et al., 2012). However, the factors influencing the behavioural intention to adopt m-commerce solutions have not been comprehensively understood, as prior studies have shown mixed findings across different countries (Al-Louzi and Iss, 2011; Jaradat and Rababaa, 2013; Zhang et al., 2012). In the context of Cameroon, little is known about the antecedents of m-commerce adoption. In order to fill this gap, this study had as objective to evaluate the factors influencing the behavioural intention to adopt m-commerce in Cameroon as well as the intention to recommend its use to others. The UTAUT2 model was selected as the appropriate theoretical framework for the study following its wide usage in studying technology adoption from a consumer perspective. The model was further extended with perceived risk and perceived trust to ensure that key factors that could influence the intention to use m-commerce were not left out. The study found that out of that social influence, facilitating conditions, hedonic motivations, perceived risk and perceived trust were significant predictors of behavioural intention while performance expectancy and effort expectancy were not. The results also showed a significant association between the behavioural intention to adopt m-commerce and the intention to recommend it to others. This study contributes to the growing literature on m-commerce and the extensions of the UTAUT2 model while also providing practical guidance for m-commerce providers in Cameroon on how to improve user adoption of their systems. References Alalwan, A.A., Dwivedi, Y.K., Rana, N.P., 2017. Factors influencing adoption of mobile banking by Jordanian bank customers: extending UTAUT2 with trust. Int. J. Inf. Manage. 37, 99–110. Alhinai, Y.S., Kurnia, S., Smith, S.P., 2010. The adoption of mobile commerce services by individuals: a current state of the literature. Retrieved from:. Pacific Asia Conference on Information Systems Proceedings. Al-Louzi, B., Iss, B., 2011. Factors influencing customer acceptance of m-commerce services in Jordan. J. Commun. Comput. 9, 1424–1436. Arenas-Gaitán, J., Peral-Peral, B., Ramón-Jerónimo, M.A., 2015. Elderly and Internet Banking: an application of UTAUT2. J. Internet Banking Commerce 20 (1), 1–23. Byrne, B.M., 2008. Structural Equation Modeling with EQS: Basic Concepts, Applications, and Programming. Psychology Press, New York, NY. Chan, S.C., Ngai, E.W., 2007. A qualitative study of information technology adoption: how ten organizations adopted Web-based training? Inf. Syst. J. 17 (3), 289–315. Chong, A.Y., 2013. Predicting m-commerce adoption determinants: A neural network approach. Expert Syst. Appl. 40, 523–530. Davis, R., 2010. Conceptualising fun in mobile commerce environments. Int. J. Mobile Commun. 8 (1), 21–40. Dwivedi, Y.K., Rana, N.P., Janssen, M., Lal, B., Williams, M., Clement, M., 2017. An empirical validation of a unified model of electronic government adoption (UMEGA). Gov. Inform. Q. 34 (2), 211–230. Faqih, K.M., Jaradat, M.R., 2015. Assessing the moderating effect of gender differences and individualism-collectivism at individual-level on the adoption of mobile commerce technology: TAM3 perspective. J. Retailing Consum. Serv. 22, 37–52. Fornell, C., Larcker, D., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Mark. Res. 18 (1), 39–50. Hair, J.F., Black, W.C., Barbin, B.J., Anderson, R.E., 2010. Multivariate Data Analysis, 6th ed. Prentice Hall, New Jersey. Henseler, J., Hubona, G., Ray, P., 2016. Using PLS path modelling in new technology research: updated guidelines. Ind. Manage. Data Syst. 116 (1), 2–20. Henseler, J., Ringle, C.M., Sinkovics, R.R., 2009. The use of partial least squares path modelling in international marketing. Adv. Int. Marketing 20, 277–319. Herero, A., Martin, H.S., Salmones, M.G., 2017. Explaining the adoption of social networks sites for sharing user-generated content: a revision of the UTAUT2. Comput. Hum. Behav. 71, 209–217. Houser, R.A., 2015. Counselling and Educational Research: Evaluation and Application. SAGE Publications Inc, Thousand Oaks, CA. Huang, C., Kao, Y., 2015. UTAUT2 Based Predictions of Factors Influencing the Technology Acceptance of Phablets by DNP. Math. Problems Eng Article ID 603747. International Telecommunications Union. (2014). Measuring the information society report. Retrieved from: https://www.itu.int/en/ITU-D/Statistics/Documents/publications// MIS2014_without_Annex_4.pdf. Jaradat, M., Rababaa, M., 2013. Assessing key factor that influence on the acceptance of mobile commerce based on modified UTAUT. Int. J. Bus. Manag. 8 (23), 102–112. Ko, E., Kim, E.Y., Lee, E.K., 2009. Modelling consumer adoption if mobile shopping for fashion products in Korea. Psychol. Marketing 26, 669–687. Lallmahomed, M.Z., Lallmahomed, N., Lallmahomed, G.M., 2017. Factors influencing the adoption of e-Government services in Mauritius. Telematics Inf. 34 (4), 57–72. Liebana-Cabanillas, F., Marinkovic, V., Kalinic, Z., 2017. A SEM-neural network approach for predicting antecedents of m-commerce acceptance. Int. J. Inf. Manage. 37, 14–24. Macedo, I.M., 2017. Predicting the acceptance and use of information and communication technology by older adults: an empirical examination of the revised UTAUT2. Comput. Hum. Behav. 75, 935–948. Miltgen, C.L., Popovič, A., Oliveira, T., 2013. Determinants of end-user acceptance of biometrics: integrating the “Big 3” of technology acceptance with privacy context. Decis. Support Syst. 56, 103–114. Morosan, C., DeFranco, A., 2016. It’s about time: Revisiting UTAUT2 to examine consumers’ intentions to use NFC mobile payments in hotels. Int. J. Hospitality Manag. 53, 17–29. Oliveira, T., Thomas, M., Baptista, G., Campos, F., 2016. Mobile payment: Understanding the determinants of customer adoption and intention to recommend the technology. Comput. Hum. Behav. 61, 404–414. Rana, N.P., Dwivedi, Y.K., Williams, M.D., Weerakkody, V., 2015. Investigating success of an e-government initiative: Validation of an integrated IS success model. Inf. Syst. Front. 17 (1), 127–142. Ringle, C.M., Wende, S., Becker, J.M. 2015. SmartPLS 3. Boenningstedt: SmartPLS GmbH. Slade, E., Williams, M., Dwivedi, Y., Piercy, N., 2015. Exploring customer adoption of proximity mobile payments. J. Strategic Marketing 23 (3), 209–223. Sohn, S., 2017. A contextual perspective on consumers’ perceived usefulness: the case of mobile online shopping. J. Retailing Consum. Serv. 38, 22–33. Venkatesh, V., Thong, J., Xu, X., 2012. Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology. MIS Quarterly 36 (1), 157–178. Vogelsang, K., Steinhüser, M., Hoppe, U. 2013. A Qualitative Approach to Examine Technology Acceptance. Retrieved from: https://aisel.aisnet.org/cgi/viewcontent.cgi? article=1183&context=icis2013. Wei, T.T., Marthandan, G., Chong, L.Y., Ooi, K., Arumugam, S., 2009. What drives Malaysian m-commerce adoption? an empirical analysis. Ind. Manag. Data Syst. 109 (3), 370–388. Wu, P.F., 2012. A mixed methods approach to technology acceptance research. J. Assoc. Inf. Syst. 13 (3), 172–187. Zhang, L., Zhu, J., Liu, Q., 2012. A meta-analysis of mobile commerce adoption and the moderating effect of culture. Comput. Human Behav. 28, 1902–1911.
10