The Enabled Research Through Use Of Ecommerce
The main issue of e-commerce research is transaction safety, no matter from consumer or business point of view. For consumer research, it is important to know why people shopping on the vertical platform instead of physical and why they trust information from the Internet (Gefen, Karahanna, & Straub, 2003; Pavlou & Fygenson, 2006; Rodriguez-Ardura & Meseguer-Artola, 2010). Rodriguez-Ardura & Meseguer-Artola (2010) summarized e-commerce of B2C adoption researches into macro and micro perspective. They found that macro orientation studies tend to reflect the environmental dynamic but ignore factors to do with the e-commerce technology per se. Micro-level studies is the very opposite of macro, mainly identified factors to do with the technology or the organizational area, but not consider with environment or institution in. For B2B research, e-commerce provides an opportunity for a broader type of business model (Iyer, Germain, & Claycomb, 2009) therefore e-commerce triggered electronic data interchange (EDI) implements and online service. More and more product information and advertisement may found on the Internet influences price transparency. Still, trust and risk avoided issue is the main concerned. Whitten et al. (2000) pointed out that e-commerce could provide necessary infrastructure under which disparate business activities can be effectively planned and coordinated.
However research on how e-commerce impacts on performance remain split. From resources based point of view (RBV), e-commerce re-optimize business processes and practices, could leads to increased efficiency and improved performance (Sarkar and Jagjit, 2006). From strategy research point of view, IS usage needs to be ‘matched’ with firm’s information needs, than makes performance improved (Premkumar, Ramamurthy, & Saunders, 2005). However, these research limited performance evaluation as financial performance (Iyer, Germain, & Claycomb, 2009; Khazanchi, 2005), such as market share (Kim, Song, & Koo, 2008) or how many OEM’s orders adds (Chang, Wang, & Chiu, 2008) rare provides a full picture of organizational process improvement. The effects on performance and organizational process improvement need to be re-exam more precisely.
In this research, we uses Premkumar, Ramamurthy, and Saunders (2005) concept of ‘‘fit’’ on information processing view of organizations. We applied the ‘‘fit’’ concept as a mediator variable (Venkatraman, 1989) examines the fit between e-commerce activity and environmental dynamic and the effect of the fit on performance, using longitudinal data. The objective of this research is to (1) improve more clear understanding of the contribution of e-commerce to firm performance; (2) develop and test hypotheses on the role and effect of e-commerce activities using longitudinal data; and (3) use longitudinal data to see e-commerce activities trace and its impacts on performance. To answer these questions, a longitudinal study was conducted over three year period from 2007 to 2009, and a model named ‘The FIT Model of E-Commerce Performance’ was proposed in this research. The contribution of this research is that (1) proposed an integrated model of e-commerce enabled research; (2) provide a clearly perspective of e-commerce activities trace; (3) exams e-commerce activities with a full and clear picture reflect the real business phenomenon; (4) delineate the concepts of the organizational process by adopt Balanced Scorecard in our research, data was collected from CIO or who could made information decision of The Largest 5,000 Corporations in Taiwan has its representative.
2. Literature Review
2.1 E-commerce processing capability
Pervious research on e-commerce is more concerned about transaction safety and why individual or organization adopts virtual platform. Many research mentioned trust, perceived ease of use (PEOU), perceived usefulness (PU), supportive institutional environment, when individual and organization adopts e-commerce (Oxley and Yeung, 2001; Rodriguez-Ardura & Meseguer-Artola, 2010). Gefen, Karahanna, and Straub (2003) using TAM model and trust theory discuss how customer trust can be maintained. Sung and Gibson (2005) mentioned that security of systems, privacy of information, technical e-commerce expertise, quantity of information about supply, and variety of goods and services influence intention of e-commerce use. However few researches discussed about information capability carried by firms. Contents and online function provided on Internet decided by information capability. With rich information capability, the contents and function user could browse on the Internet is more richness. For example, Ordanini and Rubera (2010) suggested firm IT capabilities, partners IT readiness, and service providers IT capabilities is the main IT resources that could influences on firm performance after e-commerce application. Chu, Leung, Hui, and Cheung (2007) evaluated the function of e-commerce Web sites from 1993 to 2001 as a longitudinal study. They suggested four level of e-commerce evolution, from pre-web, reactive web, interaction web to integrative web. The function of e-commerce is more complexity and more customization. E-commerce processing capability is defined as the level of IS support for various activities of the e-commerce, customer touch point, customer data analysis, and online advertising methods. Most e-commerce studies examined the use of a single technology (e.g., electronic data interchange [EDI]). Conceptualizing various interorganizational interactions using the e-commerce processing capability and examining IS support using a range of technologies, such as Automated Interactive Voice Response (IVR), for each activity in e-commerce is a unique contribution in our research.
2.2 Environmental Dynamic
From resource-based perspective, the market environment is continuous and unpredictable (Miller and Friesen, 1983; D’Aveni, 1995) that often render enterprise strategies invalid (Bourgeois and Eisenhardt, 1988; Majumdar, 2000). Firms are facing multiple uncertainties from both supply chain and customer, even technical revolution. For example, Premkumar et al. (2005) summarized eight kinds of uncertainty into two categories, environment and relationship. The eight kinds of uncertainty include product description complexity, technology, demand, supply, product, trust, and investment from firm and supplier. They defined these two uncertainties as information processing needs. To reduce these uncertainties, many activities are selected and adopted in business process, which defined as information processing capability. In our research, we emphasize the role of stakeholder and competitors plays in e-commerce environment. Stakeholders in the market make it difficult to analyze and predict market trends, threats, opportunities, and what and how firms do business (Daft and Weick, 1984). Market pressure comes from competitors (Pflughoeft et al., 2003; Wymer and Regan, 2005; Al-Qirim, 2005, 2007) refers to the degree of market dominance by a few market players affects firms’ strategies (Xu et al., 2004; Zhu et al., 2006). To compete with the dominant competitors, non-dominant firms adopt new IS to transfer market data into useful information quickly, reduce costs of production, administration, service, and distribution. Therefore firms adopt e-commerce to manage uncertainties and compete in the virtual word by offering personalized products and services and satisfy customer demands.
2.3 The concept of FIT
The concept of FIT is a connection between two specific variables, for example: fitness between strategy and managerial characteristic (Gupta and Govindarajan, 1984). Although FIT is easy to describe as the relationships using phrases and words such as matched with, but not easy to down to analytical level. An operational definition in empirical test with an appropriate statistical procedure about FIT is a complex issue. Galbraith and Nathanson (1979) commented: "although the concept of fit is a useful one, it lacks the precise definition needed to test and recognize whether an organization has it or not". Therefore, many researches try to category and give an operational definition about FIT. One remarkable overview of various forms of FIT is Venkatraman (1989). His research provides a category of six forms of FIT (moderation, mediation, matching, gestalts, profile deviation, and covariation), statistical methods used for analysis, and the implicit assumptions made in the theoretical formulation and empirical analysis. Among six, moderation and mediation perspective are traditional methods to test for causal effects and are extensively used. Umanath (2003) identifies the concept of Fit into three broad categories: congruence, contingency, and holistic. The category of contingency is also an interaction-specification modeled as moderation or mediation effect.
In IS research, many studies use the deterministic perspective of innovation adoption that new ICT systems are inherently good for organizations, but the notion of fit between information processing needs and information processing capability remain unclear. From organizational information processing view, Premkumar, Ramamurthy, & Saunders (2005) proposed an integrated model to estimate interorganizational supply chain performance. They link information process needs and information process capability with the concept of FIT. In their study, they defined "FIT as matching" and “match between two related variables” and an analysis of variance (ANOVA) model is used to test the impact of interaction effects (information processing needs and information processing capability) on procurement performance (Venkatraman, 1989; Premkumar, Ramamurthy, & Saunders, 2005). Firms with FIT capabilities to constantly evolve, adapt to task changes, and accommodate the dynamic environment. Such firms will pursue various opportunities by undertaking modification of their routines and resource configurations in pursuit of improved effectiveness (Porter, 1985; Zott, 2003). Only when the intra-organizational business activities match the extra-organizational environment dynamic that firms can increases market responsiveness (Zahra et al., 2006) and enhances performance (Wu, 2006, 2007).
2.4 IS Enabled Evaluation
To define and measure the impacts of information systems (IS) adoption on performance could category in many ways, for example: financial index, marketing and sales growth, structure and process redesign, and efficiency enhancement (Bharadwaj, 2000; Melville, Kraemer, and Gurbaxani, 2004; Powell and DentMicallef, 1997). Powell and DentMicallef (1997), categorized performance into two instruments: performance and financial performance. The first performance was designed to measure executives' perceptions about information technology performance, such as: new information technologies have dramatically increased our productivity, competitive position, sales, profitability, and overall performance. The financial performance was designed as a subjective measure of financial performance itself, consisting of questions about the firms' overall profitability and sales growth, such as: over the past 3 years, our financial performance has been outstanding, has exceeded our competitors', our sales growth has been outstanding, have been more profitable than our competitors, and our sales growth has exceeded our competitors’.
The other way to measure ICT impact is using a set of financial ratios, such as return on investment (ROI) and return on assets (ROA) or volume measures such as revenue and sales growth (Weill and Olson, 1989; Bharadwaj, 2000; Santhanam and Hartono, 2003). For example, Santhanam and Hartono (2003) find that firms with superior IT capability indeed exhibit superior performance when compared to average industry performance in financial indicators, such as return on assets or operating income to assets. Francalanci and Morabito (2008) use ROE and ROI indexes to test organizational absorptive capacity and analyze its effects on the relationship between IT and business performance in small and medium enterprises. However there is a debate in business research (Venkatraman and Ramanujan, 1986; Tippins and Sohi, 2003). The weakness of accounting measures is that they focus only on the economic dimensions of performance, neglected other important goals of the firm, and not convincing enough to answer whether or not IS improves the overall productivity due to lack of systematic observation and a strategic framework to evaluate important aspects for enterprises; also, the data are often unavailable, unreliable, and many respondents are hesitant to reported (Dess and Robinson, 1984; Tippins and Sohi, 2003; Lee, Chu, and Tseng, 2010). The accounting index may be not provided by the owner, especially in small business for many reasons, such as the avoidance of corporate and personal income taxes (Sapienza, Smith, and Gannon, 1988; Raymond, Pare, and Bergeron, 1995). Therefore, Tippins and Sohi (2003) suggested that impacts of IS adoption on performance cannot be measured directly, but can only be quantified by examining the indirect effect. Therefore it is better to asked to report how well their firm performed during the last 3 years relative to all other direct competitors in terms of profitability, ROI, customer retention, and sales growth (Powell and DentMicallef, 1997; Spanos and Lioukas, 2001).
Moreover, the concepts of IS adoption influence on performance is compare to other competitors in the market. Than, the research more focus on inter-organizational improvement. For example, Melville, Kraemer, and Gurbaxani (2004) define performance in terms of efficiency, such as enhanced cycle time and cost reduction. Premkumar, Ramamurthy, and Saunders (2005) also estimate interorganizational supply chain performance with six items measuring various dimensions of procurement, developed by Norris, Hurley, Hartley, Dunleavy, and Balls (2000) and Handfield and Nichols (2002). The indicators are: reduced ordering costs, reduced order cycle time, improved inventory turnover, helped us get better prices on the procured products, improved coordination with the suppliers, and provided better information for decision making. The shortage of financial ratios index and suggestion of indirectly performance measurement inspires us to adopt the Balanced Scorecard approach from four different perspectives: financial indicators, customer satisfaction, internal processes, and innovation and improvement activities (Kaplan and Norton, 1993). The benefit of Balanced Scorecard is that it measures not only on growth and recession on quantitative sense (finical reports), but also offers a more complete pictures about the impacts of IS on organization. Balanced Scorecard has been widely use in many companies, such as Rockwater, Apple co., Advanced Micro Devices, and Mobil North American Marketing (McDevitt et al., 2008; Nikkel, 2008; Niven, 2008; Stewart, 2008). This measurement benefits an enterprise by assessing its performance from the strategic vision and indicating how managers improve performance (Kaplan and Norton, 1993).
3. HYPOTHESES and METHODS
Based on the literature review, we thus develop the following hypotheses:
Hypotheses 1a-1d: Corporate internal process performance (IPP) has significantly positive impacts on internal learning and growth (ILG) (1a) and customer satisfaction (CS) (1b), internal learning and growth in turn has significantly impacts on customer satisfaction (CS) (1c), and then customer satisfaction has significantly impacts on corporate financial performance (CFP) (1d).
Hypotheses 2a-2b: FIT (FIT) has significant positive impacts on internal process performance (IPP) (2a) and corporate financial performance (CFP) (2b).
Hypotheses 3a-3b: Dynamic market environment (DME) (3a) and e-commerce processing capability (EPC) (3b) has significant positive impacts on FIT.
Figure 1 presents the FIT model of e-commerce performance proposed by this study. This model examines the fit concept between e-commerce activity and environmental dynamic and the effect of the fit on performance. It assumes that the FIT ability of firms as a mediator between dynamic market environment and e-commerce activity. Later, the FIT ability impacts internal process and financial performance.
[Insert Figure 1. here] The FIT Model of E-Commerce Performance
The preceding theoretical arguments and our proposed model can be formally stated by the following system of simultaneous regression models.
CFP = w1* FIT + w2*CS +ε1.
CS = w3* ILG + w4* IPP +ε2.
ILG= w5* IPP +ε3.
IPP = w6* FIT +ε4.
FIT = w7* DME + w8* EPC +ε5.
Based on the preceding theories and hypotheses, this section details the subsequent constructs and questionnaire items, and the survey settings to collect the empirical data from the senior information systems managers of organization units that make independent decisions on information technology systems.
4.1. Instrument Development
The final questionnaire, developed partially in accordance with the Business Information Technologies project, includes two major parts. Table 1 and appendix 1 shows the detailed constructs in the proposed FIT model of e-commerce performance model and e-commerce activities we adopt. The first part is composed of fifteen questions as shown in Table 1. The second part contains information about the surveyed company, including size, industry type, and capitals. The items selected for the constructs in our model are mainly adapted from prior studies to ensure content validity. For example, dynamic environment refers to unpredictability and volatility of the business environments. Two items modified from literature (Anand and Ward, 2004; Pavlou, 2004) and are used to measure the instability and turbulence of the environments. We adopt the balanced scorecard approach (Kaplan and Norton, 1993) and view the corporate performance from four perspectives: internal business process organizational learning and growth, customer, and financial perspectives. A number of key performance indices, such as standardization and production automation, customer satisfaction, innovative proposals, revenue, etc. are listed in Table 1.
[Insert Table 1 here]: Detailed Constructs and Questionnaire Items
4.2. Survey Settings
The survey was sent to Chief Information Officers (CIOs) or senior information systems managers of various organizations in 2007, 2008, and 2009. For each year, questionnaires were mailed to over 2,000 companies by systematically sampling the database of The Largest 5,000 Corporations in Taiwan. Of all 2,000 questionnaires distributed, 436, 416, and 284 valid questionnaires were returned. This low response rate is not uncommon in the management area of obtaining responses from top management (Ferratt et al., 1999). Among three year, manufacturing is the largest industry number which contains 47.9, 51.7, and 51.8%, respectively, followed by service industry, nearly 20% in each year. We demonstrate capital and employee situation by quarter in each year, shows in table2. The capital of these surveyed organizations range from below US$9 million to above US$ 0.4 billion. The range of employment is from fewer than 75 to more than 600 workers.
[Insert Table 2 here]: The sample sizes and demographics of each year
A more rigorous approach to modeling relationships between constructs, structural equation modeling (SEM) may offer a more holistic picture of FIT modeling, as relationships between FIT and business performance can be examined simultaneously in one unified model. In addition to descriptive statistical summaries, multiple-group structural equation modeling was used to test the model. The SPSS 12.0 and AMOS 16.0 software packages were used for the statistical analysis.
4.1 Evaluations of the individual measurements
We evaluated three level of IS support for e-commerce activities, customer touch point, customer data analysis, and online advertising methods. Each level has a check list with six to nine items, see appendix 1 for detail. The results show quite insufficient and divergence of IS implements. The dynamic market environment instrument is more converge, firm agrees that there are many competitors in their kind of industry and industry is very competitive. Firms are confident of their FIT capability, could efficiently integrate and reallocate existing resources to meet environmental change and re-adjust according to competitive environments. The positive response was reported in the performance instrument. There is a consistence results show that most response agrees that technology impacted their organization with positive influence. However, on the corporate financial performance, firms do not agree that technology impacts on profits and margins positively.
4.2 Quality of the Proposed Model
Scale reliability and construct validity are tested with confirmatory factor analysis (CFA) (Bentler, 1995) before assessing the hypothesized relationships shown in Figure 1. It is also necessary to identify equivalent phenomena when conducting multi-grouping research (Calantone et al., 1996). A separate CFA that involves raw data as input is performed for each of the years (Anderson and Gerbing, 1988). All constructs in the three samples are composed of identical items. Based on CFA results, we eliminate items that load on multiple constructs or have low item-to-construct loadings (Anderson and Gerbing, 1988). Then, the fit of the measurement model for each of the three years (Bagozzi and Yi, 1988) is assessed. As Table 3 shows, the loadings of items on their respective factors are highly significant (p < 0.01). The goodness of fit index (GFI), comparative fit index (CFI), and Bollen’s fit index (IFI) range between 0.96 and 0.99 for each of the three years. The root mean square of error approximation (RMSEA) values is 0.035, 0.053, and 0.048 for 2007, 2008, and 2009, respectively. The results suggest that the data converge and the CFA model for all four countries fits the data adequately (Bentler 1995, Bollen 1989).
[Insert Table 3 here]: Summary of reliability and validity index
To test if there are true multi-grouping differences in structural relationships between scale scores for further comparison, measurement invariance is conducted to rule out differences caused by systematic response biases or problems with scale artifacts, reliability, or nonequivalence (Mullen, 1995, Steenkamp and Baumgartner, 1998). Following Mullen (1995) and Steenkamp and Baumgartner (1998), we perform a test of unconstrained model (Model 1). It produces χ2 =588.05 with 332 degrees of freedom (df) (Table 4). CFI is found to be 0.97 and RMSEA is 0.018. These fit indices indicate that the six-factor model is supported in all groups, meaning that the four countries exhibit the same simple factor structure. Each sample population can therefore use the same baseline model. The second model (Model 2) assesses whether the factor covariances are equal across the four countries by constraining the λ to be equal across groups. The results (Table 4) demonstrate that the chi-square difference between the simple structure model (Model 1) and the equal variance model (Model 2) is 9.41 with 24 df (p > 0.05), the chi-square test is insignificant which means two model in consistence. The chi-square test is insignificant (an indication of a good fit), research relying exclusively on chi-square difference tests may suffer “from the same well-known problems as the chi-square test for evaluating overall model fit” (Steenkamp and Baumgartner, 1998) (also see e.g., Anderson and Gerbing, 1988, Marsh and Grayson, 1990). Therefore, we also examine the root mean square error of approximation (RMSEA); the consistent akaike information criterion (CAIC), the comparative fit index (CFI); and the tucker-lewis index (TLI) or nonnormed fit index (NNFI) for model fit indication, suggested by Steenkamp and Baumgartner (1998). Overall, the factor loadings, factor correlations, and factor variance are invariant across years.
[Insert Table 4 here]: Measurement invariance tests through constrained at several levels
4.3 Results for Hypotheses Testing
Table 5 shows path coefficients of the three years. There are several consistences phenomenon. First, information processing capability has significant and positive impacts on FIT, with a standardized path coefficient of 0.14, 0.23, and 0.28, respectively. Second, FIT has significant and positive impacts on internal process performance, with a standardized path coefficient of 0.63, 0.70, and 0.71, respectively. At last, internal process performance has significant and positive impacts on internal learning and growth, with a standardized path coefficient of 0.60, 0.82, and 0.64, respectively. Dynamic market environment has only significant and positive impacts on FIT in 2007, with a standardized path coefficient of 0.11. Internal process performance has significant and positive impacts on customer satisfaction, with a standardized path coefficient of 0.36 and 0.67, in 22007 and 2009. Internal learning and growth has significant and positive impacts on customer satisfaction, with a standardized path coefficient of 0.51 and 0.22, in 22007 and 2009.
[Insert Table 5 here]: Results of structural equation model analysis of individual country models
Moreover, the integrated model from three year data converged as well as the others. All paths are significant and positive, except path from FIT and customer satisfaction to corporate financial performance. We also found several notable and unanticipated results. First, path coefficient from FIT to corporate financial performance is only negative and significant in 2007 and negative in 2008, however, positive in 2009. Second, path coefficient from customer satisfaction to corporate financial performance is negative and significant, with standardized path coefficients of -0.14, -0.32, and -0.39, respectively. An explanation for why customer satisfaction does not have a positive impact on corporate financial performance might be that (1) we estimate profit instated of revenue; (2) there are mixed effects between profit and customer satisfaction.
There has been plenty of evidence of e-commerce adoption research, but no study has empirically assessed these changes systematically based on a theoretical framework and its impacts on performance. The study provided empirical support for the FIT concepts on environmental dynamic, e-commerce processing capability, and performance. The study also provided a longitudinal research of how organizations evolve in adopting the e-commerce activities. Consistent with the prior research (Whitten et al., 2000), e-commerce could provide necessary infrastructure under which disparate business activities can be effectively planned and coordinated. It should be noted that the impacts of FIT and customer satisfaction on corporate financial performance was not supported in this research. Despite all the care given to this study, there are several limitations of the present study that should be noted and addressed in any future research. One of the shortcomings of the study is in the sampling process, although various measures were taken to ensure that the data represent a wide range of organizations, but capital and employee satiation is not that consistent. Second, instead of panel data analysis, we adopt longitudinal approach, though invariance tests showed a good fit of estimation. Third, we adopt profit as financial performance indicator; it may have mix effects of revenue and costs from real business practice. The most important practical contribution of the study is that provides a clear picture of e-commerce enabled on organization. The FIT model should provide organizations with a strategic and macro perspective to evaluate and manage various e-commerce applications and consumer research. An important contribution of this study is that it empirically observed the trace of the e-commerce practices in business. The study provides a valuable theoretical framework for further research.
Need help with your literature review?
Our qualified researchers are here to help. Click on the button below to find out more:
In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: