Data Analysis and Results

Chapter 5 Data Analysis and Results 5.1 Introduction In this chapter, data analysis and results required to answer the research questions are presented. A total of 200 MSMEs of different firm sizes were contacted through direct contact method. Out of 200 MSMEs, 122 MSMEs agreed to participate in the survey resulting in the response rate of 61 per cent. This was higher response rate compared to similar studies. The respondents are primarily the owner/managers of the MSMEs. The data collected from 122 MSMEs is analysed using SPSS version 16. The tools used for data analysis are descriptive statistics, independent samples t-test, one-way ANOVA, chi-square tests and Logistic regression analysis. The analysis is presented in the order described below. First, profile of adopter and non-adopter MSMEs in terms of their size of the firm and industry in which they operate are presented. The stage of adoption to understand the level of participation of the adopter MSMEs is also discussed. Cross tabulation of firm size and adoption of B2B e-marketplace is done to identify any relation between adoption and firm size. Chi-square test is used to find the dependence between adoption of B2B e-marketplace and firm size. The purpose is to identify any pattern in adoption rate based on firm size as the MSME sector has firms with diverse firm sizes. Second, the awareness level of the MSMEs on various aspects of B2B e-marketplaces is analysed. Comparison of awareness level among adopters/non-adopters is done and independent samples t-test is used to identify whether there are any significant differences in means of the two groups. Comparison of awareness level among MSMEs of different firm sizes is performed and one way ANOVA is used to identify any significant difference 104

between the four groups of MSMEs based on firm size. Third, the factors affecting the adoption of B2B e-marketplaces by MSMEs are discussed. Descriptive statistics, reliability and validity test results for the factors in the research model is discussed first. Logistic Regression analysis is used to identify the most significant factors among the twelve factors that influence adoption. Based on the results of the Logistic Regression, research hypotheses are accepted or rejected. Fourth, the services of the B2B e-marketplaces used by the adopter MSMEs are analysed. Fifth, the benefits of using B2B e-marketplaces for the adopter MSMEs are presented. Finally, barriers to adoption and use of B2B e-marketplace among MSMEs are discussed. Comparison of barriers among adopters and nonadopters is done and independent samples t-test is used to identify any significant difference among the two groups. Comparison of barriers among MSMEs based on firm size is done and one-way ANOVA is used to identify any significant difference among the four groups. 5.2 Profile of the adopter and non-adopter MSME respondents in the sample The sample has almost equal representation from the MSMEs of different firm sizes and different prime industry sectors of the Karnataka state. The sample composed of MSMEs from sectors such as apparel/textile, auto/auto components, food/marine products, electrical and electronic and Pharma/Biotech that form some of the core manufacturing sectors in the state. The others category comprised of MSMEs that produced variety of products such as plastics products, construction related products, toys, utensils and the like. There are 66 adopters and 56 non-adopters in the sample. The sample profile of the final sample is shown in Table 5.1. 105

Table 5.1: Sample Profile Characteristics Firm Type Frequency (N) Firm size Micro (<= 25 lakh) 30 Small-Group1 (> 25lakh <= 1 crore) 28 Small-Group2 (> 1 crore <= 5 crore) 33 Medium (> 5 crore and <= 10 crore) 31 Industry Apparel/Textile 15 Auto/Auto components 15 Electrical/Electronics/Hardware 20 Food/Marine 22 Pharma/Biotech 15 Others 35 B2B E-marketplace Adoption Non Adopter 56 Adopter 66 Total MSMEs in the sample: 122 The adopter MSMEs are in different stages of adoption and their details are shown in Table 5.2. It is observed from the adopters profile that majority of the adopters (56 per cent) are in exploratory stage followed by 27 per cent of adopters in trial stage and only 17 per cent of the adopters are in commitment stage. As the adoption of B2B e-marketplace in India is in growth stage, MSMEs have registered with the B2B e-marketplaces and exploring the possibilities of using the e-marketplace. 106

Table 5.2: Profile of adopter MSMEs Adoption Stage N Per cent Exploratory Stage 37 56 Trial Stage 18 27 Commitment Stage 11 17 Total adopters in the Sample: 66 B2B e-marketplace adoption and firm size Previous studies (Lehman, 1985; Chen and Fu, 2001; Lee and Xia, 2006) on adoption of innovation have reported differences in adoption of innovation based on firm size. Therefore, the pattern of adoption based on firm size is explored to understand the influence of the firm s level of investment on the adoption of e-marketplace. A cross tabulation of firm size wise adoption is shown in Table 5.3. It is observed that out of the total 30 micro firms, eighty per cent of the micro firms are non-adopters. Table 5.3: Cross tabulation representing firm size and adoption of B2B e- marketplace Firm Size Adopters (n=66) Frequency (column percentage) Non-Adopters (n=56) Frequency (column percentage) Micro 6 (9) 24 (43) Small-Group1 10 (16) 18 (32) Small-Group2 26 (39) 7 (12.5) Medium 24 (37) 7 (12.5) 107

It is observed that, out of 66 adopters, 39 per cent of the adopters belonged to Small-Group2 (investment range: more than one crore rupees and less than five crore rupees) and 37 per cent of the adopters are medium enterprises (investment range more than five crore rupees and less than ten crore rupees). This group of adopters (firm s investment more than one crore rupees and less than ten crore rupees) constituted for more than 76 per cent of the adopters. The remaining 24 per cent of the MSME adopters belonged to group less than one crore rupees investment. Out of the total non-adopters, as shown in the Table 5.3, majority of the firms (43 per cent) are micro firms, followed by Small-Group1 with 32 per cent. This showed that total 75 per cent of the non-adopters are firms with investment less than one crore rupees. The remaining 25 per cent of nonadopters belonged to the group Small-Group2 and medium enterprises in equal proportions. The chi-square test, testing dependency between firm size and adoption, shows that size of the firm has significance influence on adoption of B2B e-marketplace (chi-square=32.748, df=3, p<0.01). This finding further consolidates the findings in previous studies that reported firm size has a significant influence on the adoption of technology. In this study, it has been observed that as the size of the firm is higher, the adoption rate is higher than their proportion in the sample. It can be inferred from data on firm size wise adopters and nonadopters of B2B e-marketplace (refer Table 5.3) that MSMEs can be classified into two categories based on adoption rate. The first category is the MSMEs with investment level less than one crore rupees (micro and Small- Group1 together) and the second category is the MSME that has investment level more than one crore rupees. These two categories have varying pattern of adoption of B2B e-marketplaces with first category having lower adoption rate. 108

5.3 Awareness level of B2B e-marketplace among MSMEs The first objective of the research is to explore the awareness level of the MSMEs about B2B e-marketplaces. Awareness level of MSMEs on different aspects of e-marketplace such as e-marketplaces relevant to their business, competitor use of e-marketplaces, opportunities and threats of use of e-marketplaces, business models and service provided by e-marketplaces, and awareness on benefits of e-marketplaces are analysed. Comparison of awareness level among adopters and non-adopters is done. Comparison of awareness level firm size is also done to identify the extent of MSMEs awareness. The findings of the analysis are discussed in the following sections. 5.3.1 Comparison of awareness level of adopters and non-adopter MSMEs Comparison of means of awareness level on various aspects of B2B e- marketplaces is done between two groups: adopters and non-adopters. The Table 5.4 shows the means and standard deviation of the adopters and nonadopters with reference to their awareness level. The independent samples t- test is used to compare the means of adopters and non-adopters and the results are included in the table. It is observed that the adopter MSMEs have higher degree of awareness level on all the aspects. Adopters group had positive scores with respect to awareness of B2B e-marketplaces relevant to their business, competitor use of e-marketplaces, opportunities and threats of using e- marketplaces, services and business models of e-marketplaces and benefits of using e-marketplaces. aspects: The non-adopter MSMEs have lowest awareness in the following two Recognizing the opportunities and threats of e-marketplaces Understanding services and business models of e-marketplaces. 109

The independent samples t-test results (refer Table 5.4) revealed that awareness level differed significantly (p<0.01) between the two groups: adopters and non-adopters in all the aspects of awareness. This indicates that non-adopters are aware of the B2B e-marketplaces relevant to their business, but do not have in-depth understanding of possible opportunities that can be explored through B2B e-marketplaces. Table 5.4: Comparison of B2B e-marketplace awareness level (means) among adopters and non- adopters Adoption Non Adopter n=56 Aware of e- markets relevant to business Aware of competit -or's use of e- market Firm Recognizes the opportunit -ies and threats of e-markets Firm understan -ds services and business models of e-markets Firm underst -ands benefits of e- markets Mean 3.54 3.32 2.96 2.88 3.12 S.D. 1.293 1.377 1.525 1.502 1.415 Adopter n=66 Mean 4.65 4.56 4.56 4.55 4.53 S.D..480.682.611.532.561 Total N=122 Mean 4.14 3.99 3.83 3.78 3.89 S.D. 1.093 1.223 1.377 1.370 1.254 t-value 6.111* 6.128* 7.348* 7.912* 6.981* *p<0.01 110

The Table 5.5 shows the frequencies and percentages of MSMEs that agree / disagree / neither agree nor disagree on awareness of various aspects of B2B e-marketplaces. It is observed that out of 66 non-adopter MSMEs, 61 per cent of non-adopters do not understand the services and business models of B2B e-marketplaces. It is found that 57 per cent of nonadopters have indicated that they are not aware of opportunities and threats of using B2B e-marketplaces. Forty one per cent of the non-adopters are not aware of the benefits of B2B e-marketplaces. Table 5.5: Awareness level (frequencies) of adopters and non-adopters Awareness Adoption Agree N (%) Aware of e-marketplace relevant to business Non Adopter (n=56) Adopter (n=66) Neither Agree Nor Disagree N (%) Disagree N (%) 39 (70) 2 (3) 15 (27) 66(100) 0 0 Total (N=122) 105(86) 2 (2) 15(12) Aware competitor's use of e- marketplace Non Adopter 29(52) 6(11) 21(37) Adopter 63(96) 1(1) 2(3) Total 92(75) 7(6) 23(19) Recognizes the opportunities and threats of e-marketplace Non Adopter 24(43) 0 32(57) Adopter 64 (97) 1(1) 1(1) Total 88 (72) 1(1) 33 (27) Understands services and bus models of e-marketplace Non Adopter 22 (39) 0 34 (61) Adopter 65 (99) 1(1) 0 Total 87 (71) 1 (1) 34 (28) Firm understands benefits of e- marketplace Non Adopter 25(45) 8(14) 23(41) Adopter 64(97) 2(3) 0 Total 89 (73) 10 (8) 23(19) 111

5.3.2 Comparison of awareness level firm size wise The awareness levels of MSMEs of different firm sizes such as micro, small-group1, small-group2 and medium enterprises are compared to understand if there are any observed differences. One-way ANOVA analysis is performed to understand the significance of the observed differences. The findings are shown in Table 5.6. Firm Size Micro n=30 Small- Group1 n=28 Small- Group2 n=33 Medium N=31 Table 5.6: Comparison of awareness level (firm size wise) Aware of e-mkt relevant to business Aware of competitor's use of e-mkt Recognizes opportunities and threats of e-mkts Understands services and business models of e- mkts Mean 3.33 3.13 2.77 2.53 2.87 S.D. 1.241 1.383 1.455 1.383 1.332 Mean 4.04 3.86 3.75 3.75 3.79 S.D. 1.38 1.777 1.295 1.295 1.197 Mean 4.42 4.27 4.24 4.24 4.33 S.D..792 1.039 1.146 1.062.92 Mean 4.71 4.65 4.48 4.52 4.48 S.D..643.709.962.811.890 Understands benefits of e-mkts F-statistic 11.424* 10.612* 11.809* 17.718* 13.682* *p<0.01 112

Based on the results of one-way ANOVA analysis of awareness level firm size wise, it is found that awareness level differed significantly between the four groups of MSMEs based on firm size. One-way ANOVA results also revealed that awareness level of micro firms significantly differed from other groups in all the cases. Table 5.6 shows that micro firms have lesser awareness levels than other firms. 5.4 Factors affecting the adoption of B2B e-marketplace among MSMEs The second objective of the research is to identify the factors that influence the adoption of B2B e-marketplace among Micro, Small and Medium Enterprises. Based on the review of literature and the exploratory stage of the research, twelve factors were identified to influence B2B e-marketplace among MSMEs as shown in the conceptual framework in Chapter 3. Research Hypotheses were formulated to test the influence of these factors on the MSME adoption of B2B e-marketplaces. The Table 5.7 summarises the twelve research hypotheses. Descriptive statistics of the research variables is explored for preliminary analysis. Before testing the hypothesis, all the twelve factors were subjected to reliability and validity tests to confirm its inclusion in the multivariate analysis. Test for multicollinearity is done before performing the regression analysis. Based on the confirmation of results of reliability and validity tests and multicollinearity test, Logistic regression analysis is performed to test the hypotheses and to identify the most significant factors influencing adoption. 113

Table 5.7: Research Hypotheses Hypothesis 1 Organisation resources is positively related to the B2B e- marketplace adoption by MSMEs Hypothesis 2 Mimetic pressures is positively related to B2B e-marketplace adoption by MSMEs Hypothesis 3 Adoption among suppliers is positively related to B2B e- marketplace adoption by MSMEs Hypothesis 4 Perceived dominance of supplier adopters is positively related to B2B e-marketplace adoption by MSMEs Hypothesis 5 Adoption among customers is positively related to the B2B e- marketplace Hypothesis 6 Hypothesis 7 Hypothesis 8 Hypothesis 9 Perceived dominance of customer adopters is positively related to B2B e-marketplace adoption by MSMEs Product characteristics negatively influence the MSME adoption of B2B e-marketplace Demand uncertainty is positively related to B2B e-marketplace adoption by MSMEs Market volatility is positively related to B2B e-marketplace adoption by MSMEs Hypothesis 10 Perceived relative advantage is positively related to B2B e- marketplace adoption by MSMEs Hypothesis 11 Hypothesis 12 Perceived complexity of the e-marketplace is negatively related to MSME Adoption of B2B e-marketplace Compatibility is positively related to the B2B e-marketplace adoption by MSMEs 114

5.4.1 Results of reliability and validity of research instrument A measure is reliable if it provides consistent results. For multi-item scales, testing reliability of the scale is a pre-condition for instrument validity. In the present study, Cronbach Alpha is used to test the reliability of the scale and ensure inter-item consistency. Cronbach Alpha is computed for all the twelve factors in the conceptual model and is presented in the Table 5.8. The results of reliability tests (refer Table 5.8) show that the Cronbach Alpha of all the factors is in the range of 0.8209 to 1. As the acceptable cut-off is 0.7 (Nunnally, 1967), all the factors are considered in acceptable range. The constructs used in the present study has been tested in the previous studies for validity. However, the content validity is tested through review by academicians and B2B e-marketplace vendors. The construct validity is again tested using exploratory factor analysis method. Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy (MSA) is an index used to measure appropriateness of factor analysis. Therefore, KMO is computed and is found to be 0.798. High value of KMO (between 0.5 and 1) is deemed appropriate as per Malhotra (1999). Therefore, this implies factor analysis is valid in this case. Principal Component analysis method using Varimax rotation is used to perform factor analysis. Principal Component method is a widely used method and is used when the purpose is to obtain minimum number of factors needed to represent the original set of data. Several rotation methods are used to identify those variables that load on one factor and not on another and achieve a theoretically more meaning factor pattern. (Ho, 2006). Varimax is a widely used orthogonal rotation method that assumes that factors are independent and rotation process maintains the reference axes of the factors at 90 degrees. Varimax has become popular as it gives clearest separation of factors (Ho, 2006). 115

Table 5.8: Reliability measures (Cronbach Alpha) for the factors in the conceptual model S. No. Factors Cronbach Alpha 1 Organisation Resources 0.9610 2 Mimetic Pressures 0.9164 3 Adoption among suppliers 0.8141 4 Perceived Dominance of Supplier Adopters 0.9478 5 Adoption among customers 0.8209 6 Perceived Dominance of Customer Adopters 0.8878 7 Product Characteristics 0.9660 8 Demand Uncertainty 0.9365 9 Market Volatility 1 10 Perceived Relative Advantage 0.8322 11 Perceived Complexity 0.8461 12 Compatibility 0.9868 The factor analysis generates factor loadings that represent the correlations between each item/variable and each factor (Babbie, 2009). The factor loadings of all the items and results of the factor analysis is shown in the Table 5.9. All the factor loadings are above 0.5 and the values are in the range 0.592 to 0.915. These twelve factors explained 84.834 per cent of the variance. In the present analysis, the eigenvalues above 1 are only selected. Eigenvalue represents the total variance explained by each factor (Malhotra, 1999). 116

Factor Number Table 5.9: Results of Factor analysis Variables/Factors Items Factor Loading 1. Organisation Resources 2. Mimetic Pressures (Adoption Among Competitors, Perceived Success of adopted competitors) 3. Demand Uncertainty (Frequency uncertainty, Volume uncertainty) 4. Product Characteristics(Asset Specificity and Product Complexity) 5. Perceived Relative Advantage Eigen value Percentage of Variance Explained Cumulative Percentage HR1 0.856 9.021 16.706 16.706 HR2 0.871 BR1 0.840 BR2 0.915 BR3 0.877 BR4 0.898 TR1 0.854 TR2 0.804 TR3 0.725 TR4 0.796 TR5 0.617 CA1 0.773 5.944 11.007 27.713 CA2 0.710 CS1 0.849 CS2 0.884 CS3 0.883 SFU1 0.890 5.807 10.754 38.467 SFU2 0.877 SVU1 0.879 SVU2 0.855 SAS1 0.877 5.417 10.032 48.499 SAS2 0.879 SAS3 0.836 SPC1 0.938 SPC2 0.933 SPC3 0.943 PRA1 0.675 3.703 6.857 55.356 PRA2 0.679 PRA3 0.670 PRA4 0.621 PRA5 0.597 PRA6 0.701 PRA7 0.592 (Table 5.9 continued) 117

(Table 5.9 continued) Factor Number Variables/Factors Items Factor Loading 6. Perceived Dominance of Supplier Adopters 7. Perceived Dominance of Customer Adopters 8. Perceived complexity Eigen value Percentage of Variance Explained Cumulative Percentage SD1 0.895 3.701 6.855 62.210 SD2 0.901 SD3 0.873 SD4 0.889 CD1 0.805 3.256 6.030 68.240 CD2 0.817 CD3 0.848 CD4 0.832 PC1 0.724 2.610 4.833 73.073 PC2 0.734 9. Compatibility CP1 0.646 2.172 4.023 77.096 10. Adoption among suppliers 11. Adoption among customers CP2 0.684 CP3 0.685 SA1 0.782 1.802 3.337 80.432 SA2 0.798 CSA1 0.652 CSA2 0.521 1.202 2.210 82.642 12. Market Volatility SVL1 0.877 1.184 2.192 84.834 SVL2 0.877 SVL3 0.877 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 7 iterations 118

Only factors with eigenvalues of 1 or greater are considered to be significant. The rationale for this criterion is that amount of common variance explained by an extracted factor should be at least equal to the variance explained by a single variable is that factor is to be retained for interpretation (Ho, 2006). The above results confirmed reliability and validity of the data and further analysis is performed to test the hypothesis. 5.4.2 Descriptive statistics of factors in the conceptual model Descriptive statistics of the independent variables in the conceptual model are shown in Table 5.10. All the variables are measured in the five point Likert scale except for the variable market volatility as discussed in Chapter 4. The higher values indicated positive feedback as they indicate higher degree of agreement with the statement. It can be observed from the overall descriptive statistics of the sample that the e-marketplace related factors such as perceived relative advantage, perceived complexity and perceived compatibility have scores below 3 indicating disagreement. We can also find that adoption among suppliers and adoption among customers also have lower scores which implies that majority of the respondents disagree that their suppliers/customers have adopted B2B e-marketplaces. Descriptive statistics of the independent variables of adopters and nonadopters are compared to understand the significance of differences between the two groups. The findings are presented in Table 5.11. It is found that the adopters have higher scores than non-adopters in organization resources, mimetic pressures, adoption among suppliers, adoption among customers, perceived dominance of customer adopters, demand uncertainty, market volatility, perceived relative advantage and perceived compatibility. It is observed that non-adopters have higher scores in perceived dominance of supplier adopters, product characteristics and perceived complexity. This is justifiable as product characteristics and perceived complexity is negatively related to e-marketplaces, according to previous studies. 119

Table 5.10: Overall descriptive statistics of factors in the conceptual model Factor Mean (N=122) Minimum Maximum S. D. Organisation Resources 3.62 1 5 1.055 Mimetic pressures 3.20 1 5 1.135 Adoption among suppliers 2.39 1 4 0.896 Perceived dominance of Suppliers 3.51 1 5 1.108 Adoption among customers 2.48 1 5 0.955 Perceived dominance of Customers 4.14 2 5 0.816 Product characteristics 3.21 1 5 1.462 Demand uncertainty 3.81 2 5 0.921 Market volatility 1.22 1 2 0.417 Perceived relative advantage 2.63 1 5 0.929 Perceived complexity 2.52 1 5 1.030 Compatibility 2.89 1 5 1.151 120

Table 5.11: Comparison of descriptive statistics of non-adopters and adopters Factor Organisation Resources Mimetic pressures Adoption among suppliers Perceived dominance of Supplier adopters Adoption among customers Perceived dominance of Customer adopters Product characteristics Demand uncertainty Market volatility Perceived relative advantage Perceived complexity Compatibility Non-adopters (N=56) Mean (S.D.) 2.91 (0.978) 2.46 (0.914) 2.36 (0.796) 3.84 (1.023) 1.96 (0.762) 4.11 (0.755) 3.32 (1.295) 3.80 (0.999) 1.18 (0.386) 2.45 (1.008) 3 (1.044) 2.29 (0.986) Adopters (N=66) Mean (S.D.) 4.23 (0.675) 3.83 (0.904) 2.42 (0.978) 3.23 (1.107) 2.91 (0.890) 4.17 (0.870) 3.12 (1.593) 3.82 (0.858) 1.26 (0.441) 2.79 (0.832) 2.12 (0.832) 3.39 (1.036) 121

5.4.3 Logistic Regression and Hypotheses Testing Logistic regression analysis is used in the study to analyse the factors influencing the adoption of B2B e-marketplaces. Binary Logistic regression is primarily used when the dependent variable is a categorical variable (usually dichotomous) and has two outcomes such as 0 and 1. With a categorical dependent variable, logistic regression is often chosen when the predictor variables are a mix of continuous and categorical variables and/or if they are not nicely distributed (logistic regression makes no assumptions about the distributions of the predictor variables). In the present study, since the dependent variable is the adoption of B2B e-marketplace by MSMEs, which is a dichotomous variable (0: nonadopter and 1: adopter), Logistic regression is chosen. This analysis allows identifying the variables that are most significant in predicting the adoption. Logistic regression can be used to predict a dichotomous dependent variable on the basis of continuous and/or categorical independents and to determine the percent of variance in the dependent variable explained by the independents and to rank the relative importance of independent variables in predicting the dependent variable (Hair et al., 1998). In Logistic regression, since the dependent variable is dichotomous we cannot predict a numerical value for it as in linear regression. Instead, Logistic regression employs binomial probability theory in which there are only two values to predict: that probability (p) is 1 rather than 0, i.e. the event/person belongs to one group rather than the other (Burns and Burns, 2008). Logistic regression forms a best fitting equation or function using the maximum likelihood method, which maximizes the probability of classifying the observed data into the appropriate category given the regression coefficients. Logistic regression calculates changes in the log odds of the dependent, not changes in the dependent value as in linear regression. For a dichotomous variable the odds of membership of the target group are equal to the probability of membership in the target group divided by the probability of 122

membership in the other group. Odds value can range from 0 to infinity and tell you how much more likely it is that an observation is a member of the target group rather than a member of the other group. If the probability is 0.80, the odds are 4 to 1 or.80/.20 (Burns and Burns, 2008). In the present study, the independent/predictor variables are organization resources, mimetic pressures, adoption among suppliers, perceived dominance of supplier adopters, adoption among customers, perceived dominance of customer adopters, product characteristics, demand uncertainty, market volatility, perceived relative advantage, perceived complexity and compatibility as described in the previous section. In regression analysis, one common problem is multicollinearity. Multicollinearity refers to a situation in which the independent variables are correlated among themselves. The impact of multicollinearity is limited to the statistical significance of the individual regression coefficients (Hair et al., 2003). Therefore, to determine whether the present data had the problem of multicollinearity, test for multicollinearity is done. Multicollinearity can be tested in SPSS using the two collinearity statistics: Tolerance and VIF (Variation Inflation Factor). The Tolerance value is an indication of the percentage of variance in the predictor that cannot be accounted for by other predictors. The VIF is the inverse of tolerance (Ho, 2006). Hair et al. (2003) suggest VIF value larger than 10 or tolerance value smaller than 0.10 indicates the problem of multicollinearity. The results of the test for multicollinearity are shown in Table 5.12. It is observed that the VIF factors for all the twelve variables are within 2.5. Therefore, all the twelve factors were included in the Logistic regression. 123

Table 5.12: Results of test for Multicollinearity Collinearity Statistics Model Variables Tolerance VIF* Organisation resources.535 1.871 Mimetic pressures.537 1.863 Adoption among suppliers.800 1.250 Dominance of supplier adopters.773 1.294 Adoption among customers.510 1.962 Dominance of customer adopters.788 1.269 Product Characteristics.771 1.298 Demand uncertainity.439 2.280 Market Volatality.406 2.462 Perceived relative advantage.646 1.549 Perceived complexity.601 1.663 Compatibility.427 2.337 *Variation Inflation Factor Some of the assumptions of Logistic regression, according to Burns and Burns (2008) are as follows: Logistic regression does not assume a linear relationship between the dependent and independent variables. The dependent variable must be a dichotomy (2 categories). The independent variables need not be interval, nor normally distributed, nor linearly related, nor of equal variance within each group. The categories (groups) must be mutually exclusive and exhaustive; a case can only be in one group and every case must be a member of one of the groups. 124

As all of the assumptions are true in the present study, Logistic regression is chosen. Logistic Regression is performed with Forward LR Method. Forward method is a stepwise method that reports only significant variables that predict most of the cases correctly. The probability of entry is chosen as 0.20 and removal as 0.25. The options selected specifies the criteria for selecting the variable into the model (p<0.20) and removal of the variable from the model (p>0.25) in each step. The Confidence Interval CI is chosen as 95 per cent (Hosmer and Lemeshow, 2000). Logistic regression outputs are shown in Table 5.13a and Table 5.13b. The output in the Table 5.13a shows that the model is able to predict MSME s B2B e-marketplace adoption correctly in 87.7% cases. The classification table tells us how many of the cases have been correctly predicted for both adopters (value 1) and non-adopters (value 0). In this study, 83.9% are correctly classified for non-adopter group and 90.9% for the adopter group. Overall 87.7% are correctly classified. The fit statistics of the model shown such as Cox & Snell R 2 and Nagelkerke R 2 are found to have values of 0.549 and 0.733 respectively. Nagelkerke R 2 is usually greater than Cox and Snell R 2 and in this case, 0.733 indicates a moderately strong relationship of 73.3% between the predictors and the prediction (Burns and Burns, 2008). Another significant goodness of fit is the Hosmer and Lemeshow chisquare which is found to be non-significant since its significance value is more than 0.05, indicating no differences in the distribution of the actual and predicted dependent values (Hair et al., 1998). A non-significant chi-square indicates that the data fit the model well (Burns and Burns, 2008). From the Table 5.13b, we can derive the significant variables that would influence the MSME adoption of B2B e-marketplace. This is given by the Wald statistics and its corresponding significance value. The Wald statistic has a chi-square distribution. 125

Table 5.13a: Classification table for MSME adoption of B2B e-marketplace and results of goodness of fit test for Logistic Regression Classification Table a Predicted Observed Adopter Non Adopter Adopter Percentage Correct Step 4 Adopter Non Adopter 47 9 83.9 Adopter 6 60 90.9 Overall Percentage 87.7 a. The cut value is.500 Goodness of fit measures Cox & Snell - R 2 0.549 Nagelkerke - R 2 0.733 Hosmer and Lemeshow Goodness-of-Fit Test Chi-Square df Sig. 4.506 8 0.809 Table 5.13b: Results of Logistic Regression analysis Factor B S.E. Wald Df Sig. Exp(B) Organisation 2.219 0.470 22.277 1 0.000 9.199 resource Mimetic pressures Perceived dominance of customer adopters Product characteristics 1.431 0.342 17.492 1 0.000 4.183 0.717 0.376 3.632 1 0.057 2.048-0.555 0.275 4.067 1 0.044 0.574 Constant -13.549 2.739 24.465 1 0.000 0.000 126

In the Table 5.13b, the factors organisation resources and mimetic pressures have p values less 0.01, indicating that these variables are the significant predictors of the dependent variable MSME B2B e-marketplace adoption at 99% confidence level. Similarly, the factor product characteristics has p value less than 0.05, which indicates that product characteristics significantly predict the dependent variable at 95% confidence level. It can further be observed that organization resources is having the highest predictive power as the Wald s statistic is the highest in this case (22.277), followed by mimetic pressures (17.492) and product characteristics (4.067). In case of product characteristics the b coefficient of regression is negative which indicates that the variable has a negative influence on the MSME adoption of B2B e-marketplace. The chi-square test of MSME adoption of B2B e-marketplace and product characteristics reveals a negative gamma value (chi square: 10.653, p<0.05, gamma= - 0.053) that confirms that product characteristics is negatively related to the MSME adoption of B2B e-marketplace. However, the variable perceived dominance of customer adopters has p value less than 0.1, which indicates that have marginal influence on the dependent variable at 90% confidence level. Its corresponding Wald statistic is lowest (3.632). The organization resources, mimetic pressures and product characteristics are the significant predictors for the MSME adoption of B2B e- marketplaces. It means that higher the organization resources, higher the mimetic pressures (adoption among customers, perceived success of customer adopters) and lesser the product characteristics (asset specificity and product complexity), the more the probability of B2B e-marketplace adoption by MSMEs. Further, it can be seen that these are the factors which significantly discriminates adopters and non-adopter MSMEs. 127

The values represented in B column are the B coefficients of the regression model. The Logits (log odds) are the B coefficients (the slope values) of the regression equation. As our intention is not prediction, but identifying the significant dependent variables, these values are less useful for interpretation. However, the sign of the coefficient gives us an indication of the direction (positive or negative) of the relationship. In the present study, product characteristics has a negative B coefficient. The Exp(B) is the exponent of B coefficient and is called odds ratio, which estimates the change in the odds of membership in the target group for a one unit increase in the predictor (Burns and Burns, 2008). Odds ratio is a measure which is used to assess the strength of association between binary dependent and independent variable. Exp(B) value indicates that when organization resources is raised by one unit the odds ratio is 9 times as large and therefore, MSMEs are 9 more times likely to adopt the B2B e- marketplace. Based on the results of Logistic regression, Hypothesis 1 and Hypothesis2 are accepted at 99 per cent confidence level and the Hypothesis 7 is accepted at 95 per cent confidence level and Hypothesis 6 is accepted at 90 per cent confidence level. Hypothesis 3, Hypothesis 4, Hypothesis 5, Hypothesis 8, Hypothesis 9, Hypothesis 10, Hypothesis 11, Hypothesis 12 are rejected, as they are not significantly influencing MSME adoption of B2B e- marketplace. The summary of Hypotheses testing results is presented in Table 5.14. 128

Table 5.14: Summary of Hypotheses Testing results Research Hypothesis Independent Variable Relation Type # Research Hypothesis Rejected or Accepted Hypothesis 1 Organisation Resources Positive Accepted* Hypothesis 2 Mimetic pressures (adoption among competitors and perceived benefits of competitor adopters) Positive Accepted* Hypothesis 3 Adoption among suppliers Positive Rejected Hypothesis 4 Perceived dominance of Supplier adopters Positive Rejected Hypothesis 5 Adoption among customers Positive Rejected Hypothesis 6 Perceived dominance of customer adopters Positive Accepted*** Hypothesis 7 Product characteristics (asset specificity, product complexity) Negative Accepted** Hypothesis 8 Demand uncertainty Negative Rejected Hypothesis 9 Market volatility Positive Rejected Hypothesis 10 Perceived relative advantage Positive Rejected Hypothesis 11 Perceived complexity Positive Rejected Hypothesis 12 Compatibility Positive Rejected *p<0.01, **p<0.05, ***p<0.1 # Relation type of the independent variable with the dependent variable B2B e- marketplace adoption by MSMEs 129

5.5 Services of B2B e-marketplace used by the adopter MSMEs The third objective of the research is to determine the level of adoption of B2B e-marketplace and identify the purpose of use of B2B e-marketplaces. The MSMEs are divided into three stages: exploration stage, trial stage and commitment stage based on their level of participation in the B2B e- marketplace. As shown earlier in Table 5.2, out of total 66 adopters, 37 (56 per cent) are in exploration stage, 18 (27 per cent) are in trial stage and 11 (17 per cent) are in commitment stage. The Table 5.15 shows the users and non-users of various services supported by B2B e-marketplaces. Table 5.15: B2B e-marketplace services used by adopter MSMEs Purpose Users (%) Non-Users (%) Identification of suppliers for our raw materials (Supplier search, RFQ) Identification of new customers (Sell offers, Lead Generation) Providing company and product information (Electronic Catalogue) 28 (42) 38 (58) 51 (77) 15 (23) 66 (100) 0 Finding competitor information 49 (74) 17 (26) Find new business opportunities (Find new trading partners) Find good bargains for products (Buy offers/sell offers) 49 (74) 17 (26) 19 (29) 47 (71) Update on industry news trade fairs 9 (14) 57 (86) Conduct buying selling transactions 0 66 (100) Participate in auctions 0 66 (100) Get credit rating to improve brand image and trustworthiness 11 (17) 55 (83) 130

It is observed that the top four services of B2B e-marketplace used by the adopter MSMEs rank wise are Providing company information (100 per cent) Identification of new customers (77 per cent) Finding new business opportunities (74 per cent) Finding competitor information (74 per cent) It is observed that none of the MSMEs in the sample have conducted buy/sell transactions and participated in auctions. Therefore, it can be concluded that the MSMEs are primarily using B2B e-marketplace for marketing their products rather than conduct transactions. The motive behind the adoption of B2B e-marketplace is to have an online presence. Many MSMEs have standalone websites apart from electronic catalogues in the B2B e-marketplace websites. Among the 66 adopter MSMEs, 65 per cent of the MSMEs also had a website. This clearly indicates that the MSMEs adopt B2B e-marketplaces primarily for marketing their products by placing their product information on multiple portals. As the switching costs for MSMEs are low, adopter MSMEs have registered in multiple B2B marketplace websites. On an average, adopter MSMEs had registered with at least two B2B e-marketplaces. MSMEs motive is primarily to have multiple presence online and promote the company and not conducting transactions. The motives of adoption of B2B e-marketplace are primarily exploration or legitimacy and planned decision making of use B2B e- marketplace is lacking. Most of the popular horizontal B2B e-marketplaces in India such as Indiamart.com and Tradeindia.com have positioned themselves as portals that provide a database of potential buyers and sellers. Few MSMEs (42 per cent) have used B2B e-marketplaces for identifying suppliers for items such as packaging materials and for items that are not available locally. Few (29 per cent) have confirmed using buy and sell offers to fulfill their urgent requirements. 131

A small percentage of MSMEs have used services such as credit rating (17 per cent) and update on industry news and trade fairs (14 per cent). Majority of the MSMEs have their own industry associations and industry level publications which they perceive are more useful. Credit rating service that is important for enhancing trust among buyers and sellers when done by an authorized third party credit rating firm is also a service that is not used in full extent. MSMEs use their own methods to verify the supplier credentials that may be time consuming many a times. The Table 5.16 shows the services used by the MSMEs firm size wise. Some perceptions and statements of adopter MSMEs are as follows 1. We want to promote our company through every possible channel. E- marketplace is one of them. However, sales come only through our hard work and efforts. Manufacturer of Computer storage racks, Bangalore 2. We get lot of queries daily and all of them do not result in sales, Chemicals Manufacturer, Brahmavar. 3. You know it is very important for us to have online presence these days. We get lot of enquiries. Very few have actually converted to sales. We have used e-market for searching for packaging providers, Ice cream manufacturer, Mangalore 4. We have registered our company in e-marketplace, but fulfilling orders from other places becomes difficult for us as the products are bulky and there are other competitors. We could get a new supplier who suggested us an improvement in our product and we have now developed a new product with the material. Water Tank Manufacturer, Kundapur 5. We use e-marketplace for procuring materials that we do not get locally. Toy Manufacturer, Mangalore. 132

Table 5.16: Comparison of B2B e-marketplace services used by adopter MSMEs (firm size wise) Purpose Identification of suppliers for our raw materials (Supplier search, RFQ) Identification of new customers (Sell offers, Lead Generation) Providing company and product information Finding competitor information Find new business opportunities (Find new trading partners) Find good bargains for products (Buy offers/sell offers) Update on industry news trade fairs Conduct buying selling transactions Participate in Users (%) 2 (33) 3 (50) 6 (100) 2 (33) 2 (33) auctions Credit Rating 1 (17) Micro (n=6) Non- Users (%) 4 (67) 3 (50) Small-Group1 (n=10) Users (%) 2 (20) 6 (60) 0 10 (100) 4 (67) 4 (67) 0 6 (100) 0 6 (100) 0 6 (100) 0 6 (100) 5 (83) 6 (60) 4 (40) 3 (30) 1 (10) Non- Users (%) 8 (80) 4 (40) Small- Group2 (n=26) Users (%) 12 (46) 23 (88) 0 26 (100) 4 (40) 6 (60) 7 (70) 9 (90) 0 10 (100) 0 10 (100) 2 8 (20) (80) 20 (77) 20 (77) 9 (35) 4 (15) Non- Users (%) 14 (54) 3 (12) Medium (n=24) Users (%) 12 (50) 19 (79) 0 24 (100) 6 (23) 6 (23) 17 (65) 22 (85) 0 26 (100) 0 26 (100) 2 24 (8) (92) 21 (87) 23 (96) 7 (29) 4 (17) Non- Users (%) 12 (50) 5 (21) 0 3 (13) 1 (4) 17 (71) 20 (83) 0 24 (100) 0 24 (100) 6 18 (25) (75) 133

The majority of the users who are using the services (identifying new customers, finding new business opportunities, finding competitor information is more popular among companies with larger firm size. There is lesser usage of these services among micro and Small-Group1. Since more usage also means more commitment of resources such as marketing expenses and dedicated staff to be deployed to identify leads, follow up and convert the enquiries into sales, smaller firms seem to use these services lesser. 5.6 Benefits of B2B e-marketplace use for adopter MSMEs The fourth objective of the research is to assess the benefits that have been realized by the adopter MSMEs. The Table 5.17 summarises the findings. The results show that MSMEs primarily agree (68 per cent) that the e- marketplace has enabled them to improve their brand image. Few MSMEs have agreed on benefits such as update product information instantly and cost effectively (44 per cent). As majority of the MSME adopters are in exploratory or trial stage and e-marketplace is yet to become a part of their business. The MSMEs in India are in the experimentation stage of using B2B e-marketplace. Therefore, the benefits are still not clear. MSMEs are unable to confirm any quantifiable benefits of B2B e-marketplace use. Forty one percent of MSMEs have agreed that it has enabled them to get new customers and increase sales. But MSMEs opine that the volume of sales through the e-marketplace is very low and many a time they are unable to execute the order due to overhead costs of executing small orders. Since e-marketplace gives lot of choice to buyers, not all MSMEs are able to convert their enquiries into sales. The MSMEs also have to invest considerable amount of resources to undertake marketing through e- marketplaces and follow up the enquiries. 134

Table 5.17: Benefits of B2B e-marketplace use for adopter MSMEs B2B e-marketplace Benefits Mean (S.D.) Improve brand image 3.46 (1.168) Enabled us to get new 2.73 customers and increase sales (1.144) Update product information 2.65 instantly and cost effectively (1.504) Helped us in finding new 2.45 suppliers (1.383) Information on competitors at 1.77 single source (.891) Flexibility in administration and 1.65 communication (.832) Improve customer services 1.62 (.873) Enter supply chains of large 1.52 companies (.561) Percentage of MSMEs that agree Percentage of MSMEs that disagree 68 21 41 54 44 55 33 67 8 85 4 86 8 89 3 97 Thirty three percent of the MSMEs have agreed that it has helped them in finding suppliers. The MSMEs that used e-marketplace for identifying suppliers agreed that they used it to identify suppliers for indirect goods such as packaging material supplier or for a component that is not available in the local market. However, none of the MSMEs used B2B e-marketplaces for identifying suppliers for their raw materials. These are sourced through trusted suppliers with whom the firm has long term relationships. Majority of the MSMEs have disagreed with the benefits such as flexibility in administration and communication, improve customer services, information on competitors at single source, and enter supply chains of large companies. 135

5.7 Barriers to adopt and use B2B e-marketplace by MSMEs The fifth objective of the research is to identify the barriers to adopt and use B2B e-marketplace for both adopters and non-adopters. The Table 5.18 summarises the findings. The top three barriers among adopters and nonadopters are Service providers do not understand our needs, Dependent on traditional intermediaries in trading and Business partners are not ready. Table 5.18: Barriers to adopt and use of B2B e-marketplace by MSMEs Barriers Mean S.D. Service Providers Do not understand our needs Dependent on traditional intermediaries in trading 3.74 1.341 3.61 1.196 Business partners are not ready 3.18 1.247 Not suitable for our product as they have to customized 2.84 1.438 Do not trust transactions 2.68 1.228 Lack of technology standards 2.49.920 Complex To use 2.21 1.294 Expensive 2.15 1.204 Not aware 2.10 1.463 136

MSMEs perceived that B2B e-marketplace service providers did not understand their product, industry requirements and needs of the firm. According to them, sales representatives who visited them once in a year for renewal are only interested in renewal, rather than providing tailor made recommendations for the firm. MSMEs that exported had already established traditional channels and intermediaries and are dependent on them. MSMEs also perceived that the several members in their supply chain are not ready for online transactions. Service providers do not understand our needs and Dependent on traditional intermediaries in trading are common barriers to adopt/use among both adopters and non-adopters. The independent samples t-test results (shown in the Table 5.19) showed that there are no significant differences between the means adopters and non-adopters for these two barriers. Among non-adopters, apart from the first two barriers, Business partners are not ready, Not suitable for our product as they have to customized, do not trust transactions and complex to use emerged as the top barriers. MSMEs that had highly specialized industrial products or OEM manufacturers had large buyers as their regular customers and they worked closely with them to develop customized solutions. For these barriers, there is significant difference between adopters and non-adopters. Some of the nonadopter MSME s perceptions are as follows We get regular orders from branded apparel manufacturers. All the shirts and trousers are customized based on their requirements. Will the e-marketplace assure me orders? Apparel manufacturer, Bommanahalli, Bangalore We supply to customized products to Government organisations. E- marketplace is not relevant to our products. Transformer manufacturer, Bangalore. 137

Table 5.19: Comparison of barriers to B2B e-marketplace adoption and use among adopters and non-adopters Barriers Adopters Mean (S.D.) Non-adopters Mean (S.D.) t-test Service Providers Do not understand our needs Dependent on traditional intermediaries in trading 3.70 (1.488) 3.79 (1.155) -0.363 (p=0.717) 3.7(1.150) 3.5 (1.25).899 (p=0.376) Business partners are not ready 2.55 (1.230) 3.93(0.759) -7.588 (p=0.00) Not suitable for our product as they have to customized 2.41 (1.381) 3.36 (1.341) -3.839 (p=0.00) Do not trust transactions 2.14(1.162) 3.32(.974) -6.127 (p=0.00) Lack of technology standards 2.15 (.899) 2.89 (.779) -4.881 (p=0.00) Complex To use 1.50 (.864) 3.05 (1.212) -8.236 (p=0.00) Expensive 1.71 (1.019) 2.66 (1.210) -4.634 (p=0.00) Not aware 1.20 (.613) 3.16 (1.462) -9.936 (p=0.00) The barriers are compared between MSMEs of different firm sizes. The means of the barriers firm size wise is shown in Table 5.20. One-way between groups ANOVA with post-hoc comparisons is performed to understand if there are any differences in barriers among MSMEs of different firm sizes. The results of the ANOVA analysis (F-Statistic and the corresponding significance level) are also shown in Table 5.20. 138

It is observed that there are significant differences among MSMEs in different firm sizes in all the barriers, except, service providers do not understand our needs, dependent on traditional intermediaries on trading and do not trust transactions. Therefore, these are common barriers across MSMEs in different firm sizes. Post Hoc tests and Tukey s HSD showed that there are significant differences between micro firms and other groups in the barriers complex to use, not aware, expensive. Therefore, these barriers are specific barriers to micro firms. Table 5.20: Comparison of barriers (means) to B2B e-marketplace adoption and use among MSMEs (firm size wise) Barriers Micro Small- Group1 Small- Group2 Medium F (sig.) Service Providers do not understand our needs Dependent on traditional intermediaries in trading Business partners are not ready Not suitable for our product as they have to customized 3.53 4.04 3.84 3.55 0.974 (0.408) 3.57 3.57 3.82 3.45 0.531 (0.662) 3.77 3.50 2.79 2.74 5.814 (0.001) 3.03 3.32 2.79 2.29 2.881 (0.039) Do not trust transactions 3.00 2.75 2.33 2.68 1.609 (0.191) Lack of technology standards 2.83 2.68 2.45 2.03 4.762 (0.004) Complex to use 3.13 2.32 1.73 1.74 9.807 (0.00) Expensive 2.93 2.18 1.85 1.68 7.587 (0.00) Not aware 3.20 2.14 1.64 1.48 10.666 (0.00) 139