E commerce adaption e-commerce e-commerce

Electronic commerce adoption: an empirical study of small and medium US businesses

Elizabeth E. Grandon b , J. Michael Pearson

a,*

a Department of Accounting and Computer Information Systems, School of Business, Emporia State University, Emporia, KS 66801, USA

b Department of Management, College of Business and Administration, Southern Illinois University, Carbondale, IL 62901, USA Accepted 22 December 2003

Available online 9 April 2004

Abstract By combining two independent research streams, we examined the determinant factors of strategic value and adoption of

electronic commerce as perceived by top managers in small and medium sized enterprises (SME) in the midwest region of the US. We proposed a research model that suggested three factors that have been found to be influential in previous research in the perception of strategic value of other information technologies: operational support, managerial productivity, and strategic decision aids. Inspired by the technology acceptance model and other relevant research in the area, we also identified four factors that influence electronic commerce adoption: organizational readiness, external pressure, perceived ease of use, and perceived usefulness. We hypothesized a causal link between the perceived strategic value of electronic commerce and electronic commerce adoption. To validate the research model, we collected data from top managers/owners of SME by using an Internet survey. # 2004 Elsevier B.V. All rights reserved.

Keywords: e-Commerce adoption; Strategic value of e-commerce; SMEs

1. Introduction access narrow markets segments that are widely dis- tributed while buyers can benefit by accessing global

Electronic commerce (e-commerce) has been markets with larger product availability from a variety defined in several ways depending on the context

of sellers at reduced costs. Improvement in product and research objective of the author. For this study,

quality and the creation of new methods of selling we have taken two definitions of e-commerce [44,58]

existing products are also benefits [13] . and adapted them in a B2C context: ‘‘the process of

The benefits of e-commerce are not only for large buying and selling products or services using electro-

firms; small and medium sized enterprises (SMEs) can nic data transmission via the Internet and the www.’’

also benefit from e-commerce [52] . In addition, it can E-commerce provides many benefits to both sellers

‘‘level the playing field’’ with big business, provide and buyers; e.g. Napier et al. [43] pointed out that

location and time independence, and ease communi- by implementing and using e-commerce sellers can

cation [16,29,38,50] . However, in spite of the many potential advantages of e-commerce, its adoption by

* Corresponding author. Tel.: þ1-620-341-5685;

SMEs remains limited. For example, a survey con-

fax: þ1-620-341-6346.

ducted by Verizon [20] found that 36% of small

E-mail address: grandone@emporia.edu (E.E. Grandon).

businesses established web sites primarily to advertise

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

and promote their business, compared to 9% who and argued that executives rely on their perceptions established one to sell or market online. Similarly,

in determining whether a particular IT investment in a survey of 444 SMEs during 2002, Pratt [47] found

creates value for the firm.

that many SMEs were reluctant to conduct transac- The majority of the research has proposed a direct tions on line; more than 80% were only using the

causal link between IT investment and firm perfor- Internet to communicate (via e-mail) and gather busi-

mance. However, Li and Ye [37] empirically tested the ness information. Does this mean that top managers/

moderating effects of environmental dynamism, firm owners of SMEs do not realize the strategic value

strategy, and CIO/CEO relationship on the effect of IT e-commerce to their business or does this mean that

investment on firm performance and found that IT they encounter significant barriers to implementing it?

investment appears to have a stronger positive impact Here, we focused our attention on this ‘‘under-

on financial performance when there are greater envir- studied’’ segment of business organizations [19]

onmental changes, the strategy of the company is more where research findings on large businesses cannot

proactive, and closer CIO/CEO ties. In a similar line of

be generalized; e.g. Welsh and White [66] identified inquiry, Lee [36] created a multi-level value model important differences in the financial management

that connects the use of IT to a firm’s profit; she of small and large businesses while Ballantine et al.

pointed out that the effect of incorporating IT should [5] identified unique characteristics of SMEs as lack

not be considered alone and argued that there are other of business and IT strategy, limited access to capital

variables that can influence the relationship. Her IT resources, greater emphasis on using IT and IS to

business value model incorporated other variables, automate rather than informate, influence of major

such as origination cost, cycle time, loan officer customers, and limited information skills. Similar

retention, control over external partners, and market- assertions and findings are given in other papers

ing effort and she found that IT can reduce cycle time [14,18,32,40,46,51] .

and cost, and change the way business is run. She concluded that ‘‘one has to know what other variables to manage and how to manage them in order to make

2. Literature review

IT investments profitable.’’

Few studies have focused on the perceptions of top This study represents a fusion of two independent

management regarding the strategic value of e-com- research streams: the strategic value of certain infor-

merce. Amit and Zott [4] is one of the few that has mation technologies to top managers and factors that

tried to deal with this and even though they focused influence the adoption of IT. The former has been

on e-business, their results can be generalized to studied by Subramanian and Nosek [60] and others

e-commerce [28] . They examined how 59 American (e.g. [6,11] ) while the latter has been investigated by

and European publicly traded e-business firms create Davis [21] and others (e.g. [1,30,35,65] ) primarily

value. Approximately, 80% were SMEs (with less through the technology acceptance model (TAM).

than 500 employees). They developed a value-drivers model which included four factors found to be sources

2.1. Perceived strategic value of IT of value creation: transaction efficiency, complemen- tarities, lock-in, and novelty. Some of these factors are

Many studies have focused on the relationship also found in Saloner and Spence’s [56] work. between IT investment and firm’s performance in large

Through an empirical study of 73 firms (some of corporations. For example, Hitt and Brynjolfsson [27]

them SMEs), Subramanian and Nosek identified three investigated how IT affects productivity, profitability,

factors that were found to create strategic value in IS: and consumer surplus. They found that IT increases

operational support, managerial productivity, and stra- productivity and consumer surplus but not necessarily

tegic decision aid. In each of these factors they utilized business profits. Barua et al. concluded that the pro-

different items that were found to have high conver- ductivity gains from IT investments have generally

gent validity and reliability. Their factors seem to be been neutral or negative, while Tallon et al. [62]

applicable to e-commerce. Due to a lack of research measured IT payoffs through perceptual measures

in identifying factors that create strategic value of

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

e-commerce, their model was used as the basis for the Igbaria et al. determined the factors affecting strategic value portion of this study.

personal computer acceptance in small businesses. Among the factors that directly influence personal

2.2. Information technology adoption computer acceptance were perceived ease of use and perceived usefulness. The intra-organizational (inter-

Davis proposed TAM, a model that has been tested nal computing support and training, and management in many studies (e.g. [26,59,61] ). Lederer et al. sum-

support) and extra-organizational (external computing marized sixteen articles that tested the model for

support and training) variables were hypothesized to different technologies (e.g. ATM, e-mail, Netscape,

influence adoption through perceived usefulness and Access, Internet, Word, and Excel). In their model,

ease of use. Inconsistent with research in large firms, they considered beliefs about ease of use and per-

relatively little support was found for the influence of ceived usefulness as the major factors influencing

internal support and training on perceived ease of use attitudes toward use, which, in turn, affected intentions

and usefulness. However, perceived ease of use turned to use.

out to be an important factor in explaining perceived Many other studies have attempted to describe the

usefulness and system usage. It was also found that factors influencing IT adoption in SMEs. For example,

perceived usefulness is a strong antecedent of system Iacovou et al. studied factors influencing the adoption

usage.

of electronic data interchange (EDI) by seven SMEs in In order to develop an integrated model of IS different industries; they included perceived benefits,

adoption in SMEs, Thong [63] specified four contex- organizational readiness, and external pressure. To

tual variables as primary determinants of IS adoption. measure perceived benefits they used awareness of

He highlighted the fact that the technological innova- both direct and indirect benefits. Variables measuring

tion literature has identified many variables as possible organizational readiness were the financial and tech-

determinants of organizational adoption but this nological resources. In order to measure external

‘‘suggest that more research is needed to identify pressure, they considered competitive pressure and

the critical ones’’ and provided four groups of vari- its imposition by partners. The results suggested that

ables: CEO, IS, organizational characteristics, and

a major reason that small firms become EDI-capable is environmental characteristics. due to external pressure (trading partners). In a similar

Based on the literature, Premkumar and Roberts study, Chwelos et al. [17] considered the same factors

[49] identified the use of various communication influencing the adoption of EDI in 286 SMEs. They

technologies and the factors that influence their adop- considered the trading partner as influencing external

tion in small businesses located in rural US commu- pressure and readiness while external pressure was

nities. The technologies studied included EDI, online considered to be influenced by the dependency on

data access, e-mail, and the Internet. The factors trading partner and enacted trading partner power. As

studied as potential discriminators between adopters in the case of Iacovou et al., external pressure was the

and non-adopters of communication technologies most important factor contributing to intent to adopt

were grouped into three broad categories: innovation, EDI. Kuan and Chau [34] determined the factors

organizational, and environment characteristics. influencing the adoption of EDI in small businesses

Within the innovation factor, they included relative using a technology, organization, and environment

advantage, cost, complexity, and compatibility. Orga- framework. The technology factor incorporated per-

nizational characteristics included top management ceived direct and indirect benefits of EDI. The orga-

support, and IT expertise. Finally, within the environ- nization factor consisted of perceived financial cost

mental characteristics variable, competitive pressure, and perceived technical competence. The environ-

external support, and vertical linkages were consid- ment factor was similar to external pressure in

ered. The results suggested that relative advantage, top Iacovou et al.’s study but included a new variable:

management support, and competitive pressure were perceived government pressure. There, perceived

factors influencing the three communication technol- indirect benefits were not found to be a significant

ogies. Compatibility, complexity, external pressure, factor.

and organizational size were found to be significant

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

discriminators between adopters and non-adopters of executives on the adoption of e-commerce. They online data access technology. Cost was found to be an

found that all the component items of the normative important discriminant factor only for the adoption

and control beliefs differentiated between adopters of the Internet. IT expertise was not found to be an

and non-adopters. In the behavioral beliefs (attitude) important factor that discriminates between adopters

group, however, only some items (e-commerce and non-adopters. Finally, vertical linkage was found

enhances the distribution of information, improves to be an important discriminant factor for online data

information accessibility, communication, and the access and the Internet adoption.

speed with which things get done) were found The adoption of the Internet was also studied by

to differentiate adopters from non-adopters. Table 1 Mehrtens et al. [41] . In order to develop a model of

summarizes the factors involved in the process of Internet adoption, they conducted a case study on

technology adoption.

seven SMEs. First, they considered four SMEs that had adopted the Internet. Based on Iacovou et al.’s

2.3. Causal link

work and the results of the preliminary analysis, they devised their model using perceived benefits, organi-

Support for the causal link between perceptions zational readiness, and external pressure as determi-

of strategic value and adoption comes from different nant factors. Then, to provide theoretical replication

studies that associate individual perceptions and beha- they considered three non-IT SMEs, of which two had

vior. The theory of planned behavior (TPB) is a well adopted the Internet and one had not. All the factors

established intention model that has been proven were found to affect Internet adoption by the small

successful in predicting and explaining behavior firms. Chang and Cheung [12] also determined factors

across a wide variety of domains, including the use that influence Internet/www adoption with similar

of information technology [2] . In general terms, the results.

TPB establishes that perceptions influence intentions In a more recent study and following a similar line

which in turn influence the actual behavior of the of inquiry, Riemenschneider et al. [55] studied the

individual. By considering the intention to adopt factors that influence web site adoption by SMEs.

e-commerce as the target behavior, the use of intention They proposed a combined model using the theory

models theoretically justifies the causal link between of planned behavior (TPB) [3] and TAM. They tested

perceptions and adoption of e-commerce. individual models, partially integrated models, and

This causal link has been studied and the results fully integrated models by using structural equation

indicate that managers’ perception and attitudes modeling. They found that the combined model pro-

toward other types of IT are strongly associated with vided a better fit.

its use; e.g. hypotheses developed by Jarvenpaa and The emerging field of e-commerce has not been

Ives [31] suggested a CEOs involvement in IT and ignored in the analysis of adoption. Mirchandani and

active personal participation in IT management were Motwani [42] investigated the factors that differentiate

associated with a firm being progressive in its use of adopters from non-adopters of e-commerce in small

IT. They defined involvement as ‘‘CEOs perceptions businesses. The relevant factors included enthusiasm

and attitudes concerning IT’’ and participation as of top management, compatibility of e-commerce with

‘‘the CEOs activities or substantive personal inter- the work of the company, relative advantage perceived

ventions in the management of IT.’’ They conducted a from e-commerce, and knowledge of the company’s

survey study involving 83 firms from four different employees about computers. The degree of depen-

industries and found strong support for the relation- dence of the company on information, managerial

ship between involvement (favorable perceptions of time required to plan and implement the e-commerce

IT) and progressive use of the IT. However, the link application, the nature of the company’s competition,

between CEOs participation in IT management and as well as the financial cost of implementing and

progressive use of the IT was found to be moderate. operating the e-commerce application were not

Similar research was performed by Sanders and influencing factors. Similarly, Riemenschneider and

Courtney [57] , Reich and Benbasat [53] , and Bush McKinney [54] analyzed the beliefs of small business

et al. [8] .

Table 1 Summary of factors of IT adoption in SMEs

E.E. Source

Influencing factors

IT studied

No. SMEs a Industries Gr andon,

Iacovou et al. [29] External pressure, perceived benefits, organizational

7 different industries readiness Chwelos et al. [17]

EDI adoption practices

7 SMEs (n < 200)

Manufacturing, services, J.M.

Readiness, external pressure, perceived benefits

EDI adoption

268 SMEs (n/a)

government, etc. P earson Kuan and Chau [34]

Not specified Igbaria et al. [30]

Technology organization environment

EDI adoption

575 SMEs (n < 100)

Intra-organizational factors, extra-organizational

Manufacturing and engineering factors, perceived ease of use perceived usefulness

Personal computer

203 SMEs (n < 100)

Information / Thong [63]

Not specified characteristics, environmental characteristics Premkumar and

CEO characteristics, IS characteristics, organizational

IS adoption

166 SMEs (n < 100)

Manufacturing, retail sales and Roberts [49]

Relative advantage, top management support,

Online data access, e-mail,

78 SMEs (n < 90)

organizational size, external competitive pressure

and the Internet

wholesale trade, service, finance, & insurance, others

Manage Mehrtens et al. [41]

Perceived benefits, organizational readiness,

IT industry, clothing manufacturer, external pressure

Internet adoption

7 SMEs (n < 200)

entertainment, transport Mirchandani and

Not specified ment Motwani [42]

Enthusiasm of top management, compatibility,

E-commerce adoption

62 SMEs (n < 200)

relative advantage knowledge of the company’s employees about computers

42 Riemenschneider and

Defense, agriculture, oil and (2004) McKinney [54]

Attitude, subjective norm, perceived behavioral control

E-commerce adoption

184 SMEs (n < 500)

gas, manufacturing Riemenschneider

Service/sales, government, retail, 197 et al. [55]

Attitude, subjective norm, perceived behavioral

Web site adoption

156 SMEs (n < 500)

banking, medical, manufacturing –216 a n represents the maximum number of employees considered in the criteria to define a SME.

control, perceived usefulness, perceived ease of use

(web presence)

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Organizational Readiness Organizational Support

External Pressure

Managerial

Percep. of

Productivity

Strat. Value

Adoption

Perceived Ease of Use

Strategic Dec. Aids

Perceived Usefulness

Fig. 1. The proposed research model.

3. Research model face validity into four different variables: ‘‘organiza- tional readiness, external pressure, perceived ease of

Based on the literature review, we proposed a use, and perceived usefulness’’ (see Table 2 ). research model ( Fig. 1 ).

Organizational readiness was assessed by including two items about the financial and technological

3.1. Perception of strategic value of e-commerce resources that the company may have available as well as factors dealing with the compatibility and

We considered three major variables as sources of consistency of e-commerce with firm’s culture, values, strategic value of e-commerce: ‘‘operational support,

and preferred work practices (existing technology managerial productivity, and strategic decision aids.’’

infrastructure; and top management’s enthusiasm to Since the instrument utilized by Subramanian and

adopt e-commerce). Such items were found relevant in Nosek was found to have high reliability (Cronbach

other research [7,15,48,64] .

alpha ¼ 0:82) with convergent and discriminant valid- External pressure was assessed by incorporating ity, we used their items to measure the strategic value

five items: competition, social factors, dependency on construct. Operational support measures how e-com-

other firms already using e-commerce, the industry, merce can reduce costs, improve customer services

and the government.

and distribution channels, provide effective support We considered a subset of Davis’ instrument to mea- role to operations, support linkages with suppliers, and

sure perceived ease of use as modified to make them increase ability to compete. Managerial productivity

relevant to e-commerce. We utilized the six items for per- suggests how e-commerce can enhance access to

ceived usefulness of Davis as modified to fit our research. information, provides a means to use generic methods in decision-making, improves communication in the

3.3. Research questions

organization, and improves productivity of managers. Finally, strategic decision aids defines how e-com-

The questions we explored were used to validate the merce can support strategic decisions of managers,

proposed two-step model and understand the relation- support cooperative partnerships in the industry, and

ship between these two steps. provide information for strategic decisions.

1. What are the determinant factors of the perceived

3.2. Factors influencing adoption of e-commerce strategic value of e-commerce in SME?

2. How do the perceptions of strategic value, as viewed We identified the factors found significant in prior

by top managers/owners of SMEs, influence their research and grouped them based on similarity and

decision to adopt e-commerce?

203 Table 2

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Summary of adoption factors in the current study Factor in the current study

Factors in previous studies

Source

Organizational readiness

Organizational readiness

Iacovou et al. [29]

Readiness

Chwelos et al. [17]

Organization

Kuan and Chau [34]

Organizational readiness

Mehrtens et al. [41]

Facilitating conditions

Chang and Cheung [12]

Compatibility with company

Mirchandani and Motwani [42]

Compatibility

Thong [63]

Igbaria et al. [30] External pressure

Intra/extra organizational factors

External pressure

Iacovou et al. [29]

External pressure

Chwelos et al. [17]

Environment

Kuan and Chau [34]

Social factors

Chang and Cheung [12]

External pressure

Mehrtens et al. [41]

External competitive pressure

Premkumar and Roberts [49]

Subjective norm

Riemenschneider and McKinney [54]

Riemenschneider et al. [55] Perceived ease of use

Subjective norm

Perceived ease of use

Davis [21]

Perceived ease of use

Igbaria et al. [30]

Riemenschneider et al. [55] Perceived usefulness

Perceived ease of use

Perceived usefulness

Davis [21]

Perceived ease of use

Igbaria et al. [30]

Perceived ease of use

Riemenschneider et al. [55]

3. What are the factors involved in the decision to

4.2. Data collection

adopt e-commerce by top managers/owners of SMEs?

The data were gathered by means of an electronic survey administered during Spring 2002. The process was carried out in three steps. First, a sample of 1069

4. Methodology small and medium size businesses were identified from various sources that focus on SMEs. We identi-

4.1. Subjects fied the company name, a contact person, an e-mail address for that person, an address, and a telephone

We targeted top managers of small and medium size number. The contact person was typically the owner of business from a variety of industries in the midwest

the business or a top-level manager. Second, an initial region of the US. In our study, we considered the

e-mailing that identified the purpose of the study, a number of employees as the principal criterion in

request to participate, and an opt-out feature was sent determining whether a firm qualified as an SME since

to all potential respondents; 136 of these messages other categorizations involving revenue, total capital

were returned due to incorrect addresses or that the and/or other types are more difficult to apply and can

organization was no longer in business. An additional result in misleading classifications. The number of

101 individuals indicated that they were unable or employees varies according to the agency providing

were unwilling to participate. the definition. For example, the US Small Business

Thirdly, approximately one week after the initial Administration (http://www.sba.gov) uses a cut-off of

mailing, a second electronic mailing was sent to fewer than 500 employees. Harrison et al. [25] and

the remaining 832 potential respondents. This elec- Iacovou et al. utilized a cut-off of 200 employees. For

tronic message directed them to the web site where this study, we have used ‘‘less than 500 employees.’’

the survey instrument was located. One hundred

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

individuals completed the survey for a response rate

Table 3

of 12%. Possible explanations for this relatively low

Demographics of study (n ¼ 100)

response rate could include the lack of relevance of the

Gender

topic to the respondent, delivery method of the instru-

Male

ment (electronic), and the time of the year at which the

Female

survey request took place.

Education

High school

4.3. Instrument development

2-year college

4-year college

Master degree

Three top managers participated in a pilot of the

Other

survey instrument. One of the authors observed the

Internet service provider in place

subjects as they completed the survey. Feedback

Yes

from them resulted in minor changes to the survey

No

instructions and questions. Respondents were required

Electronic commerce already in place

to complete the survey that had the following major

Yes

sections (see Appendix A for the complete instru-

Seven demographic questions (respondent’s gender,

age, education, years of work in present position,

and years of work in present firm).

Two general questions about the firm (number of

Industry

employees and industry.

Education

Four questions about the technology in the organi-

Finance

zation (number of PCs, presence of Internet server

Wholesale

provider, presence of web site, and utilization of

Fifteen questions asking the extent to which

Insurance

e-commerce is perceived as contributing to strategic

Organization has web site

Twenty-three questions to measure the factors

Yes

involved in e-commerce adoption.

No

A seven-point Likert scale (from strongly disagree to strongly agree) was utilized to measure the ques- tions about perceived strategic value and adoption of

5.2. Statistical analysis

e-commerce. In order to test the model, a statistical analysis was conducted in two stages. The first step employed

5. Results confirmatory factor analysis to measure whether the number of factors and loadings of items involved in

5.1. Demographics and descriptive statistics the two main constructs (perceived strategic value and adoption) conform to the proposed model. With this

The 100 surveys were returned over a 4-week analysis, we found answers to research questions 1 period. Results indicated that the top managers were

and 3.

well educated, with over 56% holding a 4-year college Since we were also interested in exploring how the degree or a masters degree. The majority were male

perceptions of strategic value influence the decision to (64%) and 36% were between 41 and 50 years of age.

adopt e-commerce (research question 2), canonical Table 3 shows other demographics.

analysis was utilized in the second step. This technique

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

involves developing a linear combination of indepen-

Table 4

dent variables (strategic value variables) and dependent

Rotated component matrix

variables (adoption variables) to maximize the correla-

Component

tion between the two sets [24] . MIS research has

benefited from the use of this multivariate technique (see, for example [9,33] ). Campbell and Taylor [10]

demonstrated that canonical analysis subsumes other

statistical procedures. 0.001

Non-response is a potential source of bias in survey

studies; it needed to be properly investigated [22] . The

potential bias was evaluated by comparing responses

between early and late respondents. Early respondents

were those who had completed the questionnaire 0.380

within the initial 2-weeks, while late respondents were

those who completed it after the specified period.

Approximately 70% of the responses were from early

respondents. Demographic data was utilized for this purpose: number of employees, number of years in the current position, number of years in the firm, and

The rotated component matrix in Table 4 shows that number of personal computers in the company. No

all the items loaded cleanly on their intended factors. significant differences were found between the early

Six items loaded cleanly on the organizational support and late respondent groups, suggesting no non-

factor, four items on the managerial productivity response bias.

factor, and three on the decision aids factor. Conver- gent and discriminant validity was assessed via factor

5.3. Confirmatory factor analysis analysis. Table 4 shows that all items have loading greater than 0.50 and loaded stronger on their asso-

5.3.1. Perceived strategic value construct ciated factors than on others. Thus, convergent and

A confirmatory factor analysis was run using SPSS discriminant validity were demonstrated.

10.1. All items measuring the perception of strategic Construct reliability or internal consistency was value of e-commerce were considered during the first

assessed using Cronbach’s alpha. Table 5 shows that run and resulted in one item not loading on the

the values for alpha vary from 0.88 to 0.95. The scale intended factors. This item was dropped from subse-

reliabilities are unusually good compared to the accep- quent analysis and the construct was recalculated. The

table 0.7 level for field research [45] . items in the final analysis for perceived strategic value are shown in Appendix B .

5.3.2. Adoption construct

The factor analysis used principal components in The adoption construct initially consisted of 23 order to extract the maximum variance from the items.

items. In order to test how these items loaded, another To minimize the number of items that have high

factor analysis was run. Principal component extraction loadings on any given factor, a varimax rotation

with varimax rotation and required Eigenvalues above was utilized. Using the Kaiser Eigenvalues criterion,

1.0 were considered. As in the case of the perceived we extracted three factors that collectively explained 79.4% of the variance in all items. Hair et al. provide

Table 5

guidelines for identifying significant factor loadings

Reliability analysis

based on sample size. In order to obtain a power level

of 80% at a 0.05 significant level, with standard errors Cronbach’s alpha assumed to be twice those of conventional correlation

Construct

Operational support (OS)

coefficient, a factor loading of 0.50 or higher should be

Managerial productivity (MP)

considered as a cut-off value.

Decision aids (DA)

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Organizational Readiness

Organizational Support

External Pressure

Managerial

Percep. of

Productivity

Strat. Value

Adoption

Compatibility

Perceived Strategic

Ease of Use Dec. Aids

Perceived Usefulness

Fig. 2. The revised research model.

strategic value construct, the first run of the factor of multiple dependent variables and multiple indepen- analysis resulted in items that did not load as expected

dent variables. By simultaneously considering both, on the intended factors. Two items were dropped from

it is possible to control for moderator or suppressor the analysis and the construct was recalculated.

effects that may exists among various dependent The results of this confirmatory factor analysis

variables [39] .

resulted in five factors loading cleanly with a total In canonical analysis there are criterion variables explained variance of 74.9%. Thus, we revised the

(dependent variables) and predictor variables (inde- proposed model and considered a fifth factor, which

pendent variables). The maximum number of cano- we named ‘‘compatibility’’ to better describe the items

nical correlations (functions) between these two sets used (see Fig. 2 ). The results are quite interesting.

of variables is the number of variables in the smaller Previous research has found compatibility an impor-

set [23] . In our case, the number of variables for tant factor that influences the adoption of IT. In our

the perception of strategic value construct is three study, compatibility emerged freely as a significant

while the number of variables in the adoption con- independent factor.

struct is five. Thus, the number of canonical functions The items considered in the final instrument are

extracted from the analysis is three; i.e., the smallest shown in Appendix C . The following table shows the

set.

rotated component matrix. In order to test the significance of the canonical Convergent and discriminant validity was achieved.

functions we followed the guidelines given by Hair Table 6 shows that all items have loading greater than

et al. They suggest three different measures to inter-

0.50. They also loaded stronger on their associated pret the canonical functions: factors than on others. Thus, convergent and discri-

(a) the significance of the F-value given by Wilk’s minant validity were demonstrated. Table 7 shows that lambda, Pillai’s criterion, Hotteling’s trace, and alpha values range from 0.76 to 0.95 for the perceived

Roy’s gcr;

usefulness of e-commerce factor. As in the case of the (b) the measures of overall model fit given by the strategic value construct, the reliability of the adoption size of the canonical correlations; and construct turned out to be very high. (c) the redundancy measure of shared variance.

5.4. Canonical analysis Table 8 shows the corresponding multivariate test of significance with 15 degrees of freedom while Table 9 Canonical analysis is a multivariate statistical

shows the measures of overall model fit in the three model that studies the interrelationships among sets

canonical functions. Note that the strength of the

207 Table 6

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Rotated component matrix Component 1 2 3 4 5 PU4

relationship between the canonical covariates is given are statistically significant at the 0.01 level, from by the canonical correlation.

the overall model fit ( Table 9 ) it can be concluded Even though the multivariate test of significance

that only the first canonical function is significant shows that the canonical functions, taken collectively,

(P < 0:01). This conclusion is consistent with the canonical R 2 values showed in Table 9 . For these data, in the first canonical function the independent variables

Table 7 Reliability analysis

explain approximately 42% of the variance in the dependent variables; the second canonical function

Construct

Cronbach’s alpha

explains approximately 7%, and the third one explains

Organizational readiness (OR)

0.81 only 1.5%. This is not unusual since typically the first

Compatibility (CC)

0.88 canonical function is far more important than the

External pressure (EP)

0.76 others.

Ease of use (EU)

Perceived usefulness (PU)

0.95 Even though the first canonical function was deemed to be significant, it has been recommended

that redundancy analysis be utilized to determine

Table 8 Multivariate test of significance

Table 9 Measures of overall model fit

Test name Value Approx.

Hypoth.

Error

Sig. of

F DF DF F Canonical

Canonical

Canonical F 2 -statistic Probability

function

correlation

Pillais 0.501 3.529

Hotellings 0.801 4.523

4.028 0.000 Wilks

0.986 0.448 Roys

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Table 10 Canonical redundancy analysis

Canonical Variable

Proportion of total function

Share variance

Canonical R 2 Redundancy

redundancy (%) 1 Dependent

which functions to use in the interpretation. Redun- variable is irrelevant in determining the relationship dancy is the ability of a set of independent variables, to

or that it has been partialed out of the relationship explain the variation in the dependent variables taken

because of a high degree of multicollinearity.’’ Cano- one at a time. Table 10 summarizes the redundancy

nical weights are also considered to have low stability analysis for the dependent and independent variables

from one sample to another. As in the case of weights, for the three canonical functions. The results indicate

canonical loadings are subject to considerable varia- that the first canonical function accounts for the high-

bility from one sample to another. For that reason, and est proportion of total redundancy (94.7% including

in order to increase the external validity of the find- both dependent and independent variables), the second

ings, the canonical cross-loadings method has been one accounts for 3.5%, and the third one accounts only

chosen.

for 1.8%. In addition, the redundancy indexes are These correlate each of the original observed depen- higher for the first canonical function than for the

dent variables directly with the independent canonical second. Therefore, only the first canonical function is

variate, and vice versa. Table 11 shows that almost all considered for interpretation.

of the canonical cross-loadings are significant for both In order to interpret the selected canonical function,

dependent and independent variables (cut-off >0.3) three methods were employed: canonical weights,

with the exception of organizational readiness (OR). canonical loadings, and canonical cross-loadings.

The rank order of importance (determined by the Table 11 shows the summary of these methods for

absolute value of the canonical cross-loadings) for the first canonical function considering both indepen-

the perceived strategic value of e-commerce were dent and dependent variables.

organizational support (OS), managerial productivity The interpretation of canonical weights is subject

(MP), and decision aids (DA). Similarly, the rank of to some criticism. For example, Hair et al. stated, ‘‘a

importance for the adoption construct contributing to small weight may mean either that its corresponding

the first canonical function were perceived usefulness

Table 11 Standardized canonical coefficients and canonical loadings for strategic value and adoption

Canonical cross-loading Perceived strategic

Construct

Variable

Canonical weights

Canonical loading

0.633 Value

OS

MP

0.434 Adoption

DA 0.056

OR

CC 0.206

EP

EU

PU

209 Table 12

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

considered in the perception of strategic value con-

Sensitivity analysis of the canonical correlation results

struct (organizational support, managerial producti-

Complete

Results after deletion

vity, and decision aids) were found to be significant

variate

DA in explaining the perceptions of strategic value of e-commerce. The scale reliability was found to be

Most of the factors proposed as determinants of

Independent variate (canonical cross-loadings)

e-commerce adoption: perceived usefulness, perceived

ease of use, compatibility, and external pressure were

found to be statistically significant as determinants of

Shared variance 0.658

e-commerce adoption. These results corroborate the

TAM model in the sense that perceived usefulness

Dependent variate (canonical cross-loadings)

and perceived ease of use turned out to be the most

influential factors of e-commerce adoption as per-

ceived by top managers of SMEs. The results also

confirm the studies of Igbaria et al. regarding the

factors that influence personal computer adoption

Shared variance 0.381

and Riemenschneider et al. concerning the factors

that influence web site adoption, both in the context of SMEs.

Compatibility between e-commerce and firm’s cul- (PU), ease of use (EU), compatibility (CC), and exter-

ture, values, and preferred work practices was also nal pressure (EP). Organizational readiness (OR)

found to be an influential factor in our study. Results seemed to be a non-important factor in the adoption

confirmed the earlier studies in which compatibility construct.

was considered an important factor in determining adoption. In our study, compatibility emerged freely as

5.5. Sensitivity analysis an independent factor that highly influenced e-com- merce adoption.

To validate the results, sensitivity analysis of the Regarding external pressure, our study validated independent variable set was performed. Table 12

previous research but contrary to what was expected, shows the results of the analysis after removing inde-

organizational readiness, which includes the financial pendent variables (one at a time). In this analysis

and technological resources to adopt e-commerce, was canonical cross-loading were examined for stability.

not found to be a significant factor in the decision. It can be seen from the table that the canonical cross-

Finally, from the canonical analysis it can be con- loadings are fairly stable when the independent vari-

cluded that managers who have positive attitude ables are deleted one at a time. The canonical correla-

toward the adoption of e-commerce also perceived

e-commerce as adding strategic value to the firm. fairly stable. Similarly, the shared variance and redun-

tions (R) as well as the canonical roots (R 2 ) remain

Thus, interventions toward changing managers’ per- dancy indexes were found to be stable when removing

ceptions about the strategic value of e-commerce can some of the independent variables. Thus, the sensi-

be devised in order to increase the adoption/utilization tivity analysis as a whole supported the validity of the

of e-commerce by SMEs.

canonical function.

6.1. Limitations

6. Discussion Generalizations from this research should be made with caution. The main limitation corresponds to the The confirmatory factor analysis corroborated

number of employees considered in each company. Subramanian and Nosek’s results: all the variables

Our sample is mainly of companies whose number of

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

employees varies between 10 and 200. Only five firms The canonical results reveal a significant rela- had more than 200. Thus the sample may be biased

tionship between the perceived strategic value of toward smaller firms.

e-commerce variables and the factors that influence e-commerce adoption in SMEs. This means that those top managers who perceived e-commerce as adding

7. Conclusions strategic value to the firm have a positive attitude toward its adoption. From the canonical analysis, we

Throughout this study we attempted to build a conclude that the three factors proposed as determi- model that explains how perceived strategic value

nants of perceived strategic value of e-commerce have of e-commerce influences managers’ attitudes toward

significant impact on managers’ attitudes toward e- e-commerce adoption. By studying two different

commerce adoption with organizational support and streams of research, we have proposed and validated

managerial productivity as the most influential. Over-

a predictive model that suggest three factors as deter- all, we expect that the results will help managers’ minants of the perceived strategic value of e-com-

understanding of the relationship between the percep- merce and five determinant factors for e-commerce

tions of strategic value of e-commerce and its future adoption in SMEs.

adoption.

Appendix A. The survey

Ecommerce, is defined here as ‘‘the process of buying and selling products or services using electronic data transmission via the Internet and the www.’’ Examples that do not fit this definition include electronic publishing to promote marketing, advertising, and customer support. The mere use of electronic mail or the use of a web site for electronic publishing purposes does not constitute ecommerce according to the definition above.

Section 1 : General information Gender

Male

Female

Age Education

High school

2-year college 4-year college

Other Years in present position

Master/MBA

Doctorate

Years with present firm Total number of employees

Industry in which your firm operates

Construction Transportation

Insurance

Other

Number of PCs in the firm Does your firm have an Internet service provider?

Yes

No

Does your firm have a web site?

Does your firm utilize electronic commerce?

Yes

No

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Section 2 : The following questions ask you about your perceptions of strategic value of electronic commerce. Please indicate your agreement with the next set of statements using the following rating scale.

1 2 3 4 5 6 7 Strongly Disagree

Somewhat Neutral Somewhat Agree Strongly disagree

agree In order to provide strategic value to

disagree

agree

Agree our organization, electronic commerce should help

Disagree

1 Reduce costs of business operations

2 Improve customer services

3 Improve distribution channels

4 Reap operational benefits

5 Provide effective support role to operations

6 Support linkages with suppliers

7 Increase ability to compete

1 2 3 4 5 67 In order to provide strategic value to our

Agree organization, electronic commerce should help

Disagree

8 Provide managers better access to

1 2 3 4 5 67 information

9 Provide managers access to methods and

1 2 3 4 5 67 models in making functional area decisions

10 Improve communication in the organization

11 improve productivity of managers

1 2 3 4 5 67 In order to provide strategic value to our

Agree organization, electronic commerce should help

Disagree

12 Support strategic decisions of managers

13 Help make decisions for managers

14 Support cooperative partnerships in

1 2 3 4 5 67 the industry

15 Provide information for strategic decision

Section 3 : The following questions ask you about your perceptions of adopting electronic commerce. Please indicate your agreement with the next set of statements using the same rating scale above.

Disagree

Agree

1 Our organization has the financial resources

1 2 3 4 5 6 7 to adopt electronic commerce

2 Our organization has the technological

1 2 3 4 5 6 7 resources to adopt electronic commerce Our organization perceives that electronic commerce is consistent with ...

212 E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Section 3 : (Continued )

5 preferred work practices

6 Electronic commerce would be consistent with

1 2 3 4 5 6 7 our existing technology infrastructure

1 2 3 4 5 6 7 adoption of electronic commerce

7 Top management is enthusiastic about the

8 Competition is a factor in our decision to adopt

1 2 3 4 5 6 7 electronic commerce

9 Social factors are important in our decision to

1 2 3 4 5 6 7 adopt electronic commerce

10 We depend on other firms that are already

1 2 3 4 5 6 7 using electronic commerce

11 Our industry is pressuring us to adopt

1 2 3 4 5 6 7 electronic commerce

12 Our organization is pressured by the

1 2 3 4 5 6 7 government to adopt electronic commerce

1 2 3 4 5 6 7 would be easy for me

13 Learning to operate electronic commerce

1 2 3 4 5 6 7 flexible to interact with

14 I would find electronic commerce to be

15 My interaction with electronic commerce

1 2 3 4 5 6 7 would be clear and understandable

16 It would be easy for me to become skillful at

1 2 3 4 5 6 7 using electronic commerce

17 I would find electronic commerce easy to use

18 Using electronic commerce would enable

1 2 3 4 5 6 7 my company to accomplish specific tasks more quickly

1 2 3 4 5 6 7 my job performance

19 Using electronic commerce would improve

20 Using electronic commerce in my job

1 2 3 4 5 6 7 would increase my productivity

21 Using electronic commerce would enhance

1 2 3 4 5 6 7 my effectiveness on the job

22 Using electronic commerce would make

1 2 3 4 5 6 7 it easier to do my job

1 2 3 4 5 6 7 in my job

23 I would find electronic commerce useful

24 I would like to receive the aggregated results Yes No of this survey

25 I am interested in participating further in Yes No this study

E.E. Grandon, J.M. Pearson / Information & Management 42 (2004) 197–216

Thank you for completing this survey. We recognize that your time is limited and we value your participation. Please complete the following section if you answered YES to either question 24 or 25 and you would prefer to be