Industrial Marketing Manage

Understanding corporate B2B web sites’ effectiveness from

North American and European perspective

Goutam Chakrabortya,*, Prashant Srivastavaa,1, David L. Warrenb

aCollege of Business Administration, Oklahoma State University, Stillwater, OK 74078-4011, United Statesb7174 Cottonwood Rd., Sapulpa, OK 74066, United States

Received 29 November 2003; received in revised form 5 August 2004; accepted 19 September 2004

Available online 23 November 2004

Abstract

The globalization of economic activities has led multinational business-to-business (B2B) firms to use their corporate web sites

extensively to communicate with their stakeholders. Using a field study of 368 business customers, this paper examines three factors of

information that influence corporate web sites’ effectiveness in a B2B context and whether these factors differ by web site user’s national

origin. Results indicate that informativeness of a corporate web site is important for both North American and European visitors. Quality of

information is important for European users but not an important aspect for North American users. Usability is the most important factor for

North American users but it is not important for European users. Implications of these results are discussed for managerial practices and

further research.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Corporate web site; B2B; Informativeness; Usability; Quality of information; Web site effectiveness; Internet; Multinational

1. Introduction

The globalization of economic activities and widespread

availability of Internet access across the world has led

multinational firms to use their corporate web sites to

communicate extensively with visitors from different parts

of the world. However, many companies face the challenge

of how to design a corporate web site that will be perceived

as effective in a diverse multicultural environment. This

manuscript investigates the factors that influence corporate

web sites’ effectiveness in a business-to-business (B2B)

context and whether these factors differ by the web site

visitor’s nationality (in particular, North America versus

Europe).

0019-8501/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.indmarman.2004.09.008

* Corresponding author. Tel.: +1 405 744 7644.

E-mail addresses: goutamc@okstate.edu (G. Chakraborty)8

prashant.srivastava@okstate.edu (P. Srivastava)8 warrdave@yahoo.com

(D.L. Warren).1 Tel.: +1 405 744 5192.

The Internet allows marketers to communicate with

interested customers, prospects, and stakeholders world-

wide. A firm’s corporate web site often acts as a gateway

for visitors to find critical information that shapes their

perceptions about the company as a whole as well as the

company’s offerings. If visitors to a corporate web site do

not find helpful and relevant information, they are likely

to form a poor perception about the company and may

ultimately visit and transact with competitors who

provide the necessary information. In a B2B context,

the information aspects in a corporate web site become

even more important due to the financial risk and the

complexity of the exchange process with typical B2B

transactions (Peppers & Rogers, 2001). Therefore, this

research focuses on the information content of corporate

web sites and the aspects of information customers

perceive important for corporate web site effectiveness.

While many consulting companies may have con-

ducted extensive empirical research on corporate web site

effectiveness, the results from such studies are, unfortu-

nately, not available in the public domain. There are only

ment 34 (2005) 420–429

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429 421

a handful of studies in the academic literature that looked

specifically at corporate web site issues (Palmer, 2002;

Robbins & Stylianou, 2003). Palmer (2002) used student

samples to evaluate corporate web sites from Fortune

1000 companies. His results indicated that web site

success is significantly associated with web site download

delay, navigation, content, interactivity, and responsive-

ness. Robbins and Stylianou (2003) content analyzed

corporate web sites from companies across the globe and

found web site content features to be significantly

different across cultural groups. However, in both of the

studies mentioned above, the authors did not distinguish

between B2B and B2C corporate web sites. In addition,

neither study actually measured and tracked business

customers’ perceptions about corporate web sites. Given

the importance of B2B and the web (Berthon, Ewing,

Pitt, & Naude, 2003), it is critical to specifically

investigate from business customers’ perspectives what

factors drive the effectiveness of corporate B2B web sites

and whether such factors differ between customers from

different parts of the world.

The impact of the Internet in international marketing is

expected to be much greater for B2B than for B2C

(Samiee, 1998). A recent report from eMarketer (http://

www.emarketer.com/) projects worldwide total e-com-

merce to grow to $3.2 trillion in 2004 and be dominated

by an online population outside of North America. As

per recent statistics, by 2002, 600 million people are

online worldwide (http://www.clickz.com/stats/big_picture/

geographics/print.php/151151), and the number of Euro-

pean Internet users (190 million) exceeds the number of

North American Internet users (183 million). It is

common knowledge that there are many underlying

differences between how customers in North America

perceive marketing stimuli or communication compared

to customers in other parts of the world (Lynch & Beck,

2001). However, what is unknown is how such differ-

ences in customers’ perceptions manifest in influencing

corporate B2B web site effectiveness. This research

attempts to bridge this gap with a field study of 368

business customers from two different continents—North

America and Europe.

The objective of this research is to identify the drivers of

a B2B corporate web site’s effectiveness and explore

whether these drivers are different for business customers

from different parts of the world. In the next section, a brief

review of relevant prior research is provided, followed by

hypotheses about the importance of different dimensions of

information in a corporate B2B web site and their relation-

ship to web site effectiveness. This is followed by a

discussion of the design of the survey instrument and

method of data collection. The results from a survey using a

large sample of B2B customers in the construction industry

are discussed in the next section. Finally, the implications of

the research, its limitations, and future research directions

are discussed.

2. Literature review and conceptual definitions

A B2B corporate web site is a gateway between a

company and its prospects, customers, and other stake-

holders. This differs from B2C web sites in that the

prospects and customers are other businesses rather than

end consumers. The goal of this research is to understand

the effects of prospects’ and customers’ perceptions about

different dimensions of information on a B2B firm’s

corporate web site’s effectiveness. In considering the

perceptions of prospects and customers of a B2B web

site, this paper emphasizes broader perceptual constructs

(such as information quality, usability, etc.) rather than

very specific web site design features (such as exact

color combinations, font size, etc.). Consequently, the

literature review ignores academic and popular press

literature that has focused on very specific web site

design features. Instead, the conceptual model is built by

drawing upon literature that proposes relationships among

broader perceptual constructs as shown in Fig. 1. Each of

the three perceptual constructs in Fig. 1 has been

identified by prior researchers as antecedents that

influence web site effectiveness as discussed next.

2.1. Dimensions of information in a corporate B2B web site

A B2B firm’s corporate web site is typically used to

provide general information about the company, trade or

application-specific information and news, and product

information (Robbins & Stylianou, 2003). Most multina-

tional firms provide links from their corporate web sites

to their country-specific web sites where the company

operates. Often, the actual commercial transactions

(involving order-placement, payment, tracking of ship-

ment, etc.) are carried out in those country-specific web

sites because of the idiosyncratic nature of the business

environment (laws and regulations, taxes, etc.) in each

country and varying company policies (with regard to

pricing, packaging, delivery, etc.) in different countries.

Thus, the effectiveness of a B2B firm’s corporate web site

may be more dependent on the ability of the web site to

provide relevant information to generate enough interest

to convert a visitor into a customer or partner. In this

manuscript, three dimensions of information (informative-

ness, usability of information, and quality of information)

in a corporate B2B web site are considered. These

dimensions of information were selected based on prior

research, content analysis of corporate web sites, and

expert opinions as described next.

2.1.1. Informativeness

Marketing practitioners and academic researchers con-

tend that one of the primary purposes of a company’s

web site (corporate or country-specific) is to provide

information to prospects, customers, and other stake-

holders (Chen, Clifford, & Wells, 2002; Eighmey, 1997;

Fig. 1. Informational antecedents of web site effectiveness.

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429422

Lohse, Bellman, & Johnson, 2000; Palmer, 2002; Robbins

& Stylianou, 2003). A content analysis of corporate web

sites of Fortune 500 companies indicated that a large

amount of content (information) was provided on these

sites (Robbins & Stylianou, 2003). Another content

analysis revealed that firms’ country-specific web sites

are being used as part of an integrated communication

strategy to serve higher objectives such as creating desire

and action (Sheehan & Doherty, 2001). Consequently, the

ability of a web site to make a visitor feel that the web

site has communicated something of value is viewed as

one of the most important predictors of web site

effectiveness.

Informativeness is the ability of a web site to make

information available. In this sense, informativeness may be

viewed as static information or content available on a web

site (Hoffman & Novak, 1996). Usability of information, on

the other hand, reduces consumers’ search costs by helping

to access the information available on a web site more

efficiently (Bakos, 1997). Furthermore, it must be noted that

a site may score high on informativeness regardless of the

manner in which it is presented. Thus, concerns of

information overload or formatting are unrelated to the

ability of the site to provide information of value. Lastly,

informativeness is conceptualized as a perceptual construct.

Therefore, informativeness is not the same as the actual

amount of information available on a web site, even though

these are expected to be correlated.

In an exploratory study by Keeney (1999), a need to

maximize product information was expressed as one of the

main objectives related to e-commerce. Palmer (2002) found

content (as perceived by student juries) of corporate web

sites to be a significant predictor of the web sites’ success.

Chen and Wells (1999) found perceived informativeness of a

web site to be the second most important factor in explaining

variance in visitors’ attitudes toward the web site. Therefore,

the following hypothesis is proposed:

H1. The higher the level of perceived informativeness in a

corporate web site, the higher is the web site’s effectiveness.

2.1.2. Usability

Usability relates to users’ perceived ability to utilize the

information provided by a web site. These perceptions are

often strongly influenced by the web site interface design,

specifically its navigational and interactive features (Palmer,

2002). Good navigational design and use of interactive

features in a web site help users to find relevant information

easily and quickly, thereby making information more

usable. Thus, usability is viewed at an abstract level as

pertaining to the pragmatics of how a user perceives and

interacts with a web site (Barnes & Vidgen, 2002). Many

researchers have proposed and found empirical support for

the notion that more interactivity usually results in a more

satisfying web site experience, which in turn increases the

effectiveness of web sites (Novak, Hoffman, & Yung,

2000).

Berthon, Pitt, and Watson (1996) suggested that the

usability of a web site is critical in converting site visitors

from blookersQ to bbuyers.Q In other words, usability at a

web site influences the web site’s effectiveness. Ghose and

Dou (1998) studied interactive functions in web sites and

found that the greater the degree of interactivity (leading to

higher usability of information) in a web site, the higher is

the web site’s attractiveness.

An exploratory study revealed ease of use as one of the

concerns of web site users. Specifically, maximizing ease of

user interface, making access easy and simple, and finding

desired products were mentioned as major objectives

(Keeney, 1999). Clearly, an efficiently executed web site

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429 423

design that enhances ease of use is an important factor in

determining web site effectiveness (Eighmey, 1997). On the

other hand, factors such as web page complexity that lower

ease of use also lower attitude toward the site (Stevenson,

Bruner, & Kumar, 2000). Similarly, in a study based on B2B

sites, navigability was found to improve commitment

toward the site (Bauer, Grether, & Leach, 2002). Finally,

Palmer (2002) also found interactivity and navigation to

have significant association with a corporate web site’s

success. Therefore, the following hypothesis is proposed:

H2. The higher the level of perceived usability in a

corporate web site, the higher is the web site’s effectiveness.

2.1.3. Quality of information

This refers to the quality of the content in a web site and

the suitability of the content from a user’s perspective

(Barnes & Vidgen, 2002). Just as the perceived amount of

information (informativeness) is important, the quality of

information in a corporate web site is also important in

shaping visitors’ initial perceptions about the company.

Because of advances in information technology, providing

timely and accurate information is not supposed to be

difficult for organizations. Equally important is that

customers’ expectations today are extremely high, and they

are unwilling to accept outdated, inaccurate, unbelievable

information in a web site. Empirical evidence suggests that a

web site that provides outdated or inaccurate information

leads to visitor dissatisfaction and abandonment of the web

site (Barnes & Vidgen, 2002; Palmer, 2002; Wang, 1998). In

other words, poor quality of information is detrimental to a

web site user’s experience leading to reduction of web site

effectiveness. Similarly, relevant and accurate information at

a web site has been found to improve visitors’ experience at

a web site, leading to higher web site effectiveness (Robbins

& Stylianou, 2003). Therefore, the following hypothesis is

offered:

H3. The higher the level of perceived information quality

in a corporate web site, the higher is the web site’s

effectiveness.

While separate hypotheses for different structural rela-

tionships among the dimensions of information and web site

effectiveness for customers from different parts of the world

are not proposed, a key question to explore in this research

is whether these structural relationships (shown in Fig. 1)

vary by the nationality of a corporate web site visitor. This

will be tested empirically in this study.

2.1.4. Web site effectiveness

Both behavioral and perceptual metrics have been

proposed to measure the effectiveness of web sites. Many

of the commonly used behavioral measures (such as hit rate

or unique visitors) suffer from problems due to the wide-

spread availability and use of online robots, non-uniqueness

of IP addresses, and caching of web pages by Internet

browsers (Dreze & Zufryden, 1998). Thus, in this research,

a perception-based measure of effectiveness of a web site

was used. Chen and Wells (1999) developed a valid and

reliable rating scale to measure overall effectiveness of B2C

web sites. This measure was adopted with appropriate

modifications for corporate B2B web sites.

3. Research design

3.1. Study context and sample selection

Data for the main study were obtained from corporate

web site visitors of a power tools company in the

construction industry. The focal company (who wishes to

remain anonymous) had more than $2 billion in sales in

2002, employs more than 10,000 people worldwide, and

conducts business in all continents (most revenues come

from Europe and North America). The company operates a

corporate web site (the URL ends with .com) and more than

25 country-specific web sites (with URLs ending with

country specific top-domain such as .it for Italy). As is often

the case with large multinational firms, the company uses its

corporate web site to provide general and trade-specific

information to all visitors, and basic product information for

visitors from those countries where the company does not

operate a fully integrated web site. In addition, for legal as

well as business-related reasons, the company does not offer

the ability to carry out commercial transactions on its

corporate web site. Rather, visitors who want to conduct

transactions are routed to the company’s country-specific

sites.

The company’s corporate web site was used to solicit

visitors (using a link) to participate in a web-based survey

(hosted independently by a marketing research company).

The link was active for a period of about 3 weeks in Fall

2002. As an incentive, the respondents were offered a

chance to win a high-end drill/driver (retail price about

$200) manufactured by the power tools company. The web

site visitors were given a choice to participate in the survey

in one of the two languages—English and German. These

are the two most commonly used languages by the

customers and prospects for this company. A total of 420

responses were obtained. However, there were very few

respondents from Australia, Africa, South America, or Asia.

Those responses were not used in this study. Thus, the

effective sample size is 368 (229 from North America and

139 from Europe, of which 46 are in the German version).

The characteristics of the final sample are given below.

Of the respondents’ job functions, about 57% were

construction professionals, 13% were engineers, and the

remaining 30% performed other functions. It was also found

that the corporate web site was most often used by sample

respondents to get general and technical product informa-

tion, repair or spare parts information, contact and pricing

information. These sample characteristics compare well

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429424

with the known characteristics of the company’s corporate

web site visitors.

3.2. Instrument development

Whenever possible, previously published measures were

used with appropriate adaptation for the context in the

survey. In addition, all of the measures were extensively

pretested with a focus group of customers and several in-

depth interviews of managers from the power tools

company to check for representativeness and appropriate-

ness of the items. The exact items for each construct are

shown in Table 1 and are briefly discussed below.

3.2.1. Informativeness

The items for this construct were adapted from previous

research (Chakraborty, Lala, & Warren, 2002, 2003) and

Table 1

Construct measures and validity

Construct Definitions and items Europe

Std.

Load

CR

Usability Ability of a web site to engage in

two-way communication with a

visitor and make it easy to find

needed information.

0.902

U1 The web site has an effective search

capability.

0.76

U2 The web site is easy to navigate. 0.87

U3 My interaction with the site is clear

and understandable.

0.90

U4 I find the site is easy to learn to

operate.

0.92

U5 I find the site easy to use. 0.92

Informativeness Ability of a web site to provide a

visitor with information of value.

0.820

I1 The site provides detailed technical

information about products.

0.81

I2 The site provides application or trade

specific usage information.

0.91

I3 The site provides industry-related

news/information.

0.78

Quality of

Information

Quality of content of the web site and

suitability of the information for

user’s purpose.

0.878

Q1 The information on the site is timely. 0.79

Q2 The information on the site is

accurate.

0.80

Q3 The site provides believable

information.

0.84

Q4 The site provides relevant

information.

0.85

Web site

effectiveness

Overall evaluation of the goodness or

badness of a web site.

0.849

WSE1 I am satisfied with my experience

with the corporate website.

0.88

WSE2 I feel it is useful to spend my time at

this site.

0.89

WSE3 The site makes it easy to build a

relationship with the company.

0.83

In this table, bStd. LoadQ is bstandardized loading,Q bCRQ is bcomposite reliability

had good psychometric properties. Three items were used to

capture different types of information typically offered by a

firm in a corporate B2B web site. These items are identified

as I1–I3 in Table 1. A representative item is bthe corporate

web site provides trade application or trade specific usage

information.Q Responses were measured using a seven-point

Likert-type scale with end anchors bstrongly agreeQ and

bstrongly disagree.Q

3.2.2. Usability

A five-item measure was used to tap into the domains of

this construct. These items were adapted from Barnes and

Vidgen (2002) and are also well established in the B2C

literature. Appropriate changes were made to the items as

required to account for the difference between B2B and

B2C web sites. A representative item is bthe corporate web

site is easy to navigate.Q Responses were measured using a

North America Combined Sample

AVE Std.

Load

CR AVE Std.

Load

CR AVE

0.651 0.946 0.779 0.932 0.733

0.79 0.72

0.85 0.83

0.88 0.88

0.90 0.91

0.89 0.92

0.604 0.877 0.705 0.855 0.664

0.81 0.81

0.83 0.88

0.72 0.75

0.644 0.922 0.748 0.901 0.695

0.82 0.80

0.85 0.82

0.86 0.86

0.84 0.85

0.654 0.929 0.814 0.899 0.749

0.92 0.90

0.88 0.87

0.86 0.82

,Q and bAVEQ is baverage variance extractedQ based on CFA results.

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429 425

seven-point Likert-type scale with end anchors bstronglyagreeQ and bstrongly disagree.Q These items are identified as

U1–U5 in Table 1.

3.2.3. Quality of information

A four-item scale was used to measure quality of

information. These items were also adapted from Barnes

and Vidgen (2002), who developed the measures through

extensive research and using the work on service quality as

a base. These items are identified as Q1–Q4 in Table 1. A

representative item is bthe corporate web site provides

believable information.Q Responses were measured using a

seven-point Likert-type scale with end anchors bstronglyagreeQ and bstrongly disagree.Q

3.2.4. Web site effectiveness

A three-item measure was used to tap into the domains of

this construct. These items were adapted from Chen and

Wells (1999) and later used by Chakraborty et al. (2002) by

making some modifications to the items. A representative

item is bI am satisfied with my experience with this

corporate web site.Q These items are identified in Table 1

as WSE1–WSE3. Responses were measured using a seven-

point Likert-type scale with end anchors bstrongly disagreeQand bstrongly agree.Q

3.3. Survey instrument

Respondents were asked to indicate their perceptions of

the sponsoring company’s corporate web site on a battery

of questions as described in the measures. At the end of

the survey, they were asked a few questions about their

country of residence, reasons for visiting the web site, and

their job descriptions. The survey was first produced in the

English version, then translated into German. The German

version was back-translated into English and compared to

the original English version. Discrepancies between the

back-translated version and the original version were

resolved through a number of iterations. Appropriate

modifications to questions were made during these

iterations to ensure that the integrity of the meaning and

content remained the same between the two language

versions of the survey.

Table 2

Tests of factor invariance of measurement model between North American

and European groups

Multiple groups model

Free (unconstrained) Constrained Combined

Chi-square 502.22 512.46 306.61

df 168 179 84

NFI 0.96 0.96 0.98

CFI 0.98 0.98 0.98

RMSEA 0.11 0.10 0.088

Difference in chi-square between bfreeQ and bconstrainedQ model is 10.23

with 11 df ( p-valueN0.05).

4. Results

4.1. Analysis overview

Some key issues with respect to handling of data needed

to be resolved before proceeding with the analysis. The first

question that needed to be addressed was whether or not the

German version (all of these were from European respond-

ents) of the survey data could be pooled together with the

English version of the survey data (from European

customers only) for creating a larger European group. A

comparison of all key measures (the three independent and

the dependent) between these two groups resulted in no

statistically significant differences. The two groups were

also not statistically different in important characteristics

such as high-speed Internet access and the number of visits

they made to the company’s corporate web site in last three

months. Therefore, it was decided to create a European

group by pooling the data from both English and German

surveys.

The second critical issue was to determine whether the

factor structure of the measures were similar between the

European and the North American group. A test of factor

invariance between the two sub-samples (North America

and Europe) was carried out as recommended by

Durvasula, Andrews, Lysonski, and Netemeyer (1993).

To achieve this, all measures (independent and dependent)

were analyzed in three confirmatory factor analysis (CFA)

models. For each of the three CFA models, the items-to-

construct mapping was identical and was based on the

conceptualization described earlier. First, a multiple groups

CFA model with all parameters free (unconstrained) was

estimated. Second, the same multiple groups CFA with the

factor loadings constrained to be equal across the sub-

samples was fitted. Finally, a single CFA model that

combined the two sub-samples was estimated. The

relevant details from these CFA models are reported in

Tables 1 and 2, and the factor invariance test is described

below.

4.2. Test of factor invariance and results of the measurement

model on the North American and European sample data

To test for factor invariance, first the unconstrained

model was compared with the constrained model using

differences in chi-square test results as well as model fit

indices (Table 2). The comparison clearly indicated a

preference for the constrained model as evidenced by

comparable fit indices and non-significant difference in

chi-squares. Then the unconstrained model was compared

with the combined model using the fit indices. Again, this

comparison of model fit indices across the models clearly

showed that the combined model produced a better fit

(reported in Table 2). Thus, factor invariance across the sub-

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429426

samples was achieved. In other words, the underlying factor

structure of the constructs being measured is identical across

the North American and the European customers.

The overall fit measures for the confirmatory factor

analysis models demonstrate a fairly good fit, as shown in

Table 2. Additionally, all factor loadings were statistically

significant at the 5% level, suggesting a good model.

Measures for construct validity were also tested. Measures

of the level of internal consistency between items of a single

construct, the differences between items of different con-

structs, and convergent validity were assessed for each of

the four constructs. Item reliabilities, tests of composite

reliability, and average variance extracted were examined.

The composite reliabilities were high and exceeded the

commonly accepted benchmark of 0.70 (Fornell & Larcker,

1981). Average variance extracted measures the amount of

variance captured by a construct in relation to the variance

due to random measurement error. All estimates of average

variance in the data extracted exceeded the 0.50 minimum

cutoff suggested by Bagozzi and Yi (1988). These values

are reported in Table 1.

The first test of discriminant validity was to assess

whether pairs of constructs were sufficiently different from

each other or whether they could be collapsed into single

factors to yield a more parsimonious model. In order to

achieve this, a two-factor confirmatory factor analysis of

pairs of constructs was conducted twice, once by constrain-

ing the correlation between the latent variables to unity and

once by freeing up the parameter. A chi-square difference

test was then used to test whether the chi-square value of the

unconstrained model was significantly lower, in which case

discriminant validity would be upheld (Anderson & Gerb-

ing, 1988). The results from this test indicated that the

discriminant validity was upheld in all pairwise tests. A final

test was whether the 95% confidence intervals of the phi’s

(i.e., the correlations between the constructs) contain the

value of unity (Anderson & Gerbing, 1988). This test was

Table 3

Relationship between antecedents and web site effectiveness

Antecedents Dependent variable: web site effectiveness

North America Europe

Std. Estimateb t-statistic Std. Estimateb

Informativeness 0.35* (0.08) 4.15 0.45* (0.18)

Usability 0.50* (0.09) 5.26 0.16 (0.12)

Quality of Information 0.06 (0.10) 0.59 0.40* (0.23)

SMC 0.71 0.91

Chi-square 305.52 183.87

df 84 84

RMSEA 0.117 0.092

CFI 0.98 0.98

a See Cohen and Cohen (1983).b The numbers in parentheses are standard errors from models run separately oc Standard error of estimate from pooled structural model.*Indicates that the coefficient is statistically significant ( p-valuesb0.05). One-

effect in hypothesis was predicted.

also found to be satisfactory for all the constructs. There-

fore, the measurement model seems to be acceptable.

4.3. Structural equation models

Having validated the measurement model, hypotheses

were tested using a structural equation model with web site

effectiveness as the endogenous variable and the three

antecedents in Fig. 1 as the exogenous variables. The model

was estimated separately on the two groups—North

America and Europe. The overall results are reported in

Table 3. As the numbers in the Table 3 show, the structural

model is performing satisfactorily for both groups (North

America—m2=305.52 with 84 df; CFI=0.98, NFI=0.97,

IFI=0.98; RMSEA=0.117; Europe—m2=183.87 with 84 df;

CFI=0.98, NFI=0.96, IFI=0.98; RMSEA=0.092). The inde-

pendent variables together explained 71% of the variance in

web site effectiveness for the North American sample and

91% for the European sample. Based on the t-statistics of

standardized estimates in the model, informativeness and

usability had statistically significant ( p-valueb0.05) and

positive effects on web site effectiveness for the North

American group. Hence, predictions in H1 and H2 were

supported for the North American group. However, the

quality of information was non-significant, though its effect

was positive as hypothesized. H3 was not supported for the

North American group. For the European group, both

informativeness and quality of information had statistically

significant ( p-value b0.05) and positive effects on web site

effectiveness. Hence, predictions from H1 and H3 were

supported for European group. Although usability had non-

significant effects in the European sample, its direction was

positive as hypothesized. Thus, prediction from H2 was not

supported for the European group. Taken together, H1 is

supported both in the North American and European

samples, H2 is supported in the North American sample

only, and H3 is supported in the European sample only.

Difference t-testsa

t-statistic Difference in Std. Estimates

between groupct-statistic (based on

pooled error)

2.54 0.10 (0.08) 1.25

1.34 0.34* (0.07) 4.86

1.72 0.34* (0.10) 3.40

n each group.

sided test was used only for testing those relationships where a directional

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429 427

While the structural models for both groups seem to fit

well and establish the relationships between the inde-

pendent and the dependent measures, one of the research

questions was to examine whether the relationships

among the constructs vary by nationality of corporate

web site visitors. In other words, the goal was to assess

whether the structural relationships (coefficient estimates

of informativeness, usability, and quality of information)

in the North American and European groups were

different. Table 3 shows the standardized parameter

estimates and the statistical tests for differences in the

parameter estimates between the two groups. The pattern

of the numbers in Table 3 identifies interesting differences

in how the structural relationship differs for the North

American and European groups. The coefficient for

informativeness does not differ between the two groups.

The standardized coefficient estimate for informativeness

was 0.35 for the North American group, and 0.45 for the

European group. The difference in the standardized

estimate for informativeness between the two groups is

not statistically significant ( p-valueN0.05). However, the

difference in the standardized coefficients for usability

(0.50 for the North American group and 0.16 for the

European group) and quality of information (0.06 for the

North American group and 0.40 for the European group)

are statistically significantly different ( p-valueb0.05)

between the groups (Cohen & Cohen, 1983).

5. Discussion and managerial implications

Based on the review of academic and trade press

literature, three dimensions of information (informative-

ness, usability, and quality of information) were identified

that might influence a corporate B2B web site’s effective-

ness. Using prior research and pretests, reliable and valid

perceptual measures for these constructs were developed.

The significance of these three factors in explaining a B2B

web site’s effectiveness for North American and European

respondents was tested simultaneously. Several interesting

results emerged from this study. First, the measurement

model factor structure was found to be invariant between

the North American and European groups, suggesting that

the underlying factor structure for the constructs inves-

tigated in this research is identical between the two

groups.

Second, the structural models built separately for each

group produced good fit indices and interesting insights. For

instance, informativeness had a significant effect on

corporate web site effectiveness for both the groups.

Interestingly, it was the most important variable for explain-

ing web site effectiveness for the European group but the

second-most important variable for the North American

group. Quality of information had a significant effect for the

European group, but it was non-significant for the North

American group. The usability had a significant effect (and

it was the most important variable in explaining web site

effectiveness) for the North American group but not for the

European group.

Third, it was found that the nationality of a web site

visitor moderates the structural relationships between the

exogenous and the endogenous constructs in the model. The

effects of this moderation manifest in the structural relation-

ship between usability and web site effectiveness, as well as

between quality of information and web site effectiveness.

This is an important finding and suggests that managers of

B2B corporate web sites need to be very careful in

designing their sites and not try to use a bone shoe fits

allQ approach when dealing with business customers from

North America and Europe.

This study represents the first effort in academic

literature to determine the differences in factors that

contribute to the effectiveness of corporate B2B web sites

between different geographical regions in the world using

business customers’ perceptions. There is an ongoing

debate in marketing and advertising literature about

whether communication and marketing practices can be

standardized due to the emergence of a global economy

and perhaps even global customers. The advent of the

Internet has been argued to tilt this debate towards

adopting standardized global communication practices.

As pointed out earlier, there is a lack of research looking

at this issue for corporate web sites that are used by

companies to communicate with customers from different

parts of the world. The results from this study suggest that

the factors that are important in customers’ perceptions of

a corporate web site’s effectiveness differ across geo-

graphic regions of the world, and companies need to

consider this issue in their design of corporate web sites.

Given that it is technologically possible for a company to

figure out (based on the IP address of a visitor or using

cookies) which part of the world the visitor is coming

from, perhaps a company can use different versions of its

corporate web site (emphasizing different factors) that can

be created on the fly using dynamic portal technologies

(such as those provided by Broadvision, BEA systems,

etc.) to uniquely appeal to visitors from different parts of

the world.

It is also interesting that for North American customers,

usability turns out to be the most important factor in

explaining corporate web site effectiveness. This is con-

sistent with the commonly held belief that the business

customers from North America are often more pressed for

time (than European customers) and therefore finding

information quickly and easily in a corporate web site

appeals to them. On the other hand, the European business

customers may be more sensitive (than North American

business customers) to the quality of information in a

corporate web site because of the nature of the business

environment in Europe. These are post hoc conjectures, and

more research is clearly needed to explore these issues in the

future.

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429428

6. Limitations and future research

The objective of this study was to test the effects of

different dimensions of information on corporate B2B

web site effectiveness and how such effects differ for

respondents from different parts of the world using

business customers’ perspectives. Owing to this aim, this

study has a few limitations. First, this study is based on

business customers’ perceptions of a company’s corporate

web site in the construction industry. While this allowed

us to control for inter-industry differences, the findings

from this study need to be interpreted with caution as

they may not generalize to other industries. Replicating

these findings in future research using corporate web sites

from other industries would be extremely important to

increase the generalizability of these results. Also, as the

sampling frame was restricted to the sponsoring compa-

ny’s customer base, data gathering was restricted to only

two language versions (English and German) of the

survey (these two are the most commonly used languages

by customers of this company.) Data analyses were

restricted to only North American and European samples,

as the company has very small customer bases in other

regions of the world (about 85% of the company’s

revenues are generated from North America and Western

Europe). Thus, the results may not be generalizable to

other parts of the world.

Second, even though indirect effects were not addressed

in this model, it is possible that a hierarchical relationship

exists among the antecedent variables investigated in this

research. These types of hierarchical models need to be

investigated in future research. Related to this issue is the

possibility of the existence of interaction effects among the

independent variables (informativeness, usability, and qual-

ity of information) used in this study. This needs to be

investigated more in future research.

Third, only perceptual measures were used for both

independent and dependent variables. While perceptual

measures are clearly important for generating insights into

how customers perceive and value corporate B2B web sites,

the bottom line orientation of marketers demands that future

research use other behavioral constructs for the dependent

measure, such as the number of leads generated for country-

specific web sites from traffic originating at corporate web

sites.

Acknowledgements

This research was conducted with support from a large

power tools company in the midwest region who wishes to

remain anonymous. Authors acknowledge helpful com-

ments from Vishal Lala and editorial assistance from Bev

Dunham.

References

Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in

practice: A review and recommended two-step approach. Psychological

Bulletin, 103(3), 411–423.

Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation

modeling. Journal of the Academy of Marketing Science, 16(1), 74–94.

Bakos, J. Y. (1997). Reducing buyer search costs: Implications for

electronic marketplaces. Management Science, 43(12), 1676–1692.

Barnes, Stuart J., & Vidgen, Richard T. (2002). An integrative approach to

the assessment of E-commerce quality. Journal of Electronic Commerce

Research, 3(3), 114–127.

Bauer, H. H., Grether, M., & Leach, M. (2002). Building customer relations

over the internet. Industrial Marketing Management, 31(2), 155–163.

Berthon, P., Ewing, M., Pitt, L. F., & Naude, P. (2003). Understanding B2B

and the web: The acceleration of coordination and motivation.

Industrial Marketing Management, 32(7), 553–561.

Berthon, P., Pitt, L. F., & Watson, R. T. (1996). The world wide web as an

advertising medium: Toward an understanding of conversion efficiency.

Journal of Advertising Research, 36(1), 43–54.

Chakraborty, G., Lala, V., & Warren, D. (2002). An empirical investigation

of antecedents of B2B web sites’ effectiveness. Journal of Interactive

Marketing, 16(4), 51–72.

Chakraborty, G., Lala, V., & Warren, D. (2003). What do customers

consider important in B2B web sites? Journal of Advertising Research,

43(1), 50–61.

Chen, Q., Clifford, S. J., & Wells, W. D. (2002). Attitude toward the site II:

New information. Journal of Advertising Research, 42(2), 33–45.

Chen, Q., & Wells, W. D. (1999). Attitude toward the site. Journal of

Advertising Research, 39(5), 27–37.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation

analysis for the behavioral sciences, (2nd ed.). Hillsdale, NJ7 Lawrence

Erlbaum Associates.

Dreze, X., & Zufryden, F. (1998). Is internet ready for prime time

advertising? Journal of Advertising Research, 38(3), 7–18.

Durvasula, S., Andrews, J. C., Lysonski, S., & Netemeyer, R. G. (1993,

March). Assessing the cross-national applicability of consumer behav-

ior models: A model of attitude toward advertising in general. Journal

of Consumer Research, 19, 626–636.

Eighmey, J. (1997). Profiling user responses to commercial web sites.

Journal of Advertising Research, 37(3), 59–66.

Fornell, C., & Larcker, D. F. (1981, February). Evaluating structural

equation models with unobservable variables and measurement error.

Journal of Marketing Research, 18, 39–50.

Ghose, S., & Dou, W. (1998). Interactive functions and their impacts on the

appeal of internet presence sites. Journal of Advertising Research,

38(2), 29–43.

Hoffman, D. L., & Novak, T. P. (1996). Marketing in hypermedia

computer-mediated environments: Conceptual foundations. Journal of

Marketing, 60(3), 50–68.

Keeney, R. L. (1999). The value of internet commerce to the customer.

Management Science, 45(4), 533–542.

Lohse, G. L., Bellman, S., & Johnson, E. J. (2000). Consumer buying

behavior on the internet: Findings from panel data. Journal of

Interactive Marketing, 14(1), 15–29.

Lynch, P. D., & Beck, J. C. (2001). Profiles of internet buyers in 20

countries: Evidence for region-specific strategies. Journal of Interna-

tional Business Studies, 32(4), 725–748.

Novak, T. P., Hoffman, D. L., & Yung, Y. F. (2000). Measuring the

customer experience in online environments: A structural modeling

approach. Marketing Science, 19(1), 22–42.

Palmer, J. W. (2002). Website usability, design, and performance metrics.

Information Systems Research, 13(2), 151–167.

Peppers, D., & Rogers, M. (2001). One to one B2B. New York, NY7

Currency Books.

G. Chakraborty et al. / Industrial Marketing Management 34 (2005) 420–429 429

Robbins, S. S., & Stylianou, A. C. (2003). Global corporate websites: An

empirical investigation of content and design. Information and

Management, 40, 205–212.

Samiee, S. (1998). The internet and international marketing: Is there a fit?

Journal of Interactive Marketing, 12(4), 5–21.

Sheehan, K. B., & Doherty, C. (2001). Re-weaving the web: Integrating

print and online communications. Journal of Interactive Marketing,

15(2), 47–59.

Stevenson, J. S., Bruner II, G. C., & Kumar, A. (2000). Webpage

background and viewer attitudes. Journal of Advertising Research,

40(1), 29–34.

Wang, Richard Y. (1998). Perspective on total data quality management.

Communications of the ACM, 41(2), 58–65.