A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION

OF THE EVOLUTION OF E-SALES

Luca Canetta, Naoufel Cheikhrouhou and Rémy Glardon

Ecole Polytechnique Fédérale de Lausanne (EPFL)

Laboratory for Production Management and Processes, Lausanne, Switzerland

Keywords: e-Commerce, customer behaviour, decision support system.

Abstract: Many parameters have a significant influence on e-commerce evolution. This complicates the assessment of

the requirements for and the consequences of e-sales adoption. In order to support the decisions of

companies thinking about a possible e-sales channels introduction a Decision Support System (DSS) is

proposed. The relevant e-commerce success factors, which constitute the DSS input, have been identified

and their influence described relying upon a literature review. The DSS output aims at describing typical e-

commerce evolution patterns taking into account the speed of adoption and the steady state potential

diffusion (saturation level). These variables point out the considerable discrepancies between the e-

commerce evolution charactering different industrial sectors. The DSS, which is based on a system of rules,

allows to qualitatively predict the expected e-sales evolution for companies introducing a specific e-sales

channels strategy in a given environment and to explain it in terms of different e-commerce success factor

configurations.

1 INTRODUCTION

Due to the continuous and consistent growth of B2B

and B2C e-commerce transaction share, an ever

increasing number of companies is confronted to the

decision of introducing e-sales, thus adopting

electronic-enabled sales channels (e-sales channels)

to replace and/or to support their traditional sales

channels. The adoption of a multi channel sales

strategy seems promising both in terms of sales and

revenue increase. However, e-commerce is not yet

mature and thus is far from having reached its steady

state potential, as demonstrated by the high

variability of the predictions of its future evolution

(Gurunlian, 2001). Furthermore, the choice of an

adequate e-sales strategy is complicated by the high

degree of uncertainty concerning customer reactions

to e-commerce, shown for instance by the uneven

utilization of online purchasing across different

industrial sectors (Selhofer, 2004; Stansfield, 2003).

Moreover, the choice of a wrong e-strategy can have

a critical impact on the company wealth. For these

reasons e-commerce introduction is considered as

highly risky. This hinders many companies to fully

profit from the benefits of the e-revolution and

demonstrates the need for tools capable of reducing

the complexity of management decisions related to

the choice of an e-strategy.

The analysis is mainly focused on B2B e-

commerce as it accounts already for about 80% of

the total e-commerce transacted monetary value and

it is also expected to continue to grow faster than

B2C in the next future (Scupola, 2002).

Several explanations have been proposed in

order to describe the important differences

characterising e-commerce adoption, both at the

single company and at the industrial sector level.

Proposals have also been made for the identification

of the characteristics that ensure a rapid and massive

e-purchasing adoption. Many of these analyses focus

only on a subset of the factors that potentially

influence e-purchasing adoption: e-commerce

suitability of the transacted products (Hunter, 2004;

Liu, 2004; Levin, 2003; Liu, 2003; Vijayasarathy,

2002; Lowengart, 2001; Phau, 2000; Peterson,

1997), customer readiness to adopt e-commerce

(Choi, 2004; Selhofer, 2004; Zhou, 2004; Fillis,

2003; Scupola, 2002; Liang, 1998) and brand image

as a means to partially leverage customer risk

perception (Bendixen, 2004; Lim, 2003; Clemons,

2002; Mudambi, 2002).

Canetta L., Cheikhrouhou N. and Glardon R. (2007).

A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION OF THE EVOLUTION OF E-SALES.

In Proceedings of the Third International Conference on Web Information Systems and Technologies - Society, e-Business and e-Government /

e-Learning, pages 97-106

DOI: 10.5220/0001268100970106

 SciTePress

Determining customers’ e-purchasing behaviour

is essential for companies thinking about a new e-

sales channel introduction. In fact, it allows to

estimate the e-sales amount that can be potentially

reached by targeting these customers. The objective

of this work is the design and development of a

Decision Support System (DSS) linking the value

taken by the e-commerce success factors to the e-

commerce demand evolution. Compared to previous

works, the presented paper is original in that it:

 goes beyond the current approaches and

evaluates the success of a given e-strategy,

taking into account all the related aspects;

 replaces the traditional division among e-

commerce adopters and non adopters with a

more detailed prediction of the e-commerce

demand evolution;

 provides a DSS to the enterprise management

board for a better integration of e-commerce.

In section 2, starting from the analysis of the

results of previous works, a comprehensive list of

factors influencing e-purchasing adoption is

identified and a detailed explanation of their

influence magnitude and direction is provided.

In section 3, the relationships between the

identified e-commerce success factors and the e-

purchasing adoption are investigated in order to

develop a rule-based DSS for e-sales evolution

prediction. It formalises the relationship between the

description of a specific case and the e-purchasing

adoption behaviour, which is described relying upon

the Diffusion Of Innovation (DOI) theory. It

provides qualitative indications about the saturation

level (share of e-purchasing at the end of the

innovation diffusion process) and the speed of

adoption (number of periods necessary to reach the

saturation level). The e-sales evolution is described

taking into account the combined influences of

product features, customer characteristics, brand

perception and e-commerce induced purchasing

process modifications. The DSS is developed using

a system of rules implying linguistic (qualitative)

variables in order to facilitate its parameterisation.

As shown in section 4 this facilitates the DSS

utilisation, in particular for users belonging to

companies not already familiar with e-commerce

that do not have at their disposal enough reliable

quantitative data.

2 FACTORS INFLUENCING

E-COMMERCE ADOPTION

Customer perception is fundamental for explaining

e-commerce adoption process (Choi, 2004; Zhou,

2004; Levin, 2003; Lim, 2003; Liu, 2003;

Vijayasarathy, 2002; Liang, 1998). A literature

review, considering both B2B and B2C works in

order to ensure a wider article selection, has been

undertaken in order to identify all the e-commerce

success factors. This assumes that the B2C and B2B

adoption processes are sufficiently similar to allow

for an extrapolation of the results from one field to

the other.

2.1 Product Features

The moderating effect of product/service

characteristics on the willingness to adopt e-

commerce has often been studied in the literature

(Liu, 2004; Levin, 2003; Liu, 2003; Vijayasarathy,

2002; Lowengart, 2001; Phau, 2000; Peterson,

1997). The following product characteristics are

among the most cited in the analysed literature:

degree of intangibility, importance, degree of

membership to the search good class, degree of

standardization.

2.1.1 Degree of Intangibility

Intangibility, the major characteristic that

distinguishes goods from services, affects the

customer’s ability to judge the quality of the

good/service. Intangibility is mainly related to the

lack of physical evidence; thus by the extent to

which a good cannot be touched or seen, it is

inaccessible to the senses and lacks a physical

presence. The degree of intangibility has a positive

effect on e-purchasing adoption due to transaction

costs reduction made possible by instantaneous

electronic delivery of digital product (Liu, 2003).

Moreover, instantaneous electronic delivery

decreases the uncertainty about the goods quality

(Liu, 2003). Due to the positive effect on both

transaction costs and the quality uncertainty, the

degree of intangibility is considered being a very

important e-commerce success factor. Evidence of

the significant impact of the degree of intangibility

on e-commerce adoption can be found both in the

B2B (Liu, 2004) and B2C literature (Liu, 2003;

Vijayasarathy, 2002; Phau, 2000; Peterson, 1997).

WEBIST 2007 - International Conference on Web Information Systems and Technologies

2.1.2 Importance

Product importance can be defined as the “degree of

essentiality” of the supplier delivery to the

organization, as perceived by the buyer (Hunter,

2004). Product importance has a negative effect on

e-purchasing adoption because it increases perceived

risk, which in turn is recognized as an important

hindering factor of online shopping (Liu, 2004; Lim,

2003). Two elements characterize the perceived risk:

the likelihood of negative outcomes and the

magnitude of the undesirable consequences, mainly

due to financial and performance risk (Lim, 2003).

The magnitudes of both are directly influenced by

product importance because the costs and losses

resulting from quality problems, delay problems and

other uncertainties directly depend on the product

importance. In a B2B marketplace both the financial

and the performance risks are significant. This

requires a comprehensive evaluation of product

importance. In B2C, it appears that the product price

plays a major role (Lim, 2003; Vijayasarathy, 2002;

Lowengart, 2001; Phau, 2000; Peterson, 1997). The

B2B literature points out the dichotomies between

low importance products, such as commodities and

Maintenance, Repair and Operations (MRO), and

important products, such as direct customised

materials. These findings are confirmed by the wider

diffusion of e-purchasing for MRO products than for

direct materials (Hunter, 2004).

2.1.3 Degree of Membership to the Search

Good Class

A search good is a product or service with features

and characteristics that are easily observable before

the purchase. Thus it can be completely described

and assessed without the necessity of physical

inspection. On the contrary, an experience good is a

product or service for which characteristics such as

quality or performances/functionalities are difficult

to observe in advance and can only be ascertained

upon consumption or, at least, physical inspection.

Search goods can be more easily transacted online

than experience ones because all data about their

relevant characteristics can be gathered consulting a

website, before transaction fulfilment. In this way,

the customer perceived risk is drastically reduced in

case of search goods e-purchasing (Clemons, 2002).

The degree of membership to the search good class

is thus a factor having a pronounced positive impact

on e-purchasing adoption. This hypothesis coincides

with the results of previous works (Choi, 2004;

Levin, 2003; Clemons, 2002; Phau, 2000; Peterson,

1997).

2.1.4 Degree of Standardization

A high degree of standardization reduces the need

for product physical inspection (Clemons, 2002;

Liang, 1998) and the amount of information

necessary for taking a purchasing decision (Hunter,

2004). This results in a reduction of the customer

perceived risk that facilitates e-purchasing adoption

for highly standardised products. The degree of

standardization is considered having a positive effect

on e-purchasing adoption, even if the magnitude of

its effect is less pronounced than those of the other

product features.

2.2 Customer Characteristics

The relationships between customer characteristics

and e-commerce adoption have been pointed out in

various works, both in the case of B2B (Liu, 2004;

Selhofer, 2004; Stansfield, 2003; Scupola, 2002) and

B2C (Vijayasarathy, 2002; Phau, 2000). The

following four characteristics have been identified in

order to explain the behaviour of customers involved

in B2B e-commerce: ICT resources, e-commerce

experience, e-sales channel evaluation, purchasing

frequency.

2.2.1 ICT Resources

The lack of ICT resources can constitute an entry

barrier for companies interested in e-commerce,

especially for SMEs (Fillis, 2003; Stansfield, 2003).

At the infrastructure level, many basic ICT resources

(computer use, internet access and e-mail use) have

already been widely adopted, however the use of the

world wide web still lags behind for companies not

using e-procurement (Selhofer, 2004). The

differences between e-commerce adopters and non

adopters are even more significant for IT staff

recruitment, ICT training and knowledge

acquirement and ICT use at the organisation level

(Fillis, 2003; Scupola, 2002). In fact, the strongest

barrier to e-commerce take-up by SMEs appears to

be a lack of knowledge about the Internet and

electronic commerce, which provokes a lack of IT

skilled staff and hinders the great majority of small

firms, still in the early stages of Internet adoption, to

make the leap towards full integration (Stansfield,

2003). The diffusion of various ICT resources and

IT activities are estimated for four classes of

companies. These classes are obtained by splitting a

sample of 4326 European companies involved in

A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION OF THE EVOLUTION OF E-SALES

B2B transactions (Selhofer, 2004), according to their

current share of e-purchasing. A ratio of adoption

(utilisation) has been calculated for each ICT

resource (IT activity) and for each class, it is shown

in table 1 as the percentage of adopting companies.

The basic ICT resources (Computer use, Internet

Access availability, E-mail use, WWW use) include

various mature ICT resources that can be used

independently by each user. On the other hand, the

sophisticated ICT resources (Intranet, Extranet,

LAN, Wireless LAN) mainly concern the

development, management and utilisation of various

network typologies. The columns IT training and

learning and IT staff recruitment show the

percentage of companies engaged in these activities,

thus providing a measure of the importance and the

investments made for achieving the required IT

skills. Finally, the percentage of companies having

created their own website gives information about

the importance attached to the Internet as a means of

communication, collaboration and transaction.

The obtained results confirm the tendencies

previously outlined in the literature:

 Non adopters (almost 55% of the sample)

show lower utilisations than e-purchasing

adopters even for various ICT basic resources

(e-mail and World Wide Web);

 Basic ICT resources are widely used by all the

companies using e-purchasing (almost 100%

of adoption) whatever their share of e-

purchasing;

 Weak adopters (22% of the sample) lag

behind in terms of ICT resources and

utilisation, in particular while focusing on

sophisticated ICT resources (Intranet,

Extranet, LAN, Wireless LAN) and more

challenging activities (IT training, IT staff

recruiting, company website development).

Available ICT resources have a positive

influence on e-purchasing adoption without being

one of its main drivers. Their discriminative power

is decreased by the fact that the differences in ICT

resources among companies are less and less

prominent in many industrial sectors.

2.2.2 e-Commerce (e-Purchasing)

Experience

This attribute captures all the past e-commerce

experiences of the customer. The pronounced

positive impact of previous successful e-commerce

experiences is explained by the “learning effect in

electronic commerce” (Liang, 1998); the confidence

and skills developed across successful e-commerce

utilisations decrease some components of customer

perceived risk and thus increase customer

willingness to adopt e-purchasing. Vijayasarathy

states that prior experiences with online shopping

have a strong positive effect on future shopping

intentions (Vijayasarathy, 2002). Experienced

customers are thus more willing to adopt e-

purchasing than inexperienced ones. This positive

attitude is not restricted to the already known e-

suppliers, but also applies to new unknown

suppliers, as well as to the purchasing of products

not already bought on line.

2.2.3 e-Sales Channel Evaluation

While “e-commerce experience” provides a measure

of the general attitude of a customer towards e-

purchasing, the attribute “e-sales channel

evaluation” focuses on a dyadic customer-supplier

relationship. According to Choi (Choi, 2004), it

constitutes an important factor influencing the

willingness to adopt e-commerce. This attribute

takes into account the perceived usefulness (relative

advantage) and ease of use (simplicity and

compatibility) of the available e-sales channels.

The comparison of the perceived quality of

various shopping features, between traditional and e-

commerce sales channels (Levin, 2003), as well as

Table 1: B2B customer characteristics.

Basic ICT resources Sophisticated ICT resources IT training and learning

e-purch.

Use

comput

Internet

Access

E-mail www Intranet Extranet LAN

Wireless

LAN

house

Third

party

Personal

learning

staff

recruit

Website

>26% 100 100 100 98.52 63.31 32.25 86.4 36.09 54.70 62.43 81.77 41.44 82.32

5-25% 100 99.70 98.51 95.96 60.84 28.70 80.9 35.43 55.16 61.14 79.37 32.44 83.11

< 5% 100 99.47 99.05 94.01 50.37 18.19 73.6 24.08 50.79 55.42 67.61 21.66 75.50

no 100 92.65 88.51 79.22 34.38 12.42 51.6 15.96 35.14 39.23 48.44 11.91 57.56

WEBIST 2007 - International Conference on Web Information Systems and Technologies

100

among different type of e-commerce sales channels

(Choi, 2004; Lowengart, 2001), shows the strong

correlations and tradeoffs between the different

features and suggests the use of methods providing

an overall evaluation of the quality of each sales

channel. An approach based on the Conjoint

Analysis has been proposed by Canetta (Canetta,

2004) for the evaluation of online customer

satisfaction in a B2B environment. It allows

estimating the importance of purchasing process

modifications, available e-services and customer

adoption efforts. A positive e-sales channel

evaluation implies a good fit between the features of

the available channel and the customer needs; this in

turn explains the strong positive influence of this

attribute on customer willingness to adopt e-

purchasing.

2.2.4 Purchasing Frequency

The decision to adopt a particular e-sales channel is

also influenced by the comparison of its perceived

usefulness, for instance the transaction costs

reduction, with the required customer effort,

including the ICT investments and the initial loss of

productivity due to the learning effect. Thus,

customer purchasing frequency has a great impact

on the e-purchasing adoption decision, because only

a sufficiently high purchasing frequency leads to the

critical mass that makes e-purchasing adoption

profitable. Peterson introduces purchasing frequency

as one of the three dimensions of his product

classification system (Phau, 2000; Peterson, 1997).

Subramaniam (Subramaniam, 2002) indicates

purchasing frequency as one of the transaction

characteristics influencing the impact of “web-based

B2B procurement”. Consequently, a high purchasing

frequency is considered having a moderate positive

impact on e-purchasing adoption.

2.3 Purchasing Process Modifications

E-purchasing adoption can induce significant

modifications to the customer purchasing process.

Therefore, a realistic estimation of the customer

perceived usefulness (relative advantage) requires a

detailed analysis of the purchasing process

modifications caused by the adoption of a particular

e-sales channel configuration. The process

modification assessment should take into account

the production, distribution, inventory management

and order-handling processes (Clemons, 2002).

Various methods can be used for calculating the

impact of process modifications. For instance,

Subramaniam proposes a “framework to quantify

and measure the value of B2B e-commerce systems

and identify the factors that determine this value”

(Subramaniam, 2002). The evaluation of the

potential transaction cost savings obtained using

various e-sales and e-procurement channels in a B2B

market environment is also the main objective of the

work of Iliev (Iliev, 2004).

2.4 Brand Perception

The impact of brand perception on customer

satisfaction and intention to purchase has been

confirmed both in the B2B (Bendixen, 2004) and in

the B2C (Levin, 2003) marketplace. The moderating

effect of brand perception over the perceived risk is

particularly important for transactions involving new

potential customers, which are characterised by high

levels of risk and uncertainty. On the contrary,

customers belonging to the current customer base

have already acquired knowledge about the product

characteristics and the supplier performances during

their previous transactions, even if undertaken using

traditional sales channels. The following three

factors have been identified in order to describe the

influence of brand perception: brand awareness,

brand reputation, distance.

2.4.1 Brand Awareness

It is defined as the degree to which target customers

recall a brand or are aware of its existence. Brand

awareness is a common measure of marketing

communication effectiveness. High brand awareness

facilitates e-purchasing adoption, even by customers

belonging to market segments not already covered

by the current customer base. Brand awareness can

reduce perceived risk and uncertainty. For instance,

Van den Poel (Van den Poel, 1999) identifies well-

known brands as a significant risk reliever for the

adoption of e-commerce, just behind money-back

guarantee.

2.4.2 Brand Reputation

Brand reputation is defined in terms of the perceived

quality associated by a customer to a brand. Brand

reputation, also called brand image, can improve

customer trust and is considered as a moderating

factor of perceived risk and uncertainty. For these

reasons it has a great positive influence on customer

willingness to adopt e-purchasing (Lim, 2003;

Clemons, 2002).

A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION OF THE EVOLUTION OF E-SALES

101

2.4.3 Distance

Distance describes the difficulty for a potential

customer, called New Customer in the rest of the

paper, to know the seller. It is clearly influenced by

the geographical locations of the buyer and the

seller, but it also captures the lack of knowledge of a

buyer belonging to an industrial sector radically

different from those usually targeted by the seller.

Distance allows to take into account that the

perception of brand related attributes varies across

customers and purchasing situations (Mudambi,

2002). Distance has a negative impact on brand

perception and tends to hinder e-purchasing

adoption.

3 RULE-BASED DECISION

SUPPORT SYSTEM

The utility of rule-based Decision Support Systems

(DSS) in the field of e-commerce is already

demonstrated by Zhou, who points out the broad

managerial implications of the rules he identified

(Zhou, 2004). These qualitative rules can be directly

used by decisions makers to support business

strategic planning of pure e-commerce and multi-

channel retailers. A similar objective is pursued by

the DSS presented in this paper. It allows to classify

the success of an e-strategy, in terms of rapidity of

sales turnover increase and of maximum turnover,

depending on the targeted customers, the sold

products and the adopted e-sales channel. The DSS

is formalised through a series of “IF… THEN” rules

dealing with linguistic variables. The choice of

linguistic variables (e.g. low, medium, high) is

justified by the qualitative and subjective nature of

many of the factors that are linked to customer

perception. Moreover, for the company management

is simpler providing qualitative estimations rather

than gathering e-commerce quantitative data (these

are particularly difficult to obtain for activities, such

as e-commerce, that are still in their infancy).

In order to keep the sets of rules to an

understandable and manageable size, the DSS data

are structured into three categories (e-commerce

success factors, aggregated factors, output variables)

as shown on figure 1. The e-commerce success

factors, defined in section 2, appear on the left hand

side of figure 1. The aggregated factors called

Product Suitability, Customer Propensity, and Brand

Perception are obtained by combining the e-

commerce success factors. The aggregated factors

appear in the central part of figure 1. Product

Suitability characterizes the facility to which a

product can be transacted via e-purchasing.

Figure 1: Rule-based Decision Support System structure.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

102

Customer Propensity describes the impact of

customer e-readiness and of its overall evaluation of

e-purchasing in comparison with the traditional

channels. Brand Perception represents brand

awareness and reputation as well as distance impact

on the customer perceived risk (it is used only when

considering the potential transactions made by the

New Customers). Finally, Process Modification

estimates the impact of the purchasing process

modifications induced by the e-purchasing adoption

(see 2.3).

The output variables, Saturation Level and Speed

of Adoption, shown on the right hand side of figure

1, provide a complete and reliable description of

typical e-commerce evolution patterns. They are

closely related to the parameters of Diffusion of

Innovation (DOI) mathematical models, which are

already abundantly used for the description of ICT

adoption (Teng, 2002). This allows to easily enrich

the DSS, linking the qualitative description to

quantitative data. In order to limit the efforts

required from the user to fill the inputs and to render

at this level the system of rules more understandable

only three linguistic levels are used for describing

the e-commerce success factors (low, medium, high).

On the other hand, five levels are used for the

aggregated factors and the output variables (very

low, low, medium, high, very high) in order to better

differentiate the e-commerce evolution patterns.

3.1 Rule Sets for Aggregated Factors

Each set of rules is constructed starting from the

identification of the direction (positive or negative)

and of the magnitude of the influence of each e-

commerce success factor. The direction and the

magnitude of the impact of e-commerce success

factors are depicted in figure 1 using the following

notation: (++) highly positive, (+) positive, (-)

negative, (--) highly negative. As an example, in the

set of rules leading to the values of the aggregated

factor Product Suitability, three e-commerce success

factors have a positive influence (degree of

intangibility, degree of standardization, degree of

membership to the search good class) while one

(importance) is negatively correlated to the

suitability of a product to be electronically

transacted.

The most important factors are degree of

membership to the search good class, importance

and degree of intangibility, while the factor degree

of standardization is less influent. This set of rules,

defining the values of the aggregated factor Product

Suitability, results in a full factorial combination of

the levels of the four inputs (3

=81 rules) described

in detail in table 2.

The four e-commerce success factors

characterizing the aggregated factor Customer

Propensity to adopt e-purchasing have all a positive

influence. Two factors (e-commerce experience, e-

sales channel evaluation) are considered the main

drivers of customer behaviour. The remaining two

factors (ICT resources, purchasing frequency) are

considered as less important.

The perceived risk for New Customers is clearly

higher than for customers that already know the

supplier and its products. Thus, a supplementary

aggregated factor (Brand Perception) is introduced

in the case of New Customers. It accounts for the

negative effect of distance (--) and the positive

effect of brand reputation (++) and brand

awareness (+) on the new customers’ behaviour.

3.2 Rule Sets for Output Variables

For customers belonging to the current customer

base, only three aggregated factors are considered

(Product Suitability, Customer Propensity and

Process Modification) for the determination of the

Saturation Level and of the Speed of Adoption.

Saturation Level is mainly determined by the

perceived usefulness. thus the main adoption driver

is Process Modification (++), which captures the

purchasing company long term benefits resulting

from the utilisation of the e-sales channels. The

other factors, Product Suitability (+) and Customer

Propensity (+), reinforce the positive effect of

Process Modification but contribute to Saturation

Level in a lesser manner. In fact, these two factors

are mainly related to the risk perception and tend to

have a more significant impact on short term

decisions. Product Suitability (++) is considered as

the most important factor fostering e-commerce

Speed of Adoption, because product e-commerce

suitability drastically decreases customer perceived

risk. A positive influence characterizes also the

relationship between Speed of Adoption and

Customer Propensity (+) as well as that concerning

Process Modification (+). For New Customers, the

aggregated factor Brand Perception should also be

taken into account. The latter has a positive

influence in both the sets of rules, but it is never the

most important adoption driver (+).

All the rule sets are developed and formalised

according to the format presented in table 2 and are

available upon request to the first author.

A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION OF THE EVOLUTION OF E-SALES

103

Table 2: DSS rules for the aggregated factor Product Suitability.

Inta Imp Sear Stan Pr Inta Imp Sear Stan Pr Inta Imp Sear Stan Pr

L L L L

L M L L

L H L L

L L L M

L M L M

L H L M

L L L H

L M L H

L H L H

L L M L

L M M L

L H M L

L L M M

L M M M

L H M M

L L M H

L M M H

L H M H

L L H L

L M H L

L H H L

L L H M

L M H M

L H H M

L L H H

L M H H

L H H H

M L L L

M M L L

M H L L

M L L M

M M L M

M H L M

M L L H

M M L H

M H L H

M L M L

M M M L

M H M L

M L M M

M M M M

M H M M

M L M H

M M M H

M H M H

M L H L

M M H L

M H H L

M L H M

M M H M

M H H M

M L H H

M M H H

M H H H

H L L L

H M L L

H H L L

H L L M

H M L M

H H L M

H L L H

H M L H

H H L H

H L M L

H M M L

H H M L

H L M M

H M M M

H H M M

H L M H

H M M H

H H M H

H L H L

H M H L

H H H L

H L H M

H M H M

H H H M

H L H H

H M H H

H H H H

INPUT (3 levels) Imp = importance; Inta = Intangibility; Sear = Search goods class; Stan = Standardisation

OUTPUT (5 levels) Pr = product suitability for e-purchasing

LEVELS VL = very low; L = low; M = medium; H = high; VH = very high

4 ANALYSIS OF THE ICT

SERVICES SECTOR

In order to demonstrate the benefits of the proposed

DSS, it has been applied to the analysis of a

company belonging to the ICT services (ICTS)

sector. The qualitative value of the output variables,

obtained applying the DSS rules to the e-commerce

success factor values, are compared with

quantitative data concerning its current e-sales. This

allows determining if the DSS results are consistent

with the observed e-commerce evolution.

The example of Cisco Systems (CS), as the

worldwide leader in networking for the Internet

supplying both physical products (routing and

switching systems) and services, is presented and

analysed with respect to the developed DSS factors

and rules. CS mainly works for big companies in the

telecommunication sector. The characteristics of the

aggregated factors are:

Product Suitability

 High degree of intangibility of many of its

products (e.g. software) and high contribution

of services;

 Medium importance: these products/services

concern mainly consulting and outsourcing

activities and are thus not directly connected

to the customer core competencies and value

proposition;

 High degree of membership to the search

goods class;

 High degree of standardization of its products

(e.g. telecommunication) and services.

According to the DSS rules (table 2) the factor

Product Suitability takes the value Very High.

Customer Propensity

 High ICT adoption, in particular considering

the enhanced technologies (WAN, remote

access, etc.);

 Standard Internet applications (e-mail,

WWW) reaching saturation level;

 High use of broadband that renders online

transactions more pleasant;

WEBIST 2007 - International Conference on Web Information Systems and Technologies

104

 High percentage of companies already using

e-purchasing in one of its several forms;

 Very positive evaluation of the e-sales

channels implemented by CS that go beyond

the transactional aspects, also supplying end-

to-end supply chain visibility and real-time

information about lead times and available to

promise products;

 At least medium purchasing frequency,

because the customers can purchase on line

different products/services due to the wide

offer of CS (more than 22000 SKUs).

Based on these statements the aggregated factor

Customer Propensity assumes the value Very High.

Brand Perception

CS is a well-known company. Moreover, it has a

good reputation as an e-commerce champion.

Furthermore, it has at its disposal a worldwide

production and distribution network. For these

reasons its Brand Perception is Very High.

Process Modification

In general, the assessment of e-purchasing is

very positive; two third of the companies belonging

to the ICTS sector consider that it has a positive

impact on their activities. This is particularly true for

the case of CS, which provides its customers with

enhanced tools for reducing transaction process

inefficiencies. It also provides a private e-

marketplace that reduces the order cycle time while

increasing the percentage of accurate on-time

shipments. Therefore, the value Very High is

assigned to the aggregated factor Process

Modification.

Output variables

Applying the DSS rules, we obtain the

qualitative value Very High for both output

variables, Saturation Level and Speed of Adoption.

This implies that the steady state share of e-sales for

this company is close to 100% and that this share

has already reached a significant value. The

indication provided by the DSS perfectly complies

with the situation of CS that since 2001 transacts

90% of its business over the web, reaching an online

orders rate of 96% in 2006 (Kuppens, 2006).

ICT services (ICTS) sector overview

The application of the DSS gives meaningful

results to predict the e-commerce adoption process

for a given enterprise belonging to the ICTS sector

as shown in the example of CS. The validity of the

DSS results can be further assesed comparing the

overall description of the ICTS sector to those of

other industrial sectors. In fact, the analysis carried

out on the data of a survey conducted in 2003

(Selhofer, 2004) show that the ICTS sector is

characterised by the highest percentage of e-sales

adopters. The saturation level for this sector,

measured as the percentage of companies involved

in e-sales activities, is above 31%. On the other

hand, the saturation level of the overall sample,

which also includes other nine industrial sectors, is

about 17%. This value drops below 15% if the ICTS

sector is not considered. The higher e-commerce

potential (saturation level) and the faster speed of

adoption in comparison with other sectors are

moreover confirmed by the value of the e-sales share

reached in 2003, calculated as the percentage of the

monetary value of e-commerce transactions. The

latter reaches almost 4% for the ICTS sector while

the overall sample average is below 2% (1.6% not

considering the ICTS sector), thus demonstrating

how the e-commerce take-off has already happened

in this sector but still lags in other industrial sectors.

5 CONCLUSIONS AND

PERSPECTIVES

The proposed DSS has been developed in order to

qualitatively predict the success of an e-selling

strategy. It provides a complete framework

supporting the identification and the description of

the set of e-commerce success factors that

characterise a specific company. These factors are

arranged according to four dimensions: product

features, customer characteristics, process

modifications and brand perception. The choice to

rely upon various dimensions ensures a better

understanding of the company environment than that

provided by previous incomplete frameworks, which

covers only subsets of these dimensions. The

development of a rule-based DSS defined using

linguistic variables facilitates the users in the

assessment of the characteristics of the e-sales

evolution.

The proposed DSS can be considered as a first

attempt to predict the e-sales evolution and to

explain the factors influencing it. The further step,

currently under development, aims to complement

the qualitative description of the e-commerce

evolution with quantitative information. The latter,

collected through case study analysis of companies

involved in e-commerce, will be formalised relying

upon Diffusion Of Innovation theory in order to

provide mathematical models for the e-commerce

evolution quantitative prediction.

A RULE-BASED DSS FOR THE QUALITATIVE PREDICTION OF THE EVOLUTION OF E-SALES

105

ACKNOWLEDGEMENTS

The authors thank the Swiss Agency of Promotion

and Innovation for the financial support of this work,

under the grant 6066.1 (e-Fulfillment).

REFERENCES

Bendixen, M., Bukasa, K.A., and Abratt, R., 2004. Brand

equity in the business-to-business market. Industrial

Marketing Management, 33, 371-380.

Canetta, L., Cheikhrouhou, N., Glardon, R., and Iliev, N.,

2004. Supporting electronic sales channels deployment

through the analysis of customer preferences. In IEEE

Conference on Systems, Man and Cybernetics, Den

Haag, 4123-4128.

Choi, B., Lee, C., Lee, H., and Subramani, M., 2004.

Effects of web retail service quality and product

categories on consumer behaviour. In 37th HICSS

(Hawaii International Conference on System

Sciences), 177-186.

Clemons, E.K., and Aron, R., 2002. Online distribution: a

taxonomy of channel structures determinants of

outcome and determinants of strategy. In 35th HICSS

2330-2339.

Fillis, I., Johansson, U., and Wagner, B., 2003. A

conceptualization of the opportunities and barriers to

e-business development in the smaller firm. Journal of

Small Business and Enterprise Development, 10(3),

336-344.

Gurunlian, J., and Zhongzhou, L., 2001. E-commerce and

Development Report. United Nations, New York and

Geneva.

Hunter, L.M., Kasouf, C.J., Celuch, K.J., and Curry, K.A.,

2004. A classification of business-to-business buying

decisions: risk importance and probability as a

framework for e-business benefits. Industrial

Marketing Management, 33(2), 145-154.

Iliev, N., Alard, R., Canetta, L., Cheikhrouhou, N., 2004.

Evaluating Electronic Enabled Transaction Costs´

Efficiency. In International Conference on Global

Production Management, Indonesia, 21-31.

Kuppens, R., 2006. Keeping up with innovation. ELO

Innovation Challenges in Logistics, Helsinki.

http://akseli.tekes.fi/opencms/opencms/OhjelmaPortaa

li/ohjelmat/ELO/fi/Dokumenttiarkisto/Viestinta_ja_akt

ivointi/Seminaarit/ELO_loppuseminaari/20060315_Ci

sco_SCM_Helsinki_rkuppens.pdf (view 25.10.2006)

Levin, A.M., Levin, I.P., and Heath C.E., 2003. Product

category dependent consumer preferences for online

and offline shopping features and their influence on

multi-channel retail alliances. Journal of Electronic

Commerce Research, 4(3), 85-93.

Liang, T.P., and Huang, J.S., 1998. An empirical study on

consumer acceptance of products in electronic

markets: a transaction cost model. Decision Support

Systems, 24, 29-43.

Lim, N., 2003. Consumers’ perceived risk: sources versus

consequences. Electronic Commerce Research and

Applications, 2, 216-228.

Liu, J.Y., Tarn, J.M., and Wen, H.J., 2004. Adopting e-

commerce web site operation modes - the current stage

of large US firms. Information Management and

Computer Security, 12(1), 73-90.

Liu, X., and Wei, K.K., 2003. An empirical study of

product differences in consumers’ e-commerce

adoption behaviour.

Electronic Commerce Research

and Applications, 2, 229-239.

Lowengart, O., and Tractinsky, N., 2001. Differential

effects of product category on shoppers’ selection of

web-based stores: a probabilistic modeling approach.

Journal of Electronic Commerce Research, 2(4), 142-

156.

Mudambi, S., 2002. Branding importance in business-to-

business market – three buyer clusters. Industrial

Marketing Management, 31, 525-533.

Peterson, R.A., Balasubramanian, S., Bronnenberg, B.J.,

1997. Exploring the implications of the Internet for

consumer marketing. Journal of the Academy of

Marketing Science, 25(4), 329-346.

Phau, I., Poon, S.M., 2000. Factors influencing the types

of products and services purchased over the Internet.

Internet Research: Electronic Networking

Applications and Policy, 10(2), 102-113.

Scupola, A., 2002. Adoption issues of business-to-

business Internet commerce in European SMEs. In

35th HICSS, 2119-2128.

Selhofer, H., 2004. The European e-Business Report.

European Commission Enterprise Directorate General,

Bonn.

Stansfield, M., and Grant, K., 2003. An investigation into

issues influencing the use of the internet and electronic

commerce among small-medium size enterprises.

Journal of Electronic Commerce Research, 4(1), 15-

26.

Subramaniam, C., and Shaw, M.J., 2002. A study on the

value and impact of B2B e-commerce: the case of

web-based procurement. International Journal of

Electronic Commerce, 6(4), 41-72.

Teng, J.T.C., Grover, V., and Güttler, W., 2002.

Information technology innovations: general diffusion

patterns and its relationships to innovation

characteristics. IEEE Transactions on Engineering

Management, 49(1), 13-27.

Van den Poel, D., and Leunis, J., 1999. Consumer

acceptance of the Internet as a channel of distribution.

Journal of Business Research, 45, 249-256.

Vijayasarathy, L.R., 2002. Product characteristics and

internet shopping intention. Internet Research:

Electronic Networking Applications and Policy, 12(5),

411-426.

Zhou, L., Chiang, W.K., and Zhang D., 2004. Discovering

rules for predicting customers’ attitude toward internet

retailers. Journal of Electronic Commerce Research,

5(4), 228-238.

WEBIST 2007 - International Conference on Web Information Systems and Technologies

106