The impacts of E-retail on the choice of shopping trips and delivery: Some preliminary findings

Abstract

E-retail, like many other information technology-based activities (telecommuting, telemedicine etc.) offers a potential substitution of travel by telecommunications. Traditional shopping activities typically consist of a visit to a store in which product information is sought, and a decision on purchase is made. Pending that decision, the product is obtained and most often self-delivered by the consumer. Certain types of products are store-delivered to the consumer premises. In the face of E-retail, consumers can acquire information, make a purchase transaction and choose a delivery arrangement from a remote location. These options may result in a reduction of transport activity, as a delivery by the supplier is potentially more efficient than the traditional process. The current study presents a conceptual model of the decisions households make with regard to information gathering, purchase transactions and delivery mode. Data on revealed behavior and various socio-demographic and economic characteristics of shoppers was collected in the Tel-Aviv Metropolitan area in the summer of 2004.

Full text

Page 1

The impacts of E-retail on the choice of shopping trips
and delivery: Some preliminary findings
Orit Rotem-Mindalia,*, Ilan Salomonb,1
aDepartment of Geography, The Hebrew University of Jerusalem, Jerusalem 91905, Israel
bSchool of Public Policy, Department of Geography, The Hebrew University of Jerusalem, Jerusalem 91905, Israel
Received 22 November 2005; received in revised form 22 November 2005; accepted 26 February 2006
Abstract
E-retail, like many other information technology-based activities (telecommuting, telemedicine etc.) offers a potential
substitution of travel by telecommunications. Traditional shopping activities typically consist of a visit to a store in which
product information is sought, and a decision on purchase is made. Pending that decision, the product is obtained and
most often self-delivered by the consumer. Certain types of products are store-delivered to the consumer premises. In
the face of E-retail, consumers can acquire information, make a purchase transaction and choose a delivery arrangement
from a remote location. These options may result in a reduction of transport activity, as a delivery by the supplier is poten-
tially more efficient than the traditional process. The current study presents a conceptual model of the decisions households
make with regard to information gathering, purchase transactions and delivery mode. Data on revealed behavior and var-
ious socio-demographic and economic characteristics of shoppers was collected in the Tel-Aviv Metropolitan area in the
summer of 2004.
? 2006 Elsevier Ltd. All rights reserved.
Keywords: E-retail; Delivery; E-shopping; Information-gathering
1. Introduction and background
E-retail represents a small, but growing part of retail activities, which may have broad implications on the
organizational and spatial structure of retail systems, as well as shopping patterns. Such impacts depend, to a
great extent, on consumers’ response to technological change. It is often hypothesized that E-retail will sub-
stitute for physical shopping trips. Three choices that consumers make in the face of E-based options can be
examined: the collection of product information, the mode of transaction and the mode of delivery (e.g., self-
delivery, third party services).
Recent developments in information technology (IT) have expanded the range, type and number of
possible spatial interactions (Gould and Golob, 2002). These developments may also have an effect on
0965-8564/$ - see front matter ? 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tra.2006.02.007
*Corresponding author. Tel.: +972 2 5883348; fax: +972 2 5820549.
E-mail addresses: mindali@pob.huji.ac.il (O. Rotem-Mindali), msilans@mscc.huji.ac.il (I. Salomon).
1Tel.: +972 2 5880082; fax: +972 2 5881200.
Transportation Research Part A 41 (2007) 176–189
www.elsevier.com/locate/tra

Page 2

consumption-related travel. E-retail is hypothesized to create a trade-off between IT and travel, and thus sug-
gests that travel may be reduced.
Understanding the changes in households’ consumption patterns together with an analysis of the spatial
and transportation implications, in the presence of IT-based options, will assist in the formulation of transport
policy. More specifically, it can improve the forecasting of negative externalities and recommend means of
intervention to substitute travel and reduce the deterioration of the environment.
A shortcoming of the emerging body of research on tele-activities (tele-shopping included) is the lack of
behavioral-level data. To allow quantitative analysis and model estimation, the current study has collected
individual level data on shopping and the use of information technology. The data was collected in the
Tel-Aviv metropolitan area in the summer of 2004. It includes 510 valid questionnaires, based on a response
rate of about 44%.
This paper focuses on the consumers’ purchase and delivery (PD) choices, as part of a broader effort to
understand consumers’ shopping behavior. The general structure of this behavior is described in a conceptual
model shown below.
An important aspect of consumers’ choices discussed throughout this paper, is the concept of ‘‘product
class’’. It refers to a category of goods which share certain characteristics, thus making them a distinguished
group (e.g., furniture, large appliances and perishable groceries) and therefore may differentially affect individ-
uals’ shopping process.
This paper aims to give some preliminary insight on several aspects of consumption patterns in the elec-
tronic age. The paper first depicts some characteristics of the sampled population, followed by describing
the extent of E-modes’ use for purchase and information gathering and consumers’ reasoning for E-shopping.
Subsequently the paper analyses the purchase and delivery sequence, which is important for the purpose of
choice modeling. Finally the paper examines the relationship between information technologies and delivery
preferences.
1.1. Spatial considerations in shopping behavior
Individuals and households generate a spatially differentiated demand for products and services. Combined
with the demand for a certain product, service and information, consumers also exhibit a demand for access.
Accessibility may not necessarily be travel-based. It can be IT-based. Accessibility is a necessary condition for
retail activity to take place, but not necessarily a sufficient one. As retail facilities compete for customers, there
is a high premium for accessibility. This is certainly the case when physical access is necessary (road network,
parking, public transportation and sidewalks), but is also of relevance when virtual access is involved (Internet
connection, telephone etc.).
The automobile is a major factor affecting accessibility changes in metropolitan areas. The private car
reduces travel time to destinations and increases the freedom of choosing when and where to travel, but it also
increases travel time through congestion. The automobile is hypothesized, in many cases, to be the determi-
nant factor in choosing physical or virtual shopping.
With growing suburbanization, city centers have become less accessible for the private car, and their
remaining retail vitality is dependent on the accessibility provided by public transportation and pedestrian
activity. Shops in downtown areas are becoming smaller and the downtown is reshaped into a large pedestrian
precinct (Monheim, 1992). Simultaneously, unplanned sprawl of retail facilities, along major roads and junc-
tions in suburban areas, are an indication of the growing accessibility advantages of these areas, due to higher
personal mobility facilitated by the private car. These changes in accessibility, combined with the development
of expressways and beltway systems, increase the number and the scale of regional shopping centers.
1.2. E-retail
A major trend in retail, which is drawing much attention, is the employment of new technologies. This
involves a shift from some aspects of the traditional store format toward the introduction of electronic means
of performing retail activities (Mulhern, 1997). E-retail, E-commerce and teleshopping have nowadays become
common synonyms for electronic, mainly Internet-based transactions. E-retail encompasses three main
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
177

Page 3

facilities of consumption activities. Specifically, a product search facility (often referred as a product evalua-
tion or information gathering facility), an on-line purchase function and a product delivery capability (Kolesar
and Galbraith, 2000).
In the context of E-retail, accessibility has two meanings. First is the physical access, namely the consumer’s
ability to access the retail facility, or vice versa, the retailer’s ability to access the consumer’s premises. For
most product classes, physical delivery is still a necessity. The second is that of IT-based access: the telephone
or the yet less popular (but rapidly expanding) computer web access, labelled below with the generic term
‘Internet’.
Projections of the diffusion of E-shopping activity have ranged from 5 to 300 billion US$ by the year 2002
(Alba et al., 1997; Gould and Golob, 2002; Teo, 2002). The dynamics of growth in E-retail in the last decade
make data reported in scientific publications obsolete. Other sources, such as professional journals, provide
more up-to-date data. For example The Economist (2005) ran a cover story on eBay, revealing the meteoric
growth in net profit and gross merchandise volume between 1998 and 2004. Today, it seems that many fore-
casts were merely an optimistic prediction. Statistical data in both the United Kingdom and the United States
demonstrate a growth in on line retail sales, but they still remain relatively small, at or around 1–1.5% of total
retail sales (Dixon and Marston, 2002). In addition, Internet-only firms are in trouble, and what seemed to be
a golden opportunity appears to have a relatively small effect over the next few years (Rosen and Howard,
2000; Biyalogorsky and Naik, 2003; Mokhtarian, 2004). In 2003 it was estimated that there were 190 million
computer-network users around the world (Lee and Tan, 2003).
Currently, E-retail activity is concentrated in items such as books, software, music, travel, hardware, cloth-
ing and electronics, with a growing and developing sector of groceries (Gould and Golob, 2002). However, as
for the early 2000’s, people are browsing the Internet more for information than for actually buying online
(Teo, 2002; Forsythe and Shi, 2003). Using the Internet enables consumers to, fairly easily, access information
about merchandise, gather vertical information (i.e., comparing a product across suppliers) at a low cost, to
efficiently screen the offerings, and easily locate a low price for a specific item (Alba et al., 1997; Ratchford
et al., 2001).
Many studies have tried to isolate the reasons for the success or failure of E-retail as a substitute for tra-
ditional retail. Trust in online shopping is one of the main factors identified. Grabner-Kraeuter (2002) sug-
gested that trust is not only a short term issue, but is the most significant long-term barrier for achieving
the potential of E-retail. The reason for that is that buying on the Internet presents numerous risks, mainly,
transaction-related processes (Court and Dayal, 2002).
E-commerce offers increased market activity (and efficiency) for retailers in the form of increased market
access and information, and decreased operating and procurement costs. The consumers gain enhanced price
competition, expanded information on goods and services and increased choice of products (Rao, 1999; Rosen
and Howard, 2000; Mokhtarian, 2001).
1.3. Impacts on shopping-related travel
Vehicular travel generated for the purpose of shopping activity is increasingly considered to be a contrib-
uting factor to suburban congestion, and consequently harmful to the environment. Thus, it has also become a
focus for travel demand management (TDM) policy measures.
The impacts of E-retail as an acquisition mode are of interest in at least two contexts. First, there is a grow-
ing body of research addressing the question of substitution and complementarity of travel-based and E-based
options (Koppelman et al., 1991; Arnfalk, 1999; Golob and Regan, 2001; Mokhtarian and Salomon, 2002;
Teo, 2002; Lenz, 2003). If significant quantities of teleshopping activities are assumed, shopping-related travel
may be reduced. Studies of the impacts of teleshopping on transport usually assume that the delivery trip, by
the retailer or a third party, to multiple customers is more efficient than individual trips.
Second, at a more theoretical level, the option of virtual access, as opposed to real travel, raises some ques-
tions with regard to the underlying reasons of why people travel. The common assumption in travel behavior
analysis is that the demand for travel is a derived demand, where the demand is actually for activities which
can be performed at the destination. Some recent studies suggest qualifying the derived demand assumption,
by pointing to cases in which the activity of travel itself is the generator of a trip, or that travel constitutes an
178
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 4

auxiliary factor (Mokhtarian and Salomon, 2001). Shopping may very well be an activity that generates a trip,
or part of a trip, even when a purchase is not planned or performed.
2. The conceptual model
The general framework with which we suggest studying the impacts of E-retail on shopping-related travel is
shown in Fig. 1.
The model consists of six blocks (shown in Roman numerals). Block I represents the household, which
serves as the basic unit of analysis, as it is assumed to be the most basic unit of consumption. Within this
block, three elements are shown: the socio-economic and demographic (SED) attributes, the lifestyle prefer-
ences of the household members and the demand for goods, emanating from the previous two. One of the
variables that represent the SED context is consumer affinity to IT. Similar to variables that represent acces-
sibility to a private car, and car ownership, the variable affinity to IT is an indicator that defines consumers’
accessibility to technology in general and accessibility to the Internet in particular. It comprises elements of
computer ownership, Internet access and level of use.
Block II represents the retail system and there is a two directional interaction with Block I. It represents the
response to the demand which may affect the household demand. Block III represents a choice of the acqui-
sition process an individual decides to engage in. Block IV represents the travel decisions that are consequen-
tial to the acquisition process. Note the upward pointing arrow from Block IV to Block III which represents
the effect of the expected costs of travel on the choice of the acquisition mode. Block V represents the travel
and transportation impacts and Block VI represents the policy intervention options.
In the shopping process (III) we distinguish between two activities: information gathering (IG) and acqui-
sition. Acquisition, in turn, includes two components: purchase (P) and delivery (D). The shopping process
may be viewed as a two-phase process (Fig. 2). The first phase includes, in response to a need for a product,
the choice of the information-gathering mode, i.e., virtual or physical (this may depend on consumer affinity to
technology). The information collected concerns both product and transaction related information. IG can be
performed in an iterative manner, where new sources of data are consulted when prior search efforts fell
short of supplying sufficient information (Salomon and Koppelman, 1992). While there is usually a range
Retail
Structure
Environmental
Externalities
Policy
Shopping
Process
Travel
Behavior
III
II
IV
V
VI
SED
Context
I.I
Consumer
Demand
I.II
Households
I
Lifestyle
I.II
Fig. 1. General conceptual model.
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
179

Page 5

of information gathering options (e.g., by telephone, Internet, advertisement, asking a friend, going to a store),
there are fewer alternatives for purchasing procedures. Thus, the purchasing element may be the constraining
factor. When sufficient information is gained the individual decides to exit the market or stay in it i.e., make a
purchase by moving to phase 2.
The decision on how to purchase is combined with the decision on the mode of delivery. Individuals may
decide on purchase and delivery simultaneously (P = D) or in a sequence (i.e., P ! D or D ! P). These three
options constitute different choice sets and consequently different behavior patterns. It is plausible to hypoth-
esize that the P = D option indicates a situation where neither action implies a significant burden. By contrast,
the two sequential options (P ! D or D ! P) imply that one or the other is more burdensome. In the former
(P ! D), the delivery is of a lesser importance compared to the utility of purchasing in a particular setting. In
the latter (D ! P), the delivery is overriding in its importance to the individual. This may be the case of large
items, which do not necessarily fall in the option where store delivery is always available (e.g., large furniture,
large electric appliances).
The consumer who decides on a ‘‘purchase-self-delivery’’ combination must generate at least one trip to the
retailer to pick up the good. The actual purchase may be done by IT, but given that a trip is made, purchase is
most likely to be done at a store. If the consumer decides to use any other mode of delivery, a trip can poten-
tially be eliminated. In this case a delivery trip will be made by the retailer or by a third party. It is also inter-
esting to examine whether consumers who prefer non-self delivery are those who have higher affinity to
technology and use virtual modes for remote activities.
The travel made by the consumer should be characterized in terms of mode, destination, time of day and
consequent VMT. It is plausible to assume that the decision rule employed by consumers is a utility maximi-
zation process. On the other hand, a commercial delivery involves decision on optimal route and timing. That
decision is likely to be based on cost minimization or profit maximization.
The frequency of trips depends on two main characteristics: product class and consumers’ attributes. Prod-
uct class attributes, such as shelf life and perishability, or size and value, affect the frequency of acquisition.
Retail activities involve a wide range of goods (and services) that differ in attributes relevant to the current
context, travel and delivery of goods. A classification of goods along the attributes which bear upon the trans-
portation factors is warranted. There are a variety of ways in which goods and services may be classified
(Chiang and Dholakia, 2003; Hsieh et al., 2005; Klein, 1998). For the purpose of the current discussion the
relevant classification is best attained by using product class as a distinguishing factor. Product class refers
to a range of attributes that distinguish one group of products from another, by means of factors relevant
to information gathering, purchase and delivery. For example, size and weight distinguish furniture and major
Information
Product related
Transaction related
Purchase
?
Need
Exit
Phase 1
Phase 2
Mode of
IG
Purchase
and Delivery
Fig. 2. Shopping process: a two-phase model.
180
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 6

appliances from most other goods, perishable food products create a class of their own, and price may distin-
guish between a home computer and a book.
Belonging to a particular product class has implications for the manner in which individuals gather infor-
mation, purchase and choose among alternative modes of delivery. Thus, product class can be seen as an
explanatory variable. In some cases the product class attributes imply advantages or disadvantages of partic-
ular IG or delivery form and in others they may act as a constraint. Consider, for example, shopping for a high
quality sound system. Most consumers will prefer listening to candidate products, hence in-store information
gathering will be preferred over other information sources. A different example is that of the product class in
which the basic attributes are large size and weight, which in furniture and large appliances serve as a delivery
constraint, requiring professional delivery.
The second characteristic is the consumers’ attributes, as often described by socioeconomic and demo-
graphic characteristics. Economic status attributes, such as income level and leisure time availability, may
have a direct effect on when and where a product is being bought. This may create a situation where house-
holds characterized with medium and high-income levels may have the time and mobility abilities to reach
lower price retail centers, usually on the outskirts of cities to which the less mobile sectors of the population
are less accessible. The same is true for household age composition and size. Residential patterns, distance
from store, and consumers’ mobility and accessibility attributes, both physical and virtual, may create con-
straints to affect consumer behavior.
Another attribute of the individual is lifestyle. Local fads, culture and ethnic identity may shape the fre-
quency of trips. For an example, a noticeable difference between the US and Europe is in the frequency of
buying groceries. While in Europe this activity is often an everyday chore and it is a custom to have fresh
products every day, in the US it is a less frequent task, resulting in a larger amount of groceries being bought
each time (Bell et al., 1997).
3. The data
To test the hypotheses emanating from the conceptual model, it is necessary to collect data on the individ-
ual decision maker level. We chose to collect structured questionnaires from a sample of residents of the Tel-
Aviv metropolitan area. The data was collected through face-to-face interviews in the summer of 2004. Each
interview lasted some 30 min.
At present, there is a noticeable lack of quantitative studies of consumers’ behavior in the context of E-
retail. This can be attributed to at least three factors. First, much of the activity in this area lies within the
scope of the private sector, which is concerned about commercial confidentiality. Second, E-retail is at its
infancy. This implies fast dynamics, which have a negative effect on data quality due to significant noise levels.
Third, the lack of well developed theoretical constructs in this field implies that it is not quite clear which data
should be collected.
As noted earlier, one of the objectives of the research is to understand households’ consumption patterns in
view of recent technological changes and advances. Therefore, we sampled from a population with an expected
high affinity to technology, on the basis of income and education. The respondents to the survey were chosen
from five cities in the metropolitan area of Tel-Aviv, selected from a sector that stretched from the core of the
metropolitan area (Tel-Aviv) to the northeastern outer ring. These sections are considered to have, on average,
a high socioeconomic level. Under the assumption that there is some correlation between socioeconomic level
and technology adoption, these cities were selected for the study.
The respondents were either one of the heads of the household, namely a salaried adult, and were all above
the age of twenty. This was done with the intention of receiving answers from the person who makes most
consumption decisions given the household’s budget. While individual members of a household may choose
to purchase their own clothes or some food separately, the overall budget constraint and responsibility usually
lies on one or two adults in the household (Jones and Simmons, 1990).
The study uses four product classes: perishables, grocery (non perishable), small electric appliances, and
large electric appliances. For each of the product classes, questions regarding the acquisition process were
asked with the intention of understanding the differences among basic types of manufactured goods. To
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
181

Page 7

eliminate respondents’ fatigue, each questionnaire included just two product classes. Thus, six versions of the
questionnaire were used to allow a full design.
The questionnaire consists of four sections. The first provides information on experience with retail and
E-retail, as well as attitudes towards shopping. The second section reveals the respondent’s shopping process,
in which each respondent provided data for two (out of the four) product classes. The third section deals with
the choice of shopping facility, emphasizing the spatial dimension. The fourth and last section collects data on
individual’s and household characteristics. Given the complexity and length of the questionnaire there was a
clear preference for face-to-face questionnaire filling.
3.1. Characteristics of the respondents
This section introduces some background information on the sample population. The purpose of this sec-
tion is to reveal basic information on the households’ SED characteristics. Some of these characteristics, such
as age and level of education, are considered to be correlated with shopping behavior. We currently present
only initial analysis, to be articulated in a subsequent paper.
The respondents were asked to assess a large variety of aspects concerning their shopping habits, prefer-
ences as well as attributes concerning their accessibility to and usage of IT-based applications. Since one of
the main objectives of this study is to examine the impacts of technological advances on households’ consump-
tion patterns, we decided to recruit households with potentially high affinity to technology. In this section we
will present two of the attributes that exemplify the characteristics of the research population.
An important indicator for the potential adoption of new information technology is the level of education
of the respondents. Since we try to understand households’ consumption patterns, it is necessary to also con-
sider the same indicator for the partner’s level of education. Table 1 presents the data on education level of the
respondents and their partners (if existing), in comparison to the level of education of Israel’s population.
For both respondents and partners, the level of education is fairly high, in comparison to the general pop-
ulation. For the respondents, more than 50% have an education level higher than high school. The same share
of high level of education is true for the respondents’ partners, when eliminating cases of no partner or missing
values. When compared to the general Israeli population (Israel CBS, 2004), the share of education higher
than high-school in the general population is about 38%. This finding is important for the analysis, since
we prefer to test a sample in which the level of education is higher than the population average.
The age distribution is important when dealing with the adoption of new information technologies, since
high positive correlations between age and exposure to ICT, as well as usage of ICT, have been shown in pre-
vious studies.
In this survey, as shown in Table 2, a large share of the respondents (47.4%) are in the age group of 24–55,
in comparison to 42.5% of the Israeli population in the same age group (Israel CBS, 2004). The share of
respondents 65 years or older is identical for the sample and the population. The distribution of gender in
the survey is approximately the same for male and female (48.8%, 51.2%, respectively), similar to the general
population. The relatively high proportion of respondents between ages of 24 and 45 may affect the adoption
Table 1
Level of education (%)
Level of education Respondents’Partners’a
Israel’s population (CBS, 2004)
Primary/elementary school
High school (partly)
High school
Incomplete university degree/vocational
Bachelor’s degree
Master’s or higher degree
No education
Missing
7.6
9.8
27.3
17.8
25.3
12.2
9.4
7.4
30.1
15.0
28.0
10.0
15.4
43.3
13.2
24.9
3
0.2
Total100100100
aNo partner and missing values are not included.
182
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 8

of advanced information technology in the sample. But since we aim to find a population with high affinity,
which potentially uses the Internet for shopping activities; their high representation is not considered to be a
problem.2
4. Some preliminary findings
In this paper we focus on the purchase and delivery stages of the acquisition process. The motivation for
this focus is twofold. First, we seek to understand the decision sequence of the consumer when facing the need
to acquire a product (moving from the first phase to the second shown in Fig. 2). This separation will improve
the ability to model the acquisition process. Second, we attempt to identify the impacts of information tech-
nologies on delivery patterns.
In the following analysis of the purchase-delivery (PD) sequence, we employ the division to the four prod-
uct classes mentioned above.
4.1. Some general findings
The data collection instruments provided a large array of data on various aspects hypothesized to influence
E-retail behavior. In what follows, we present the preliminary findings and analysis of the data.
Evidence on the use of IT-based options for shopping, indicates that about 15% of the respondents use this
option ‘rarely’ or ‘occasionally’, while the remaining 85% do not use IT for shopping at all. Some 47% of the
cases report that they are highly exposed to technological advances, in comparison to only 30% with no access
to IT. Despite the popular notions about a rapid growth in the popularity of E-retail, none of our respondents
reported using IT-based options ‘often’ or ‘always’. This may hint to the possibility that the growth of E-retail
is more in population size than in intensity.
There is a clear difference in the use of IT-based modes of acquisition, among various product classes.
Table 3 demonstrates that the tendency to use IT-based modes for purchasing products is considerably stron-
ger for the electronic goods and appliances compared to groceries and perishable goods. Although only a
small proportion of the sample declared they use IT-based options for purchasing products, we find it useful
to analyze their consumption behavior. The consumers composing this small proportion of IT users are ‘‘early
birds’’ in their behavior. Though their behavior may be different from the mean behavior of the total
population, they can serve as an indicator for future trends.
Using the Internet for information gathering, as shown in Table 4 and in comparison to Fig. 2, reveals that
the use of IT as an electronic mode for starting the acquisition process is preferred. There is a larger propor-
tion of respondents that do use the Internet for collecting information.
Fig. 3 shows the stated reasons for purchasing products using IT for groceries and large appliances,. The
most popular response for both cases was product price. However, for electric appliances price was the reason
mentioned by almost 60% of the respondents, while for groceries only 34% considered this to be the main
Table 2
Respondents’ age distribution
Age distribution Percentage (n = 510)Israel’s population (CBS, 2004)
Less than 24
24–45
46–64
65–74
74 and more
Missing
10.6
47.4
26.1
8.6
7.1
0.2
12.9
42.5
28.9
8.6
7.1
Total 100100
2Nowadays the proportion of people who uses electronic modes of shopping is fairly small. Therefore we wanted to oversample younger
adults, having a potentially higher affinity to ICT.
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
183

Page 9

reason. A time constraint, as represented by ‘no time’, was the second most popular reason. Time constraints
were identified as an important advantage when electronic modes of shopping were considered.
It seems that time constraints play an important role when groceries are concerned. For grocery shopping
via the Internet, more than 27% chose this as the main reason for E-shopping, and another 7.8% chose ‘no
car’, which may also represent a time constraint, as the most important reason for E-shopping. For electric
appliance shopping, ‘no time’ was selected by only 12.6% and ‘no car’ by 4.3%. These findings represent some
of the differences that diverse product classes offer.
4.2. Purchase and delivery (PD) sequence
When considering the whole data set, clearly the preferred sequence is that of P ! D. Some 60.7% of the
respondents choose that order.3Ranking second, the option of simultaneous PD was preferred by 32.6%
(Fig. 4).
Table 3
Use of IT-based option for purchasing products by product class (%)
Perishable Grocery Large electricSmall electric General
Never
Sometimes
90.3
9.7
90.6
9.4
74.1
25.9
83.3
16.7
84.6
15.4
Total 100100100100100
3This percentage is out of 852 responses. The remaining 168 responses are respondents with no preference or those who did not want to
answer the question.
0.0
no time
10.0
20.0
30.0
40.0
50.0
60.0
70.0
large distance
urgent shopping
as leisure
large shopping
product price
guarantee
diversity
no car
no reason
percentage (%)
ELECTRIC APPLIANCESGROCERIES
Fig. 3. Reasons for E-shopping.
Table 4
Use of IT-based modes for information gathering by product class (%)
Perishable GroceryLarge electricSmall electric General
Never
Sometimes
Often
93.1
5.8
1.2
88.2
8.6
3.1
61.6
17.3
21.2
72.1
16.7
11.2
78.8
12.1
9.1
Total100.0100.0100.0100.0100.0
184
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 10

About 74% of the respondents expressed a preference for self-delivery. When this group is eliminated from
the data set, different PD decision structures are revealed.
In an ANOVA test we identified two, intuitively acceptable, groups of product classes that are significant.
In the first group one can find grocery (GRO), perishables (PERI) and small electric appliances (SE), while the
other group included only large electric appliances (LE). Most of the respondents (70%) who preferred non-
self-delivery referred to the large electric appliances product class (LE).
When analyzing the frequencies of PD sequence decisions across product classes, one can see that in the
first group of products (GRO, PERI and SE) the distribution of the decision on PD sequence is altered
(Fig. 5).
Specifically, for the SE product class the most desired option (43%) was the P = D ‘sequence’, while the
sequential options (P ! D or D ! P) have the same share of preferences. In the GRO and PERI product clas-
ses, the preferred option is the P ! D sequence (42%, 43% respectively). The difference between the two prod-
uct classes is in the second and third preferred option. In GRO, the second choice was the D ! P sequence
(33%), while in PERI, it was the P = D ‘sequence’ (38%).
GRO and PERI products can generally be found in the same store, and consequently one may expect them
to exhibit similar PD sequences. However, as seen in Fig. 5, there is a noticeable difference in the second most
popular mode of decision. The fact that in the PERI product class the D ! P sequence is the least preferred,
probably stems from the fact that consumers are less comfortable to rely on others when perishable products
are concerned. We have found that when perishable products are considered there is a significant difference
between men and women. Women were found to worry more about the goods delivered in comparison to
men. Whereas when grocery products (non-perishables) are considered, neither men nor women are concerned
about the goods delivered.
In the LE product class the results were similar to those obtained in the total sample; 64% selected the
P ! D sequence. A possible explanation may lie in the fact that most LE products cannot be self-delivered.
Probably, almost all selling locations provide the service of delivery. Consequently, it does not play a role in
affecting the decision on the location of purchase. A delivery constraint that was found to act as a consumers’
restriction for non-self-delivery is the size of the shipment. We have found that 62.7% of the respondents iden-
tify this attribute of the product as a reason for using non-self delivery. Special conditions required, such as the
need for a technician to install the product, was selected by only 14.1% of the respondents.
6.7
60.7
32.6
D->P
P->D
P=D
Fig. 4. Purchase and delivery decision.
24
38
7
29
43
42
64
29
33
21
29
43
0% 20%40% 60%80%100%
perishable (24)
grocery (33)
large electric (157)
small electric (7)
Product class
percentage from non-self delivery (n=221)
D->P P->DP=D
Fig. 5. PD decision according to product class.
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
185

Page 11

In analyzing the sequence of decision on shopping activities, we found that consumers usually decide first
on the location of purchase. The delivery decision tends to follow and correspond to the first decision. This
result suggests that people grasp the shopping process options as a function of the location preferred. However
for the small fraction of consumers who uses self delivery, the results are not as clear. We find variation in the
sequences for the different product classes. This finding suggests that when delivery services are recognized as
important the choice set of the shopping process changes.
4.3. Delivery decision and affinity to technology
E-retail is technology dependent, and users must be technology-literate to properly and independently
adopt it. Therefore, we have hypothesized that the exposure to computers and to the Internet can contribute
to the explanation of E-retail usage patterns.
This section presents some initial analysis results of the relationship between the delivery decision and new
information technologies. In particular, we examine the impact of affinity to technology on the use of non-self
delivery with growing exposure to information technology. Affinity to IT is defined as an indicator composed
of several technology-related attributes which characterize consumers.
The underlying factor seems to be the quality of access. It, in turn, is determined by a number of technol-
ogy-related attributes which we explored in the questionnaire. First, we asked about whether or not the
respondent had any experience with purchasing or gathering information using electronic means, and the
intensity of such experience. Second, the respondents were asked about access to computers and the Internet,
at home and at work. In order to describe in one indicator all these aspects, a new variable was calculated as a
composite of the above-mentioned variables.
Three clusters of affinity to ICT were developed and their attributes are shown in Table 5. The low affinity
cluster is a group with a majority of respondents who does not own a computer (0 – do not own a computer;
1 – own a computer). This cluster has no accessibility to the Internet at home (1 – no accessibility; 2 – slow
(dialup connection via modem); 3 – fast (broadband)), do not use the Internet at work (0 – do not use; 1 – use),
and most of the respondents in this group do not have any experience using the Internet (0 – no experience; 5 –
more than two years). Next is the medium affinity cluster. The respondents in this cluster have a computer at
home and the majority of respondents also have a broadband connection to the Internet. The mean of fre-
quency using the Internet, not work-related is 1–3 times a month (0 – do not use; 5 – almost every day and
more), and they do not use the Internet at work. The third and final cluster is the high affinity cluster. This
cluster is composed of respondents who have a home computer, with fast connection to the Internet. The
respondents in this cluster use the Internet very frequently, the mean frequency is 1–4 times a week, with most
of the respondents using the Internet almost every day. The respondents in this cluster use the Internet also at
work and have experience of more than two years.
The central focus of analyzing different IT affinity consumers is to evaluate how consumers respond to
delivery options. We examine if there are differences in delivery behavior between clusters. An ANOVA anal-
ysis is performed to identify if, and what, dissimilarities exist both for groceries and electric appliances (Figs. 6
Table 5
The three clusters of consumers’ ICT affinity
IT affinity clusterComputerInternet accessibility Internet frequencyInternet workInternet experience
LowMean
Std. dev.
N
0.287
0.453
334
0
0
334
0
0
334
0
0
334
0.012
0.109
334
MediumMean
Std. dev.
N
1
0
244
1.852
0.355
244
3.320
1.816
244
0
0
244
0.303
1.161
244
HighMean
Std. dev.
N
0.917
0.276
362
1.624
0.731
362
3.829
1.855
362
1
0
362
4.751
0.758
362
186
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 12

and 7). In the analysis we found that for groceries we have not observed dissimilarities between the clusters, as
shown by the horizontal line cutting through the three diamonds in Fig. 6.
Dissimilarity between clusters does occur in the case of delivery of electric appliances (Fig. 7). The clusters
of medium and high IT affinity indicate higher use of delivery by others for purchasing electric appliances,
while the low IT affinity cluster uses only self delivery.
5. Conclusions
This paper is part of a growing body of research which focuses on the question: Are information technol-
ogies likely to substitute travel and thereby reduce the negative externalities of transportation systems?
In the last two decades research efforts have addressed this question mainly in the context of telecommuting
(telework) and to a lesser degree in teleshopping. At present, two major conclusions have been brought for-
ward. First, that substitution, if evident, is of very limited magnitude. Second, that much of the expectations
with regard to behavioral changes (e.g., substitution) are based on rather simplistic assumptions. The transi-
tion from physical travel to virtual movement cannot be explained solely by cost saving or by the availability
of the technology.
The adoption of E-retail depends on a variety of factors which may entail significant changes in life styles. It
seems that our ability to forecast profound changes in behavior is very limited. In the present context, the role
of the sequence of choosing purchase and delivery is a case in mind.
Fig. 6. Oneway analysis of delivery and ‘IT affinity’ cluster for groceries.
Fig. 7. Oneway analysis of delivery and ‘IT affinity’ cluster for electric appliances.
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
187

Page 13

An underlying assumption of the current research is that IT expands the choice set which individuals con-
sider in planning their (shopping) activities. The wider choice set opens up possibilities for new patterns of
shopping which were not available heretofore, including ‘virtual travel’ and innovative distribution and logis-
tics techniques. It is not unusual nowadays to observe situations in which the simple daily trip to the street-
corner grocery store is being replaced by a more or less complex bundle of shopping activities. Such activities
may take advantage of IT in response to temporally and spatially differentiated specialized needs.
Moreover, the expanded choice set in shopping behavior opens the way for other changes in the activity
program of the individual. Bypassing the trip to the store, or frequenting the store but eliminating the need
to carry goods may relax some common constraints and open the way for more flexible activity planning,
including, of course, non-shopping activities. In the long ran, such changes may even lead to a change in
the size of automobiles people purchase.
The complexity of behavioral patterns which can be attributed, in part, to the expanded choice set is grow-
ing too. As a result, it becomes somewhat less amenable for analysis, but more robust in terms of the theo-
retical basis. The conceptual model lying at the basis of this study calls for testing a number of choice
situations. These include multiple elements of shopping behavior: the structure of the shopping process,
and the consequent decisions on choice of destination, frequency and timing, mode of travel, trip chaining,
alternative delivery arrangements and more.
One of the implications of the complex activity pattern is the need for detailed, tailored-made data collec-
tion efforts. Our present data base will enable analysis and model estimation for some of the above-mentioned
questions.
At present, it seems that despite the rapid growth and widespread publicity, teleshopping is still in its
infancy. A growing number of suppliers, along with increasing household access to computer networks, facil-
itate the popularization of E-retail. But, it is difficult to forecast the impacts of E-retail on travel. One reason
lies in the fact that suppliers and much more so, consumers, are still experimenting with E-retail. Service qual-
ity, security of transaction and trust between the consumers and suppliers must be built. Second, access to the
necessary technology, both hardware and software, is improving quite rapidly. Generally, usage is becoming
more user-friendly, and the learning time on how to search and purchase is decreasing. All of these, combined,
imply that the dynamics of change seem to be very rapid, making forecasting more complicated. When the
environment is characterized by such dynamics it may become more important to develop sound theory-
based, behavioral models that can reveal the underlying factors which encourage or hamper the use of new
technology. By adopting a detailed behavioral model, it is suggested that an understanding of both retail
and E-retail can be fostered.
References
Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., Wood, S., 1997. Interactive home shopping: consumer, retailer, and
manufacturer incentives to participate in electronic marketplaces. Journal of Marketing 61 (3), 38–53.
Arnfalk, P. 1999. Information technology in pollution prevention: teleconferencing and telework used as tools in the reduction of work-
related travel. Licentiate dissertation, Lund University, Sweden, The International Institute for Industrial Environmental Economics.
Bell, R., Davis, R., Howard, E., 1997. The changing structure of food retailing in Europe: the implications for strategy. Long Range
Planning 30 (6), 853–861.
Biyalogorsky, E., Naik, P., 2003. Clicks and mortar: the effect of on-line activities on off-line sales. Marketing Letters 14 (1), 21–32.
Chiang, K.P., Dholakia, R.R., 2003. Factors driving consumer intention to shop online: an empirical investigation. Journal of Consumer
Psychology 13 (1–2), 177–183.
Court, D., Dayal, S., 2002. Beyond behavioral bounds. Marketing Management 11 (5), 28–32.
Dixon, T., Marston, A., 2002. UK retail real estate and the effects of online shopping. Journal of Urban Technology 9 (3), 19–47.
Forsythe, S.M., Shi, B., 2003. Consumer patronage and risk perceptions in Internet shopping. Journal of Business Research 56 (11), 867–
875.
Golob, T.F., Regan, A.C., 2001. Impacts of information technology on personal travel and commercial vehicle operations: research
challenges and opportunities. Transportation Research C 9 (2), 87–121.
Gould, J., Golob, T.F., 2002. Consumer E-commerce, virtual accessibility, and sustainable transport. In: Black, W.R., Nijkamp, P. (Eds.),
Social Change and Sustainable Transport. Indiana University Press, Indiana, USA, pp. 279–285.
Grabner-Kraeuter, S., 2002. The role of consumers’ trust in online-shopping. Journal of Business Ethics 39 (1–2), 43–50.
Hsieh, Y.C., Chiu, H.C., Chiang, M.Y., 2005. Maintaining a committed online customer: a study across search-experience-credence
products. Journal of Retailing 81 (1), 75–82.
188
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189

Page 14

Israel CBS, 2004. Statistical abstract of Israel, No. 51.
Jones, K., Simmons, J., 1990. The Retail Environment. Routledge Publication, London.
Klein, L.R., 1998. Evaluating the potential of interactive media through new lens: search versus experience goods. Journal of Business
Research 41 (3), 195–203.
Kolesar, M.B., Galbraith, R.W., 2000. A services-marketing perspective on E-retailing: implications for E-retailers and directions for
further research. Internet Research-Electronic Networking 10 (5), 424–438.
Koppelman, F., Salomon, I., Proussaloglou, K., 1991. Teleshopping or store shopping? A choice model for forecasting the use of new
telecommunications-based services. Environment and Planning B: Planning and Design 18, 473–489.
Lee, K.S., Tan, S.J., 2003. E-retailing versus physical retailing – a theoretical model and empirical test of consumer choice. Journal of
Business Research 56 (11), 877–885.
Lenz, B., 2003. Will electronic commerce help to reduce traffic in agglomeration areas? TRB – 82nd Annual Meeting, 12–16 January 2003,
Washington. TRB – 82nd Annual Meeting, Washington.
Mokhtarian, P.L., 2001. The Impacts of B2C E-Commerce on Transportation and Urban Form. Prepared for the Conference on
Electronic Commerce (B2C) and its Consequences for Urban Development and Transportation. University of Stuttgart.
Mokhtarian, P.L., 2004. A conceptual analysis of the transportation impacts of B2C E-commerce. Transportation 31 (3), 257–284.
Mokhtarian, P.L., Salomon, I., 2001. How derived is the demand for travel? Some conceptual and measurement considerations.
Transportation Research A 35, 695–719.
Mokhtarian, P.L., Salomon, I., 2002. Emerging travel patterns: Do telecommunications make a difference? In: Mahmassani, H.S. (Ed.), In
perpetual motion: travel behavior research opportunities and application challenges. Pergamon, Amsterdam, pp. 143–182.
Monheim, R., 1992. Town and transport planning and the development of retail trade in metropolitan-areas of West-Germany. Landscape
and Urban Planning 22 (2–4), 121–136.
Mulhern, F.J., 1997. Retail marketing: from distribution to integration. International Journal of Research in Marketing 14 (2), 103–124.
Rao, B., 1999. The Internet and the revolution in distribution: a cross- industry examination. Technology in Society 21 (3), 287–306.
Ratchford, B.T., Talukdar, D., Lee, M.S., 2001. A model of consumer choice of the Internet as an information source. International
Journal of Electronic Commerce 5 (3), 7–21.
Rosen, K.T., Howard, A.L., 2000. E-retail: gold rush or fool’s gold? California Management Review 42 (3), 72–100.
Salomon, I., Koppelman, F.S., 1992. Teleshopping or going shopping? An information acquisition perspective. Behaviour and
Information Technology 11 (4), 189–198.
Teo, T.S.H., 2002. Attitudes toward online shopping and the Internet. Behaviour and Information Technology 21 (4), 259–271.
The Economist, 2005. Meg and Power of Many. June, 11th.
O. Rotem-Mindali, I. Salomon / Transportation Research Part A 41 (2007) 176–189
189