The effects of in-store themed events on consumer store choice decisions

Abstract

Retailers are increasingly using in-store events to provide shoppers with unique experiences that will enhance shopping value and help differentiate their stores from competitors. However, relatively little is known about how consumers respond to experiential retail events in terms of their store choice decisions. The purpose of this research was to find out how in-store retail events affect consumers in terms of their store choice decisions. The paper presents findings from a discrete choice experiment that manipulated the presence of different types of in-store themed events in a do-it-yourself (DIY) category. Participants were 312 randomly recruited residents of Melbourne, Australia, who had recently shopped at a hardware store. The experiment was implemented as a mail-back survey. Using logit models the authors assess the effects of the in-store events along with those of various traditional store attributes, including store appearance, price image and distance, on consumer store choice decisions.

Full text

The effects of in-store themed events on consumer store choice decisions
Final manuscript;
published in 2009 in the Journal of Retailing and Consumer Services, 16, 386-395
Doi: 10.1016/j.jretconser.2009.05.001
Dr Sean Sands*1
Research Fellow
The Australian Centre for Retail Studies
Monash University
PO Box 197
Caulfield East, VIC 3145
Australia
Phone: +61 (0)3 9903 2753
Fax: +61 (0)3 9903 1558
Email: sean.sands@buseco.monash.edu.au
Professor Harmen Oppewal
Department of Marketing
Monash University
PO Box 197
Caulfield East, VIC 3145
Australia
Phone: +61 (0)3 9903 2360
Fax: +61 (0)3 9903 1558
Email: harmen.oppewal@buseco.monash.edu.au
Professor Michael Beverland
Department of Economics, Finance and Marketing
RMIT University
Building 108, Level 12
Melbourne, VIC 3000
Australia
Phone: +61 (0)3 9925 1475
Email: michael.beverland@rmit.edu.au
*) Acknowledgement – see final page
1 Corresponding author

Abstract
Retailers are increasingly using in-store events to provide shoppers with unique experiences that
will enhance shopping value and help differentiate their stores from competitors. However,
relatively little is known about how consumers respond to experiential retail events in terms of
their store choice decisions. The purpose of this research was to find out how in-store retail
events affect consumers in terms of their store choice decisions. The paper presents findings
from a discrete choice experiment which manipulated the presence of different types of in-store
themed events in a DIY category. Participants were 312 randomly recruited residents of
Melbourne, Australia, who had recently shopped at a hardware store. The experiment was
implemented as a mailback survey. Using logit models the authors assess the effects of the in-
store events along with those of various traditional store attributes, including store appearance,
price image and distance, on consumer store choice decisions.
Keywords: retail atmospherics, themed events, store choice, store format

1
The effects of in-store themed events on consumer store choice decisions
1. Introduction
Retailers are increasingly attempting to create in-store environments that not only provide a
pleasant ambience but that also provide uniquely enriching experiences to the customer (Babin
and Attaway, 2000; Baker et al., 2002; Kozinets et al., 2002; Pine and Gilmore, 1999). Such
unique customer experiences are not solely dependent on the individual atmospheric components
of the store. They depend also on themed events that retailers provide to encourage and include
the engagement of the consumer in a variety of ways, from the ludic and playful to the
educational. The atmospheric components and event themes help to prime the consumer and
provide cues as to how customers can engage with the retail product (Kozinets et al., 2002).
The creation of unique environments within the retail setting occurs across a variety of store
settings, including: manufacturer brands (i.e. Niketown, Apple, Nokia –T5), department stores
(i.e. Harrods, Dover St Market), food retail (i.e. Whole Food Market), general retail (i.e. Toys R
Us), and fashion retail (i.e. Louis Vuitton, Prada) to name a few. Peñaloza (1998) notes that at
retailers such as Niketown shoppers are presented with opportunities to physically engage in
sports of their choosing. In this setting the physical exercise and sporting environment allow the
customer to become engaged in the shopping activity. Computer retailers such as Apple also
have embraced the concept and host special events in order to educate the consumer (Baron et
al., 2000). The general premise is that such retail experiential events result in benefits to retailers
(Pine and Gilmore, 1999; Schmitt, 1999), such as increased sales and positive word of mouth.
However, only limited research has been conducted that tested these assertions.

2
This paper aims to assess how the presence of an in-store themed event affects consumer store
choice. An in-store themed event occurs when a retailer provides a recognizable activity within
the store to inform or engage the consumer. As Sit and Merrilees (2005) note, these events are
often discrete and temporary. To distinguish between different types of events we draw on Pine
and Gilmore’s (1999) four themes of aesthetics, education, entertainment, and escapism We
investigate how in-store experience-enhancing retail events impact consumer store choice
decisions using a discrete choice experimental approach (Louviere et al., 2000). We also
separately test the effect of the store’s ambience, or appearance. In choice experiments
participants are exposed to realistic but experimentally controlled shopping task conditions. We
develop a shopping scenario to test how store choice is affected by the store’s ambience and the
presence of a special event, in addition to the effects of various traditional store attributes. The
attribute effects are tested for two specific retail formats, a specialist and a general (big box)
store. Differentiating between these two formats is important as they represent very different
shopping environments within which similar environmental cues may have different influences
on consumer behaviour (Hansen, 2003).
To date there has been limited research investigating the impact of in-store retail experiences (or
events) on consumer behaviour. Both Parsons (2003) and Sit et al. (2003) have investigated the
effect of entertainment-based events on consumers, however this was in the shopping centre
context and there have been no studies that empirically investigate the impact of these events on
consumer store choice decisions. In addition, no published works have looked at the differences
between the specialist and large format stores. We therefore investigated if, and how the effects
of the store appearance and type of in-store event on store choice differ between these two

3
formats. While we find no effects for the general store our findings suggest that specialist
retailers may be able to attract customers by organizing in-store themed events. Conversely, store
appearance had only an effect for the general store, with a store appearance that is deemed
unattractive having a negative impact compared to a regular or ‘normal’ appearance. A store
presented as explicitly ‘attractive’ however did not draw more customers than a store with a
‘normal’ appearance.
The context for this study is paint retailing. Paint is a type of do-it-yourself (DIY) retailing which
provides a unique opportunity for the investigation of retail experience. In the housing market,
DIY relates to any home improvement activity that is undertaken by the home owner or occupant
rather than employing professional help (Williams, 2004). These activities may include building,
renovating, and painting to name a few. Whilst the frequency of shopping for paint is less than
that of consumer goods, paint retailing was deemed relevant as a context given recent trends
revealing the increasing role fashion has in paint retailing and the recent growth in DIY as an
activity. This is evident also from the fact that our industry partner in this project operates in this
sector and was seriously interested in exploring the possibilities of using events to create and
enhance retail experience. Fashion has become prominent in paint retailing, with new ranges
being combined with fashion mega-brands (Pepall and Richards, 2002) including Martha
Stewart, Ralph Lauren, and New Zealand fashion designer Karen Walker. Recent years have
seen a boom in DIY activities, with the growth attributed to the buoyant housing market
combined with the rise in television home makeover and property development shows (Watson
and Shove, 2005). Although the traditional assumption in retail studies is that DIY is a rational
response to an inability to pay for external labor, this perspective is not consistent with insights

4
regarding consumer decisions to engage in DIY activities (Williams, 2004), which include that
DIY can be a means to realizing effects which convey individuality and self-identity
(Woodward, 2003). Paint retailing henceforth is a suitable context for studying effects of in-store
themed events.
To achieve its aims the paper will first present an overview of the relevant shopping literature,
leading to formulating the hypotheses that are tested in our empirical study. This is followed by
the methodology and findings sections. A conclusion and discussion section ends the paper.
2. Literature review
2.1. Shopping value and store choice
In deciding where to shop consumers are said to trade-off the value offered by alternative
destinations. Shopping value is a multi-dimensional construct and is often described as
comprising utilitarian and hedonic value (Babin et al. 1994; Childers et al. 2001). Within a retail
context, utilitarian shopping value can be related to the consumer’s need to obtain some
utilitarian of functional consequences from visiting a store, for example in terms of competitive
price (Dodds, Monroe and Grewal 1991), time and effort expenditure reduction (Baker et al.
2002), or risk reduction (Chen and Dubinsky 2003; Sweeney, Soutar and Johnson 1999; Teas
and Agarwal 2000). There are a variety of utilitarian attributes which are salient in consumer
discrimination between retail locations, these include: price level, range of goods, distance from
home to store, customer service levels, and accessibility (i.e. Bell 1999; Oppewal and
Koelemeijer 2005). A store’s appearance, for instance its atmosphere, has also been found to
affect store choice (Gilmore, Margulis and Rauch 2001; Kaltcheva and Weitz 2006; Koelemeijer

5
and Oppewal 1999; Mattila and Wirtz, 2001). Similarly, various design and atmospheric
attributes have been found to affect shopping centre attractiveness (Oppewal and Timmermans
1999). Store appearance however can be regarded more a hedonic than a utilitarian attribute.
Hedonic shopping value can be related to the consumer’s need to obtain fun and pleasure and
relates to the perceived level of shopping enjoyment. Beatty and Ferrell (1998) define shopping
enjoyment as the pleasure obtained from the shopping process, which often transcends product
purchase (Alba et al. 1997; Babin et al. 1994). The hedonic attributes that retail researchers have
investigated have typically related to atmospheric variables, and are defined by Turley and
Milliman (2000) as including: the store exterior, store interior, store layout and design, the point-
of-purchase and decoration variables, and human variables. Turley and Milliman (2000)
conducted a comprehensive review of the effects of atmospherics on shopping behaviour. The
authors show that since Kotler (1973-1974) coined the term atmospherics a range of research has
investigated the impact of visual, aural, olfactory, and tactile dimensions of the retail
environment on a variety of behavioural factors. These studies have investigated the impact of
highly arousing environments, considering the effects of atmospheric variables either in isolation
(i.e. the effect of music only) or in some combination (i.e. the effects of music and lighting).
There have been mixed results in terms of how a retail environment might impact consumer
behaviour. Research findings have reported positive effects (Sherman, Mathur and Smith, 1997)
and negative effects (Milliman, 1982; Smith and Curnow, 1966). In particular, research has
shown that the impact of arousal on shopping behaviour has had mixed results (Donovan and
Rossiter, 1982; Donovan et al. 1994; Kaltcheva and Weitz, 2006). Kaltcheva and Weitz (2006)

6
demonstrated that highly arousing environments would attract some consumers, while others
might avoid them. In contrast, pleasant store environments typically do have positive effects on
shopping behaviours such as approach behaviours, unplanned spending, and duration of visit
(Donovan and Rossiter, 1982; Baker, Levy and Grewal, 1992; Kaltcheva and Weitz, 2006).
A more recent trend in this stream of research is the increased focus on the experiential aspects
of consumption (Holbrook 1999; Vargo and Lusch 2004) and the notion of experiential retailing.
The concept of retail experience is based on the notion that a consumer’s experience is not solely
about the individual atmospheric components that can be found within the store. Rather,
experience incorporates and extends atmospheric variables to encourage and include the
engagement of the consumer in a variety of ways, from the ludic and playful to the educational,
within the store. The components within the environment help to prime the consumer and
provide cues as to how customers can engage with the retail product (Kozinets et al. 2002).
Some retailers have invested heavily in this area, including the Samsung concept store in New
York City, Recreation Equipment Inc. (REI), Niketown, and Apple to name a few. In their
concept store, Samsung have implemented a permanent display store, in fact nothing is sold with
the servicescape, rather experience is created in an escapist environment where consumers can
engage and interact with products in order to see how products can be used in everyday life
(Barbaro 2007). Peñaloza (1998) notes that at sporting retailer Niketown, shoppers are presented
with opportunities to physically engage in sports of their choosing. In this setting the physical
exercise and sporting environment allow the customer to become engaged in the shopping

7
activity. Furthermore, at Apple stores, special events are hosted in order to educate the consumer
and engage them with the product (Baron, Harris and Harris 2001).
The general premise underlying these investments is that such retail experiences result in
benefits to retailers (Pine and Gilmore 1999; Schmitt 1999). However, there is only limited
empirical evidence supporting these assertions. In particular there is no empirical evidence to
suggest what the relevance or importance of retail experience is for store choice compared to
other variables, either utilitarian (i.e. access, location) or hedonic (i.e. atmospheric). In fact, a
review of the academic literature reveals that there are no studies that investigate the effect of
hosting short or long term experiential events within the retail setting.
2.2. Retail experience and in-store events
To conceptualize the possible effects of in-store experiential events we adopt the framework
developed by Pine and Gilmore. Pine and Gilmore (1999) define four potential realms of retail
experience, identifying the aesthetic, educational, entertaining, and escapist as potential event
types. They operationalize their four realms of retail experience in terms of two key dimensions.
The first dimension is customer participation, which ranges from passive (weak) to active
(strong). This dimension concerns the level of physical participation of the individual. Weak
participation means the individual cannot influence the execution of the experience. In contrast,
strong participation means that the individual can act to influence the execution of the
experience. The second dimension refers to the individual’s connection with the experience,
ranging from absorption (weak) to immersion (strong). For example, attending a live sports event
is more immersive than watching the same event on television.

8
The two dimensions span four sectors defining the four experiences. During aesthetic
experiences individuals are immersed in an environment but have little or no effect on the
environment itself. An educational experience involves active participation, for instance the
consumer can engage with the environment or products to learn but typically involves more
absorption than immersion. Entertaining experiences are similarly absorptive but are more
passive than educational experiences. Finally, the escapist experience involves greater immersion
than the entertaining or educational experiences, and involves more active participation than the
esthetic experience. Examples of environments where guests encounter escapist experiences
include theme parks, casinos, virtual reality, and even paint ball. In these experiences consumers
participate in the thrill of moment (Pine and Gilmore, 1999, 33).
Based on the literature discussed here, we predict that in-store events can potentially impact
consumers in a similar way to the hedonic aspects of the retail environment discussed above (i.e.
atmospheric variables). In-store themed events can add value to the consumer’s visit to the store
by offering a suitable entertainment or educational experience, or by offering an aesthetically
pleasing experience and/or by offering an opportunity to ‘escape’. In either case, we may expect
that consumers will be more likely to visit a store if they anticipate such in-store events to be
present. We therefore hypothesize that:
H1: The presence of an in-store retail event has a positive effect on store choice, such that
the presence of an in-store event will influence shoppers to visit the store.

9
Furthermore, research has found that the importance of retail store attributes can vary depending
on the format of the store. In a supermarket setting Lal and Rao (1997) show that stores offering
everyday-low-pricing appeal to shoppers who are time deprived, offering an attractive price
proposition for one-stop shopping. In contrast, shoppers at specialty food stores tend to value
product quality over price. Further, Hansen (2003) found that although product quality and
freshness of products were ranked the most important by specialty food store shoppers, low
prices was ranked 17 among 29 factors. Furthermore, the advantage that specialty stores have
traditionally held over large format stores, in terms of service levels, and their authentic,
interactive, and personalized ambience, is being eroded. Many large retailers are increasingly
adjusting their look and feel to better service the customer and replicate a specialty store
shopping experience (Spector, 2005). In this context, the in-store environment provides a
potentially important means of differentiation and competitive advantage for specialty retailers
(Pine and Gilmore, 1999).
We propose that the presence of an in-store themed event is an additional attribute in the retail
setting that may impact consumer store choice. Given that the importance of retail store
attributes has been shown to vary depending on the format of the store (i.e. Lal and Rao, 1997)
and that Pine and Gilmore (1999) have suggested that these types of experiential events may be
particularly beneficial to specialist retailers, we expect that there will be significant differences in
consumer evaluations of the themes event depending on the store format (specialist versus big
box) the event is presented in. Shoppers attach greater value to hedonic aspects of the in-store
experience, such as offered by in-store themed events and atmospheric features when

10
considering a specialist store than when considering a big-box store. Consequently we
hypothesize that:
H2: The presence of an in-store retail event will have a greater effect for specialist store
choice than for choice of a general (big box) store.
We expect that the context of the study (DIY retailing) will impact preference for the type of
event theme presented in-store. As previously noted, DIY relates to any home improvement
activity that is undertaken by the home owner, rather than employing professional help, and has
become increasingly common in recent years as a means of self-expression (Williams, 2004). In
the housing market, these activities may include building, renovating, and painting to name a
few. Recent years have seen a boom in DIY activities, with the growth attributed to the buoyant
housing market combined with the rise in television home makeover and property development
shows (Watson and Shove, 2005). Within the DIY context, we hypothesize the presence of an
educational in-store event theme will have a greater effect on store choice, compared to the other
in-store event types, hence:
H3: An education-themed in-store event will have a significantly higher impact on
consumer store choice within the DIY context than the other in-store event types.
3. Method
3.1 Sample
The context for this study was the DIY category. Telephone recruitment was used to approach a
random sample of households in Melbourne, Australia, for participation in a mail-back survey
about shopping for DIY products. Interviewers were instructed to recruit participants who had

11
visited a hardware store within the past 6 months. Of 488 questionnaires distributed, a total of
312 (64%) were returned. The sample consisted of 59% females, and a variety of age groups
were represented with the highest percentage of subjects being between 35 and 44 years of age
(22%). Overall, the majority of the sample was between the ages of 35 and 64 (62.9%).
3.2 Stated preference choice modeling
It is difficult to observe the types of experiential events defined by Pine and Gilmore’s (1999)
framework in situ. For this reason we adopt the use of stated preference choice modeling
methods. Stated preference choice models have been used in areas as diverse as marketing,
transportation, environmental economics, and health economics (Louviere et al., 2000), and have
been used numerous times to model choice of shopping destination (Bucklin and Lattin, 1992;
Oppewal et al., 1997; Grover and Srinivasan, 1992). Furthermore, these models have been used
in decisions as diverse as consumer demand for travel demand (Ben-Akiva and Lerman, 1985)
and unique cultural events (Louviere and Hensher, 1983).
The collection of stated preference data and discrete choice models derived from random utility
theory, have become the standard technique used to model individual choice behaviour (Hensher
et al., 2005; Louviere et al., 2000). Hensher et al. (2005) discuss the fact that models of
individual choice have their foundations in classic economic consumer theory, which is the
source of many of the assumptions within the models. Economic consumer theory states that
consumers are rational decision makers that, when faced with a set of possible consumption
alternatives, assign preferences to each of the various bundles and then choose the most preferred

12
from the set of affordable alternatives. Individuals are said to select the consumption bundle that
maximizes their utility, subject to their budget constraint.
The most common discrete choice model is the linear in parameters, utility maximizing,
multinomial logit model (MNL). During the choice process consumers attach a certain structural
utility Vi to each alternative i; this utility is a function of all attributes, and is measured with an
error εi. Each decision maker will derive an amount of utility for each of the alternatives, within
the universal (but finite) set of alternatives. The MNL model was derived by McFadden (1974)
and is specified as:
Uin = Xinβ + Vin, Vin are i.i.d Gumbel random variates with scale parameter µ (1)
yin = (2)
Equations 1 and 2 lead to the following individual choice probability:
P(yin = 1|Xn; β) = (3)
where: Cn is the choice set faced by individual n, comprising Jn alternatives; Xn is a vector
of predictor values and β is a (K x 1) vector of unknown parameters.
The probability that a consumer selects alternative i, equals the probability that Ui is larger than
the utilities Uj of all other alternatives j in the consumer’s choice set J (1,…,J). As demonstrated
in equation 2, the model is based on the notion that an individual chooses the alternative that is
perceived as having the highest utility. The model allows predicting the probabilities of an
1, if Uin = max Ujn
J
0, otherwise
exp µ (Xin β)
Σ exp µ (Xjn β)
j∈Cn

13
individual selecting each alternative, which can then be aggregated to produce forecasts for the
sample or population. The MNL model assumes that the error components in the utility function
(εi) are independently and identically distributed (IID) according to an Extreme Value or Gumbel
(0, µ) distribution, where µ is a scale parameter (see Louviere et al., 2000).
To assess how the presence of an in-store themed event can influence store choice decisions we
designed a discrete choice experiment as a section within a mail-back survey. Discrete choice
models work under the assumption that individuals assign utility to alternatives according to
their utility function, and this function (or preference structure) is revealed by the choice of the
most preferred alternative (Louviere et al., 2000). In this study we collect preference data based
on the measurement of stated preferences between hypothetical choice alternatives, based on the
event types categorized by Pine and Gilmore (1999).
Stated preference data possess a number of advantages over revealed preference data, which are
particularly useful in the development and calibration of forecasting models of store choice and
have been used extensively in business research, particularly in marketing and transportation
(Louviere et al., 2000; Louviere and Woodworth, 1983). It is particularly advantageous in that
the alternatives included in the hypothetical choice set are independent of the choice context, and
may therefore be designed to cover the range of choice alternatives to be modeled. Thus, it is
possible to include store alternatives, features, or unique events that may not exist in the current
market (Louviere and Hensher, 1983), for instance the offering of in-store events as part of a
retail marketing mix.
3.3. Event description and manipulation

14
Respondents were first asked to imagine a hypothetical trip to the DIY store they would be most
likely to visit today if they had to select a particular DIY product. They were asked to imagine
that when they reached the store they found out that the store had an in-store event running. It
was explained that some retail stores attempt to create deliberate and engaging events or
experiences for consumers within the store.
Respondents were then presented with a verbal description of the event. Respondents were
randomly allocated to one of four theme conditions. The theme scenario conditions varied the
theme of the in-store event based on the four Pine and Gilmore (1999) dimensions (aesthetic,
education, entertainment, escapist). For example in the educational condition respondents were
presented with an event description as follows:
When you reach the store you find out that the store has an in-store event running. This
event is based on the concept of education. Within this store you are presented with the
ability to seek professional advice by way of in-store presentations and/ or one-on-one
interaction with specialists.
The remaining scenarios were described in a similar manner and are presented in the Appendix.
The in-store event descriptions were later referred to when presenting the in-store experience
attribute in the discrete choice experiment.
Following the description of the in-store themed events, respondents were asked to write one
thing that they liked and one that they disliked about the event described to them. In the analysis
stage two coders independently coded these responses using a categorization scheme and
independently coded all participants’ responses. Coder agreement was calculated as the

15
percentage of responses that both coders independently classified into the same category. The
agreement rate across the four event types was 51%; hence the open-ended responses indicate
that the manipulation of the events was comprehended in the intended manner.
3.4 The choice task
The event description and manipulation task was followed by the actual choice experiment. In
the experiment the presented in-store event was referred to using a separate attribute with three
levels, ‘in-store today only’, ‘in-store for one month’, and ‘no in-store event’. The remaining
attributes in the choice experiment were identified through the literature and validated through
preliminary focus groups.
Two focus groups were conducted to determine the importance of the attributes identified in the
literature for our category, examine respondent responses to the in-store experience attribute
descriptions and assess respondent perception of the discrete choice modeling approach whilst
refining the instrument and instructions. During these groups, respondents discussed a range of
store attribute that might be salient in their retail paint store choice decisions. They were also
presented with an earlier version of the survey and asked to comment on a range of features,
from the layout to comprehension of the task. The focus group results confirmed the relevance of
the five key attributes from the literature - price level (discount), range of goods, distance from
home to store, customer service (customer service and staff friendliness), and accessibility (ease
of access).
[TABLE 1 ABOUT HERE]

16
The choice task asked respondents to imagine that they needed to visit a DIY store today. They
were then presented with two hypothetical store options, along with a delay or neither option and
asked which of the two stores they would visit, or neither. The inclusion of a neither option
allows for the creation of real life situations where the available alternatives may not suit the
needs of the shopper. So, respondents first received the event manipulation descriptions (as
detailed above) and were then presented with the multi-attribute choice task which had the event
as one of the listed attributes. On a separate piece of paper respondents received a detailed
description of each of the attributes and their respective levels. The choice task provided two
store options, which always comprised one specialist and one general hardware store, and a delay
or neither option. An example choice task is shown in figure 1.
[FIGURE 1 ABOUT HERE]
There were eight three-level and one two-level attributes defined, as shown in table 1. An
orthogonal main-effects fractional factorial design was employed to generate an economical
number of choice profiles from the full factorial design. This resulted in 27 alternatives being
generated for each of the two alternatives. This design allowed for the estimation of alternative
specific attribute effects for the two store formats. The presentation order of the alternatives was
rotated so that the specialist store was presented as the left hand alternative in the choice set for
half of the sets. The design was blocked into three subsets of nine choice sets to be presented to
respondents to limit the task load. This choice set design was nested under the four-level

17
between-subjects factor that manipulated the type of event explained to the respondent in the
earlier part of the task.
4. Data analysis and results
Logit analysis was conducted to investigate the importance of the in-store event attribute on store
choice decisions and compare the impact to other, more traditional attributes. We ran three
models for the analysis. The first MNL model tests our hypothesis that the presence of an in-
store themed event has a positive effect on store choice (H1). The model estimates the effect of
the presence and duration of the in-store event, regardless of the type of event or the store
format. The second model investigates the differential impact of the in-store event across the two
store formats and tests our hypothesis that the presence of the in-store themed event will be a
greater determinant of specialist store choice than of choice of general (big box) store (H2).
Finally, the third model investigates the event type interaction effects in order to determine if
there are specific effects relative to the type of event that respondents are presented with.
Specifically it tests our hypothesis that an education-themed in-store event will have a
significantly higher impact on consumer store choice within the DIY context than the other in-
store event types (H3).
4.1 Presence of an in-store event
Our first hypothesis was tested by estimating a generic model that constrains attribute effects to
be equal across the two store formats. Table 2 shows the estimation results for this model (model
I).

18
[TABLE 2 ABOUT HERE]
Effect coding was used for all attributes. With a value of 0.20 for McFadden’s RhoSquare, the
model fit was satisfactory (Louviere et al., 2000). The log likelihood for the base (constants only)
model was compared to the model presented in Table 2 with a Likelihood ratio test. The latter
proved a significant improvement from the base model (ChiSq = 860.61, d.f .= 16, p≤ 0.001).
Inspection of the estimated model coefficients reveals that the majority of the variables are
statistically significant at the five percent level of significance. Initial inspection of the attributes
confirms that most coefficients have signs and relative sizes as expected. For example, the
relationship between level of discount and store choice utility, for the range of discount levels
included in the survey, shows that higher discount levels result in higher levels of utility – albeit
not a perfectly linear relationship.
The attribute level coefficient for the in-store event (level = one month) is statistically significant
(β =.096, p ≤ .05) for model I. Hence, we find support for the hypothesis (H1) that the presence of
an in-store retail event has a positive effect on store choice, such that the presence of an in-store
event will influence shoppers to visit the store.
4.2 Differences between specialist and general (big box) hardware store
A second model was run to test Hypothesis 2. In this model the attribute level coefficients are
specified as being specific to each alternative (specialist store versus general store), in order to
allow for the estimation of effects for each store type. Table 3 presents the estimation results of
this model.

19
[TABLE 3 ABOUT HERE]
With a value of 0.20 for McFadden’s RhoSquare, the model fit was again satisfactory.
Likelihood Ratio tests revealed that the parameter vectors for the specialist store and the general
hardware store were significantly different, even when allowing for different alternative specific
constants (LR Chi-Sq = 2705.13, d.f.=16, p≤.001). The alternative specific attributes will now be
further discussed separately in relation to the two store formats, before a more detailed
discussion of the in-store event attribute.
4.2.1 Specialist store format
For the specialist store there are several significant effects worth noting. The coefficients
estimated for the customer service attribute are relatively large and highly significant for the high
level of customer service. An improvement from a low customer service level to a medium level
of customer service results in no significant increase in utility, however an increase from low to
high results in a significant increase in utility (β =.482, p≤ .001). The coefficients of the distance
from home attribute have the expected signs. Specialist stores that are closer to the respondent’s
home, and those having a shorter travel time, are preferred to stores that are further away. The
difference in utility between a store that is five minutes and another store that is 15 minutes drive
from home is non-significant however the difference in utility for a store that is five minutes
versus one that is 25 minutes is significant. The difference in the level of access to the store
between easy and medium, and medium and difficult is reflected in a significant decrease in
utility.

20
An increase in staff friendliness from low to high results in a large increase in utility associated
with the specialist (β =.371, p≤ .001), whereas there is no increase in utility garnered from an
increase in staff friendliness from low to medium. As mentioned above, the relationship between
level of discount and store choice utility shows a positive relationship. An increase in the level of
price discount is associated with an increase in overall utility. However this relationship is not
linear, an increase from no discount to a 2.5% discount results in a large increase in utility,
whereas a smaller increase in utility is garnered from an increase from 2.5% to 5% discount – as
would be perceived due to the relative small increase in the discounted percentage. The signs of
the estimated coefficients pertaining to the levels of the attribute product range are as perceived.
The difference in range from small to large is reflected in a large increase in utility – however
there is a non-significant change from a low product range to a medium range. Finally, the store
appearance coefficients are relatively small and not significant.
4.2.2 General (large format) store format
There were similar patterns in utility for the general hardware store format compared to the
specialist store format. In particular, the attributes for distance, access, staff friendliness,
discount level, and product range had similar coefficient signs and significance levels. There
were similar effects for the general hardware store format on most of the attributes. However, a
change from low to medium customer service levels was significant for the general format (β
=.241, p≤ .05). This would suggest that offering at least some form of customer service level
already makes a difference for shoppers at the general format, whereas a specialist store format
has to provide higher levels of customer service (with only the high level attribute being

21
significant) to meet customers’ service expectations. There were also differences in terms of the
store appearance attribute and the in-store event attribute.
4.2.3 Store appearance
Whilst the store appearance attribute effects were non-significant for the specialist store format,
the effect of changing from an ‘unattractive’ to a ‘normal’ appearance was significant for the
general hardware store format (β =.210, p≤ .05). Remarkably, further changing a general
hardware store’s appearance to fit the description ‘attractive’ tended to reduce the store’s choice
utility. These results suggest that whereas for the specialist store appearance did not seem to
matter, for the general hardware store respondents in our sample preferred a moderately
attractive appearance
4.2.4 The in-store event
The in-store event attribute had a significant and positive effect for the specialist store format (β
= .15, p ≤ .05) but only when it is present for one-month. The effect for the event when it was
only on for today tends to even be negative compared to having no event at all - but note the
differences between ‘one day’ and ‘no event’ was not statistically significant. This lack of
positive effect for the one day event suggests that respondents did not like or even dislike an
event they might not be able to make use of. The in-store event attribute was not significant for
the general hardware store format. For our respondents the presence of an event was no different,
or did not impact utility differently than having no event in-store. Overall these results suggest
that the presence of an in-store event did have an impact on respondents’ derived utility levels
for store choice decisions – however this effect was specific to the specialist store. These results

22
support our hypothesis (H2) that the in-store themed event is a greater determinant of specialist
store choice than of choice of a general (big box) store.
4.3 Differences between event types
In order to investigate the individual event type effects specified in Hypothesis 3 we estimated a
third MNL model that included the interaction of the in-store event attributes and the event types
(i.e. aesthetic, education, entertainment, and escapist) detailed in the scenarios. Interaction terms
were created for each event type (EX1, EX2, and EX3) compared to a base level – arbitrarily
chosen as the escapist event. Further, interaction terms were created to contrasts the event
interactions with the event attribute levels (i.e. today versus one month). The third model was
estimated in a similar manner to model two, the only addition was the inclusion of the event type
interaction terms. The results of this model are presented in table 4.
[TABLE 4 ABOUT HERE]
Analysis of the event interaction terms, contrasted with the base levels, shows no significant
differences overall between the event types, as the extension of the model with event type
interaction terms did not significantly improve the model fit (LR = 22.16, d.f.= 22, n.s.). Detailed
inspection however shows two particular interactions to be significant, both for the general
hardware store, which format had not shown event effects in the previous analyses. Although
such effects should be interpreted with caution, the negative parameters do suggest that
entertainment, and also education, can have even negative effects under particular conditions,
such as determined by timing of the event. Despite this, the overall results, with neither of three

23
‘main’ event interactions being significant, indicate that respondents did not reveal any specific
preference for events to be presented within their shopping situations. This finding leads us to
reject our hypothesis (H3) and we conclude that an education-themed in-store event does not
have a significantly higher impact on consumer store choice than the other in-store event types,
in the context selected for this study.
5. Conclusion and discussion
5.1 Summary of findings
With this research we set out to assess the effects of in-store themed events on consumer store
choice. Using a discrete choice experimental approach we investigated how a range of variables,
including store appearance, impact on consumer store choice decisions. Our analyses revealed
that the presence of an in-store event does have an impact on store choice decisions, however
findings from further analysis suggested that this is only the case for specialist stores. Hence, our
results suggest that specialist retailers have more to gain from hosting these discrete and
temporary in-store events. Conversely, store appearance was found to only have an effect for a
general store, with a store appearance that is deemed unattractive having a negative impact
compared to a regular or ‘normal’ appearance. A store presented as explicitly ‘attractive’ did not
draw more customers than a store with a ‘normal’ appearance, suggesting appearance may be a
mere hygiene factor in this case. Furthermore, we did not find any significant effect of the type
of event theme on store choice decisions.
5.2 Contribution and implications

24
This paper contributes to the literature in a number of ways. First and foremost, we identify the
effects of in-store experiential retail events on consumers' store choice, and investigate the
differential effects of these themed events across two distinct retail formats. To our knowledge
this is the first paper to empirically examine the interrelationship of these variables. This
research provides insight into the interrelationship between in-store experience enhancing events
and the important dependent variable of retail store choice. Specifically, our findings lend
support to the notion that retail environment can positively influence consumer store choice.
We also identified how the effects of in-store retail events on store choice vary across store
formats. For consumers shopping at a large format general hardware store, the presence of an in-
store event did not increase the store’s utility as a shopping destination; there were even
indications that the effect may be negative in particular circumstances. This suggests shoppers at
big box hardware stores do not care to engage in experience enhancing retail environments (i.e.
arousing environments). In contrast, for consumers shopping at a specialist store, in-store events
did contribute to the overall attractiveness of the store. In this case the presence of in-store events
increased the utility that shoppers expect to derive from the store (albeit only slightly) and the in-
store event was a significant predictor of these shoppers’ store choice decisions.
While we show that the presence of a themed event can be a predictor in determining some
shoppers’ store choice decisions, our results point to a complex relationship between the
utilitarian and hedonic dimensions of the retail encounter. In particular, we find support for the
notion that the utilitarian aspects may well be necessary but not sufficient dimensions for retailer
success (Jones et al., 2006). The results show that the hedonic dimension of the shopping

25
encounter (i.e. the themed event) can add to consumers’ derived utility, however the presence of
a themed event is not necessarily a retail panacea. Our results also point to the fact that retail
performance on hedonic aspects may not be sufficient if not supported by the necessary
utilitarian structure. In particular, the results revealed that the utilitarian aspects of the shopping
encounter (i.e. discount level, distance, and product range) were generally larger (in derived
utility) than the hedonic aspects (i.e. appearance, themed event).
The results indicate that there are some limitations in terms of the effectiveness of a themed
event on influencing consumer behaviour. First, the data revealed that a themed event does not
contribute toward store choice decisions if the event is available only for a short time period (i.e.
one day only) or if the event is presented in a large format / big box retail store. Second, we
found interaction effects with negative parameters which suggest that entertainment, and also
education, can have negative effects under particular conditions, such as determined by timing of
the event. Together these findings reveal that the implementation of a themed event is not a
silver bullet that will guarantee retailer success. It may be that in some instances in-store events
may impose an undesired required “effort” in order to engage in and negotiate the altered
environment that events may create in-store. Future research could examine what factors drive
these types of outcomes, for instance these effects may be different depending on shoppers’
levels of experience with the category or their motivational orientation (Kaltcheva and Weitz,
2006).
Our results build on previous research that has shown the impact of the retail environment on
consumer behaviour. From this we have three key implications for retailers. First, the creation of

26
a unique customer experience within the store can be used as a tool to encourage store choice.
Whilst academic and practitioner research has alluded to this, there has been limited empirical
research to test such assertions. We have shown that these experiences can be developed through
offering consumers either aesthetic, educational, entertaining, or escapist events within the store.
Second, it seems that such events may not be beneficial to all retailers. We proposed one
distinguishing factor is retail format; however there may be other important aspects, such as the
consumer’s shopping purpose, the shopping context or retail location. There is much work that
needs to be done in this area. Third, the cost of implementing themed events can be prohibitive,
and likely a reason that in retail-reality it are often the manufacturer brands that invest in retail
experience creation (e.g., Apple, Nike). There may be opportunities for retailers to engage
suppliers and manufacturers in the creation of themed store experiences, not unlike the catalogue
industry where the cost of production is typically covered by the manufacturers.
Furthermore, it is important for retail managers to consider issues around the implementation of
themed environments, in particular considerations relating to what these servicescapes may look
like. There are three core areas the retailer can manipulate in order to create a themed
environment (Baker et al., 2002). These three dimensions are: 1) atmospheric factors, which are
background conditions in the environment, such as scent, noise, music, and lighting; 2) social
factors, representing the human component of the environment; and 3) design factors, including
functional and aesthetic elements such as architecture, style, and layout. Together these factors
can be manipulated to form a theme in the servicescape which may aid in the creation of
experience. The manner in which these variables are combined will depend on the theme that the
retailer wants to convey. We have discussed four core themes which can be implemented within

27
the service setting, as developed by Pine and Gilmore (1999). For instance, an entertaining theme
would involve a high level of arousal in store, with the store features tying together to convey an
entertaining servicescape. In contrast, an aesthetic theme might be more subdued, with a lower
level of emphasis on excitement arousal, and more emphasis on stimulating consumers through
sound and sight.
The results from this research should be interpreted relative to certain limitations. Whilst the data
were analyzed across different retail formats, we also recognize that the scope of the research is
limited to the DIY category. How the presence of experiential events affects store choice
decisions in different retail categories is an issue for further investigation. Additional issues
warranting further investigation include how perceptions of shopping value change across event
types and store formats. Additionally, how do shoppers react to the presence of such events when
they actually encounter them in-store? Empirical studies are needed to examine such topics.
Studies that purposively create and manipulate in-store events along the dimension investigated
here would help advance our knowledge of experiential retailing and extend empirical research
results as presented here.

28
Acknowledgments
The authors gratefully acknowledge the financial support of the Australian Research Council
(LP0455193), and industry linkage partner PaintRight Ltd. The authors would like to thank the
staff of PaintRight, in particular former CEO Bruce Munday and Jason Buttigieg, for their
support and assistance throughout the research. We also acknowledge Haymes Paint for their
support at various stages throughout the research.

29
Appendix 1
Aesthetic: Within this paint store you are presented with a visually stunning in-store
environment. The store has gone to great effort to provide a space that is aesthetically pleasing
and inspirational, showcasing a combination of colour, texture, and lighting.
Education: Within this paint store you are presented with the ability to seek professional
decorating advice by way of in-store presentations and/ or one-on-one interaction with
specialists.
Entertainment: Within this paint store you are entertained. Specifically, you are able to watch a
live painting demonstration where a variety of artists are painting a mural on a large wall.
Escapist: Within this paint store you are presented with the ability to engage in a hands-on
manner by painting different textures and using different techniques on a variety of surfaces.

30
References
Alba, J.W., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, S., Wood, S., 1997.
Interactive home shopping: Consumer, retailer, and manufacturer incentives to participate
in electronic marketplaces. Journal of Marketing 61 (Summer), 38-65.
Babin, B.J., Attaway. J.S., 2000. Atmospheric affect as a tool for creating value and gaining
share of customer. Journal of Business Research 49 (2), 91–99.
Babin, B.J., Darden, W.R., Griffin, M., 1994. Work and/or fun: Measuring hedonic and
utilitarian shopping value. Journal of Consumer Research 20, 644-656.
Baker, J.D., Levy, M., Grewal, D., 1992. An experimental approach to making retail store
environmental decisions. Journal of Retailing 68 (4), 445-460.
Baker, J.D., Parasuraman, A., Grewal, D., Voss, G.B., 2002. The influence of multiple store
environment cues on perceived merchandise value and patronage intentions. Journal of
Marketing 66 (2), 120-41.
Barbaro, M., 2007. At the ‘experience store’: A new concept in retailing…loitering is the right
idea. International Herald Tribune accessed November 15, 2007, [available at
http://www.iht.com/articles/2007/03/18/business/shop.php]
Baron, S., Harris, K., Harris, R., 2001. Retail theater: The intended effect of the performance.
Journal of Service Research 4 (2), 102-117.
Beatty, S.E., Ferrell, M.E., 1998. Impulse buying: Modeling its precursors. Journal of Retailing
74 (2), 169-191.
Bell, S.J., 1999. Image and consumer attraction to intraurban retail areas: An environmental
psychology approach. Journal of Retailing and Consumer Services 6, 67-78.
Ben-Akiva, M., Lerman, S.R., 1985. Discrete choice analysis: Theory and application to travel
demand. MIT Press, Cambridge, MA.
Beverland, M., Lim, E.A.C., Morrison, M., Terziovski, M., 2006. In-store music and consumer–
brand relationships: Relational transformation following experiences of (mis)fit. Journal
of Business Research 59 (9), 982-989.
Bucklin, R.E. Lattin, J.M., 1992. A model of product category competition among grocery
retailers. Journal of Retailing 68 (3), 271-293.
Chen, Z., Dubinsky, A.J., 2003. A conceptual model of perceived customer value in e-
commerce: A preliminary investigation. Psychology & Marketing 20 (4), 323-347.
Childers, T.L., Carr, C.L., Peck, J., Carson, S., 2001. Hedonic and utilitarian motivations for
online retail shopping behavior. Journal of Retailing 77, 511-535.
Dennis, C., Newman, A., Zaman, S., 2006. Marketing images and consumers’ experiences in
selling environments. Marketing Management Journal (Fall), 515-599.
Dodds, W.B., Monroe, K.B., Grewal, D., 1991. The effects of price, brand, and store information
on buyers’ product evaluations. Journal of Marketing Research 28 (3), 307-319.
Donovan, R.J., Rossiter, J.R., 1982. Store atmosphere: An environmental psychology approach.
Journal of Retailing 58 (1), 34-57.
Donovan, R.J., Rossiter, J.R., Marcoolyn, G., Nesdale, A., 1994. Store atmosphere and
purchasing behavior. Journal of Retailing 70 (3), 283-294.
Finn, A. Louviere, J.J., 1996. Shopping centre image, consideration and choice: Anchor store
contribution. Journal of Business Research 35, 241-251.

31
Gilmore, R., Margulis, W., Rauch, R.A., 2001. Consumer’s attitudes and retailers’ images in
creating store choice. A study of two different sides of the same story. International
Journal of Value-Based Management 14 (3), 205-221.
Grover, R. Srinivasan, V., 1992. Evaluating the multiple effects of retail promotions on brand
loyal and brand switching segments. Journal of Marketing Research 29, 76-89.
Hansen, T., 2003. Intertype competition: Specialty food stores competing with supermarkets.
Journal of Retailing and Consumer Services 10 (1), 35-49.
Hensher, D.A., Rose, J., Greene, W.H., 2005. Applied choice analysis: A primer. Cambridge
University Press, Cambridge, MA.
Holbrook, M., 1999. Introduction to consumer value. In: Holbrook, M., (Ed.), Consumer value:
A framework for analysis and research. Routledge, New York, NY, pp. 1-28.
Jones, M.A., Reynolds, K.E., Arnold, M.J., 2006. Hedonic and utilitarian shopping value:
Investigating differential effects on retail outcomes. Journal of Business Research 59 (9),
974-981.
Kaltcheva, V., Weitz, B.A., 2006. The moderating influence of motivational orientation on the
relationship between shopping environment arousal and behavior. Journal of Marketing
Winter, 107-118.
Koelemeijer, K., Oppewal, H., 1999. Assessing the effects of assortment and ambience: A choice
experimental approach. Journal of Retailing 75(3), 319-345.
Kotler, P., 1973-1974. Atmospherics as a marketing tool. Journal of Retailing 49 (Winter), 48-
61.
Kozinets, R.V., Sherry, J.F., DeBerry-Spence, B., Duhachek, A., Nuttavuthisit, K., Storm, D.,
2002. Themed flagship store in the new millennium: Theory, practice, prospects. Journal
of Retailing 78 (1), 17-29.
Lal, R. Rao, R., 1997. Supermarket competition: The case of everyday low pricing. Marketing
Science 16 (1), 60-80.
Louviere, J.J. Hensher, D.A., 1983. On the design and analysis of simulated or allocation
experiments in travel choice modeling. Transportation Research Record, 11–17.
Louviere, J.J., Hensher, D., Swait, J., 2000. Stated choice methods: Analysis and applications.
Cambridge University Press, New York, NY.
Louviere, J.J. Woodworth, G.G., 1983. Design and analysis of simulated consumer choice or
allocation experiments; an approach based on aggregate data. Journal of Marketing
Research 20, 350-367.
Mattila, A., Wirtz, J., 2001. Congruency of scent and music as a driver of in-store evaluations
and behavior. Journal of Retailing 77, 289-293.
McFadden, D., 1974. Conditional logit analysis of qualitative choice behavior. In: Zarembka, P.,
(Ed.), Frontiers of Econometrics. Academic Press, New York, NY.
Milliman, R.E., 1982. Using background music to affect the behavior of supermarket shoppers.
Journal of Marketing 46 (2), 86-91.
Oppewal, H., Koelemeijer, K., 2005. More choice is better: Effects of assortment size and
composition on assortment evaluation. International Journal of Research in Marketing 22
(1), 45-60.
Oppewal H., Timmermans H.J.P., 1999. Modeling consumer perception of public space in
shopping centers. Environment and Behavior 31 (1), 45-65.
Oppewal H., Timmermans H., Louviere J.J., 1997. Modelling the effects of shopping centre size
and store variety on consumer choice behaviour. Environment and Planning A 29, 1073-
1090.

32
Parsons, A.G., 2003. Assessing the effectiveness of shopping mall promotions: customer
analysis. International Journal of Retail and Distribution Management 31 (2), 74-79.
Peñaloza, L., 1998. Just doing it: A visual ethnographic study of spectacular consumption
behavior at Niketown. Consumption, Markets and Culture 2 (4), 337-400.
Pepall, L.M., Richards, D.J., 2002. The simple economics of brand stretching. Journal of
Business 75, 535-552.
Pine, J.B., Gilmore, J.H., 1999. The experience economy: Work is theatre and every business a
stage. Harvard Business School Press, Boston, MA.
Schmitt, B.H., 1999. Experiential marketing. Journal of Marketing Management 15 (1-3), 53-67.
Sherman, E., Mathur, A., Smith, R.B., 1997. Store environment and consumer purchase
behaviour: Mediating role of consumer emotions. Psychology & Marketing 14 (July),
367-378.
Sit, J., Merrilees, B., 2005. Understanding the experiential consumption of special event
entertainment (SEE) at shopping centres: An exploratory study. ANZMAC Conference,
Wellington, New Zealand.
Sit, J., Merrilees, B., Birch, D., 2003. Entertainment-seeking shopping centre patrons: the
missing segments. International Journal of Retail and Distribution Management 31 (2),
80-94.
Smith, P.C., Curnow, R., 1966. Arousal hypothesis and the effects of music on purchasing
behavior. Journal of Applied Psychology 50 (June), 255-256.
Spector, R., 2005. Category killers: The retail revolution and its impact on consumer culture.
Harvard Business School Press, Boston, MA.
Sweeney, J.C., Soutar, G.N., Johnson, L.W., 1999. The role of perceived risk in the quality-value
relationship: A study in a retail environment. Journal of Retailing 77 (1), 75-105.
Teas, K.R., Agarwal, S., 2000. The effects of intrinsic product cues on consumers’ perceptions of
quality, sacrifice, and value. Journal of the Academy of Marketing Science 28 (2), 278-
290.
Turley, L.W., Milliman, R.E., 2000. Atmospheric effects on shopping behavior: A review of the
experimental evidence. Journal of Business Research 49 (2), 193–211.
Vargo, S.L. Lusch, R.F., 2004. Evolving to a new dominant logic of marketing. Journal of
Marketing 68 (January), 1-17.
Watson, M., Shove, E., 2005. Cultures of consumption. Working paper series, Materialising
consumption: Products, projects and the dynamics of practice, ESRC-AHRC Research
Programme.
Williams, C., 2004 A lifestyle choice? Evaluating the motives of Do-It-Yourself (DIY)
consumers. International Journal of Retail and Distribution Management 32 (5), 270-278.
Woodward, I., 2003. Divergent narratives in the imaging of the home amongst middle-class
consumers – Aesthetics, comfort and the symbolic boundaries of self and home. Journal
of Sociology 39(4), 391-412.

33
Figures
Figure 1. Single choice task
Option 1
Option 2
Option 3
Store Feature
Store A
Store B
Neither
Customer Service Levels
High
High
Distance from Home
Medium (15 mins)
Far (25 mins)
Postpone
Ease of Access
Medium
Difficult
my trip
Friendliness of Staff
High
Low
In-store Event (as previously
described)
One Month
Only
Today
Only
Neither
Price Discount
None
2.5% off
Option 1 nor
Product Range
Medium
Small
Option 2
Store Appearance
Not especially
attractive
Attractive
are attractive
Store Format
Specialist
Store
General
Hardware Store
Which ONE option would
you choose?
Option 1
¨
Option 2
¨
Option 3
¨

34
Tables
Table 1. Store attributes and levels
Attribute
Level 1
Level 2
Level 3
Customer service levels
Low
Medium
High
Distance from home
5 minutes
15 minutes
25 minutes
Ease of access
Easy
Medium
Difficult
Friendliness of staff
Low
Medium
High
In-store event
No event
Event today only
Event for one month
Price discount
No discount
2.5% discount
5% discount
Product range
Small
Medium
Large
Store Appearance
Unattractive
Normal
Attractive
Store Format
Specialist store
General hardware
-

35
Table 2. Model I estimation results
Variable
Coefficient
s.e.
t-stat
Alternative specific constant (specialist)
0.082
0.050
1.616
Alternative specific constant (general)
-0.185**
0.052
-3.532
Customer Service – low
-0.408
-
-
Customer Service – medium
0.067
0.035
1.904
Customer Service – high
0.341**
0.039
8.755
Distance – 5 minutes
0.606
-
-
Distance – 15 minutes
0.071
0.042
1.682
Distance – 25 minutes
-0.677**
0.041
-16.375
Access – easy
0.375
-
-
Access – medium
0.260**
0.039
6.600
Access – difficult
-0.635**
0.037
-17.174
Staff friendliness – low
-0.481
-
-
Staff friendliness – medium
0.092*
0.043
2.160
Staff friendliness – high
0.389**
0.039
9.965
No in-store event
-0.061
-
-
In-store event – today only
-0.035
0.041
-0.857
In-store event – one month
0.096*
0.038
2.497
No discount
-0.343
-
-
Discount level – 2.5%
0.067
0.045
1.479
Discount level – 5%
0.276*
0.044
6.202
Product range – low
-0.479
-
-
Product range – medium
0.053
0.042
1.283
Product range – large
0.426*
0.036
11.984
Appearance – unattractive store
-0.112
-
-
Appearance – normal store
0.130*
0.039
3.355
Appearance – attractive store
-0.018
0.035
-0.499
Summary Statistics
Observations
4,572
Log Likelihood
-3305.48
Rho Squared
0.196
Rho Squared (Adjusted)
0.193
Note: * p ≤ .05, ** p ≤ .001

36
Table 3. Model II estimation results
Specialist Store
General (big box) store
Variable
Coefficient
s.e.
t-stat
Coefficient
s.e.
t-stat
Alternative specific constant
0.075
0.058
1.299
-0.242**
0.065
-3.750
Customer Service – low
-0.497
-
-
-0.469
-
-
Customer Service – medium
0.015
0.078
0.193
0.241*
0.081
2.997
Customer Service – high
0.482**
0.078
6.220
0.228*
0.086
2.655
Distance – 5 minutes
0.590
-
-
0.642
-
-
Distance – 15 minutes
0.115
0.076
1.510
0.057
0.081
0.711
Distance – 25 minutes
-0.705**
0.085
-8.310
-0.699**
0.094
-7.420
Access – easy
0.434
-
-
0.418
-
-
Access – medium
0.284**
0.075
3.808
0.198*
0.078
2.532
Access – difficult
-0.718**
0.085
-8.489
-0.616**
0.091
-6.772
Staff friendliness – low
-0.446
-
-
-0.516
-
-
Staff friendliness – medium
0.075
0.079
0.947
-0.013
0.083
-0.151
Staff friendliness – high
0.371**
0.073
5.106
0.529**
0.078
6.762
No in-store event
-0.035
-
-
-0.026
-
-
In-store event – today only
-0.117
0.076
-1.542
-0.026
0.086
-0.303
In-store event – one month
0.152*
0.073
2.084
0.052
0.078
0.664
No discount
-0.189
-
-
-0.361
-
-
Discount level – 2.5%
0.052
0.077
0.676
0.076
0.084
0.907
Discount level – 5%
0.137
0.077
1.773
0.285**
0.086
3.293
Product range – low
-0.495
-
-
-0.375
-
-
Product range – medium
0.143
0.077
1.864
-0.106
0.084
-1.265
Product range – large
0.352**
0.077
4.594
0.481**
0.077
6.232
Appearance – unattractive store
0.000
-
-
-0.150
-
-
Appearance – normal store
0.112
0.075
1.493
0.210*
0.082
2.552
Appearance – attractive store
-0.112
0.076
-1.478
-0.060
0.089
-0.676
Summary Statistics
Observations
4,572
Log Likelihood
-3290.20
Rho Squared
0.200
Rho Squared (Adjusted)
0.197
Note: * p ≤ .05, ** p ≤ .001

37
Table 4. Model III estimation results
Specialist Store
General (big box) store
Variable
Coefficient
s.e.
t-stat
Coefficient
s.e.
t-stat
Alternative specific constant
0.718
0.052
13.81
-0.197**
0.546
-3.608
Customer Service – low
-0.491
-
-
-0.404
-
-
Customer Service – medium
-0.028
0.654
-0.425
0.155*
0.066
2.338
Customer Service – high
0.519**
0.068
7.685
0.249**
0.72
3.432
Distance – 5 minutes
0.571
-
-
0.600
-
-
Distance – 15 minutes
0.075
0.067
1.128
0.032
0.068
0.477
Distance – 25 minutes
-0.646**
0.069
-9.301
-0.632**
0.075
-8.428
Access – easy
0.400
-
-
0.394
-
-
Access – medium
0.276**
0.068
4.045
0.200*
0.065
3.088
Access – difficult
-0.676**
0.067
-10.138
-0.594**
0.072
-8.260
Staff friendliness – low
-0.387
-
-
-0.528
-
-
Staff friendliness – medium
0.066
0.069
0.956
-0.028
0.065
-0.433
Staff friendliness – high
0.321**
0.059
5.467
0.556**
0.63
8.816
No discount
-0.209
-
-
-0.380
-
-
Discount level – 2.5%
0.090
0.615
1.466
0.134
0.071
1.888
Discount level – 5%
0.119
0.641
1.866
0.246**
0.071
3.471
Product range – low
-0.451
-
-
-0.341
-
-
Product range – medium
0.097
0.065
1.500
-0.132
0.067
-1.970
Product range – large
0.354**
0.062
5.701
0.473**
0.65
7.267
Appearance – unattractive store
0.004
-
-
-0.149
-
-
Appearance – normal store
0.069
0.063
1.103
0.208*
0.66
3.142
Appearance – attractive store
-0.065
0.066
-0.989
-0.059
0.073
-0.807
In-store event attribute and interactions
No in-store event
-0. 058
-
-
0.081
-
-
In-store event – today only
-0.080
0.059
-1.353
-0.138
0.070
-0.198
In-store event – one month
0.138*
0.616
2.237
0.057
0.064
0.893
EX1 (Aesthetic)
0.066
0.089
0.739
-0.007
0.092
-0.074
EX2 (Education)
0.001
0.086
0.023
0.020
0.088
0.226
EX3 (Entertainment)
-0.089
0.085
-1.057
-0.092
0.088
-1.051
Event_today*Aesthetic event
-0.850
0.091
-0.937
0.028
0.095
0.296
Event_today*Edu event
0.101
0.088
1.139
0.142
0.094
1.517
Event_today*Ent event
-0.006
0.088
-0.670
-0.343**
0.094
-3.663
Event_1month*Aesthetic event
0.162
0.090
1.801
0.091
0.087
1.041
Event_1month * Edu event
-0.037
0.885
-0.416
-0.190*
0.086
-2.216
Event_1month * Ent event
-0.059
0.865
-0.682
0.134
0.084
1.591
Summary Statistics
Observations
4,572
Log Likelihood
-3278.22
Rho Squared
0.204
Rho Squared (Adjusted)
0.198
* p ≤ .05, ** p ≤ .001