Professional Documents
Culture Documents
Customers
Malaika Brengman
Vrije Universiteit Brussel
Kim Willems
Vrije Universiteit Brussel and Hasselt University
Yannick Joye
University of Leuven
ABSTRACT
The incorporation of greenery in retail areas has scarcely been explored in environmental
psychology studies. In a 2 2 experimental design, the effect of in-store vegetation on consumer
emotions and responses toward the store was studied, considering the moderating role of the
information rate of the retail setting. While introducing foliage in the store environment did not
evoke feelings of excitement, it was found to elicit pleasure and to reduce stress in a complex store
interior. Given the impact of pleasure and stress on consumers approach/avoidance responses, these
findings support the potential of integrating greenery, particularly in spatially dense stores.
C 2012 Wiley Periodicals, Inc.
It is widely accepted in psychology that the first level of ing environments to produce specific emotional effects
response to any environment is affective (emotional) in in the buyer that enhance his purchase probability
nature (Ittelson, 1973). Empirical research has demon- and indicated that this would become an important
strated that shopping environments can also evoke marketing tool for retailers. The more recent finding
emotional responses in consumers (Machleit & Eroglu, that almost two-thirds of all purchase decisions are
2000) and that these emotions, in turn, influence shop- made in the store (Point-of-Purchase Advertising In-
ping behaviors and outcomes (Darden & Babin, 1994; stitute (POPAI), 1998) has turned the attention of
Donovan & Rossiter, 1982; Sherman, Mathur, & Smith, practitioners and marketing researchers to the point
1997). A neglected emotions-evoking element in the of purchase and store atmospherics. There is now a
shopping environment appears to be in-store greenery. growing recognition that store interiors and exteriors
Although vegetation has been demonstrated to have can be designed to create specific feelings in shoppers
positive and stress-mitigating influences in humans that can have an important cueing or reinforcing ef-
(among others, Bringslimark, Hartig, & Patil, 2009; fect on purchase (Newman & Foxall, 2003; Shapiro
Ulrich, 1993; Van Den Berg, 2009), it is unclear how and Associates, 2005). In the last two decades, there
it would affect in-store emotions and behaviors. While have been many studies concerning store atmospherics
many environmental psychology studies have inves- (Babin & Attaway, 2000; Fowler, Wesley, & Vazquez,
tigated retail settings and green environments sepa- 2007; McGoldrick & Pieros, 1998; Michon, Chebat, &
rately, the incorporation of greenery in retail areas has Turley, 2005; Sherman, Mathur, & Smith, 1997), pre-
barely been explored. Therefore, the objective of this dominantly based upon environmental psychology and
study is to determine the effect of introducing foliage specifically on the Mehrabian and Russell stimulus-
into the store environment. The role of store complex- organism-response (SOR) model (1974a), which was
ity will be considered as a potential moderating factor. first applied to the study of store atmosphere by Dono-
van and Rossiter (1982). In this context, it has been
proposed that approach/avoidance behaviors by cus-
STORE ATMOSPHERICS tomers are largely determined by individual internal
(cognitive, emotional, and physiological) responses to
Kotler (1973, p. 50) introduced the term atmospher- the store environment. Especially the pleasure and
ics to denote the conscious designing of [ . . . ] buy- arousal dimensions have been shown to be useful in
807
determining shopper behavior. As such, atmospheric & Pruyn, 2008), office environments (cf. Kweon, Ulrich,
variables have been found to influence a wide variety Walker, & Tassinary, 2008), roadside views (cf. Cack-
of consumer evaluations and behaviors (Turley & Mil- owski & Nasar, 2003), school settings (cf. Han, 2009),
liman, 2000). and home environments (cf. Hartig & Fransson, 2009;
While some authors have focused on overall per- Kaplan, 2001).
ceptions of the general store interior (e.g. Akhter, An- By contrast, only very few inquiries seem to have
drews, & Durvasula, 1994; Donovan, Rossiter, Mar- focused on the effects of greenery in commercial set-
coolyn, & Nesdale, 1994; Donovan & Rossiter, 1982; tings. Noteworthy in this context are the findings by
Grossbart, Hampton, Rammohan, & Lapidus, 1990), Kathleen Wolf (2003, 2004a, 2004b, 2005, 2008), who
others have investigated the effects of specific in-store carried out a multistudy research program, focusing
stimuli. Hence, some store interior variables such specifically on the impact of urban vegetation and trees
as music and scent have received ample attention, in streetscapes. She found that retail environments
whereas others have not received the attention they were more pleasing when there were trees in the streets
probably deserve (Turley & Milliman, 2000). With re- as compared to when the streets were treeless. Judg-
gard more specifically to point-of-purchase and dec- ments on retailers and merchandize appeared to be
oration variables (cf. the classification of Berman & more positive in streetscapes with urban greenery. This
Evans, 1995), attention has focused merely on prod- also seemed to be the case for inferences about prod-
uct displays and point-of-purchase information as well uct value, product quality, and merchant responsive-
as promotional signs, while neglecting decorative ele- ness, leading to enhanced expectations of shopping ex-
ments, such as artwork and wall decorations (Turley periences in vegetated commercial settings. Also, pa-
& Milliman, 2000). The impact in particular of one tronage intentions appeared higher in commercial set-
specific decorative element, that is, in-store greenery, tings with trees as compared to settings without trees.
which is the main focus of the current paper, strangely Although positive price increments were associated
appears to have been entirely ignored by academic with the presence of trees (+9% for small cities and
researchers. +12% for large cities), respondents claimed they would
travel farther to visit a business district with trees.
They also reported to be willing to stay longer once
there.
THE IMPACT OF GREENERY IN THE Experimentally testing the effects of, for example,
STORE ENVIRONMENT plants on consumer behavior in a shopping mall, Bu-
ber, Ruso, and Gadner (2006) found that consumers
The intuition that nature has positive and healing influ- were more inclined to enter the mall when it featured
ences on humans has been around since time immemo- vegetation. Moreover, the presence of greenery trans-
rial. These last few decades there has been a steady lated into higher exploration rates (e.g. more detailed
increase of environmental psychology research into the inquiries about offers). Increases in physical contact
positive impact of nature and greenery on human atti- with products and with other people in the shopping
tudes, moods, and behavior in different settings. For ex- mall were also apparent when plants were present.
ample, different studies show that humans prefer veg- These results corroborate and extend environmental
etated settings over artifact-dominated environments psychology results about greenery from residential and
and that greenery can reduce stress in humans. These other settings to retail environments.
effects are frequently explained as a remnant of a While Joye, Willems, Brengman, and Wolf (2010) in-
shared evolutionary history, during which human be- troduced the notion of Biophilic Store Design to denote
ings were deeply dependent on nature for their subsis- the use of (references to) nature in the design of retail
tence (e.g. Heerwagen & Orians, 1993; Ulrich, 1993). environments in order to tap into the purported ben-
According to this view, humans have evolved positive eficial effects of greenery, no empirical studies seem
affective responses toward green elements and set- so far to have investigated the impact of in-store
tings, such as [ . . . ] flowers, ripe fruits, savanna (open foliage.
forests that give easy visual access), growth and leaf Essentially two main findings from the environmen-
patterns of healthy savanna trees, [and] closed forest tal psychology literature studying the beneficial effects
canopy (shelter), [ . . . ] (Thornhill, 1998, p. 562). Hard- of greenery are of interest for the current paper (see also
wired positive affective responses to these elements Joye et al., 2010). The first is that nature consistently
and conditions conveyed an adaptive benefit to our an- leads to preference/liking reactions in humans. Differ-
cestors because they could offer them resource and ent studies have confirmed that scenes featuring green-
refuge opportunities. Colarelli and Dettmann (2003) ery are preferred over nonnatural, artifact-dominated
provide an overview of how these evolved preferences settings, even if the latter are well tended (see Ulrich,
are already commonly being used by marketers in ad- 1993, for a review). Based on these findings, it can be
vertising practice. expected that stores with in-store vegetation will be
As to research on the impact of vegetation on hu- perceived as more pleasurable compared to stores with-
man behavior, the literature provides a wide range of out vegetation, which results in the formulation of the
study contexts, such as hospitals (cf. Dijkstra, Pieterse, following hypothesis:
(1974a). Considering the potential moderating role of tic differential scale items: simplecomplex, sparse
the spatial density or complexity of the store en- dense, patternedrandom, clearconfusing, plain
vironment, this study aims to ascertain how in-store ambiguous, and uncrowdedcrowded (adapted from
vegetation affects the shoppers feelings of pleasure, ex- Mehrabian & Russell, 1974b; range: 3 to +3; Cron-
citement, as well as stress and how these evoked feel- bachs = 0.86). An independent samples t-test demon-
ings engender further responses toward the store. An strates that the information rate in the "complex"
overview of the studys research hypotheses is provided store environment was significantly higher than in the
in Figure 2. "lean" store environment (0.992complex vs. 0.444lean ,
p < 0.001). While the integration of in-store greenery
appears to increase the information rate of the envi-
RESEARCH METHODOLOGY ronment somewhat, an ANOVA with post-hoc Scheffe-
test, comparing the information rate between all con-
ditions, does not indicate any significant differences
Experimental Design between the environments with or without vegeta-
tion: the information rate of the complex store envi-
A 2 2 factorial experimental design was set up to test
ronment remains significantly higher (1.06with vegetation
these hypotheses, within a fashion retail context. Ma-
and 0.93without vegetation ) than the information rate of
nipulations included two conditions of store environ-
the lean store environment (0.33with vegetation and
mental complexity (lean vs. complex), with and with-
0.57without vegetation ). An independent samples t-test also
out in-store vegetation (see Figure 3). Two separate
does not reveal any significant difference in informa-
pretests of the stimuli confirmed the adequateness of
tion rate between store environments with or with-
the manipulations (reported underneath). The exper-
out vegetation (0.35with vegetation vs. 0.25without vegetation ,
iment that was subsequently conducted on the basis
p = 0.764).
of these stimuli was monadic in nature (i.e. each re-
According to OKeefe (2003), there is no need to check
spondent evaluated only one store). Participants were
the manipulation of inserting "greenery in the store envi-
exposed to a picture of the store and were asked to
ronment," because whether respondents noted the pres-
imagine shopping in that store. This is a procedure
ence of vegetation or not, potential effects of integrating
that has been shown to accurately simulate reactions
greenery may occur unconsciously, without consumers
toward real environments (Bateson & Hui, 1992). A
even noticing its presence. Still, in the pretest respon-
meta-analysis by Stamps (1990) found a 0.86 correla-
dents were asked to give their opinion with respect to
tion between environmental preferences obtained by
the greenery present in the photographs of the store en-
photographs and preferences obtained by exposures to
vironments. The response categories concerned (1) "If
real environments.
there were plants, I didnt notice them at all," (2) "If
there were plants, I didnt notice them," (3) "If there
were plants, they are in harmony with the environ-
Manipulation Checks
ment," (4) "Plants were quite prominently present,"
In order to check whether the experimental manipu- and (5) "Plants were very prominently present." An
lations were successful, an online pretest of the four independent samples t-test indicates that vegetation
photographic stimuli was conducted in a between- was noted more in environments with versus with-
subjects design using a convenience sample of 81 re- out vegetation (1.94with vegetation vs. 1.10without vegetation ,
spondents. The average respondent age was 34.17 years p < 0.001). Still the pre-test findings also show that
old (range: 2066) and 39.5% of the participants were even in the condition with vegetation, 52.8% of respon-
male. dents indicated not to have noticed the plants at all. An
The "information rate of the store environ- ANOVA with post-hoc Scheffe test reveals that plants
ment" was measured by means of six seman- appeared somewhat more prominently present in the
Sex
Male 437 40.5 454 44.2 453 41.9 494 44.6 2 1838 42.8 8,205,443 49.5
Female 642 59.5 573 55.8 627 58.1 613 55.4 0.173 2455 57.2 8,372,169 50.5
Age
Min./Max 18 69 18 69 18 69 18 69 18 69 0 >99
Mean/SD 41.4 13.1 42.3 12.8 42.3 13.4 42.8 13.0 42.2 13.1 39,2 -
< 20 43 4.0 41 4.0 34 3.1 30 2.7 2 148 3.4 3,927,975 23.7
2040 493 45.7 420 40.9 480 44.4 458 41.4 0.181 1851 43.1 4,195,201 25.3
4065 512 47.5 540 52.6 536 49.6 588 53.1 2176 50.7 5,916,067 35.7
6580 31 2.9 26 2.5 30 2.8 31 2.8 118 2.7 1,890,472 11.4
>80 - - - - - - - - - - 647,890 3.9
Educational level
Lower 599 55.5 602 58.6 604 55.9 587 53.0 2 2392 55.7 6,905,000 63.4
Higher 480 44.5 425 41.4 476 43.0 520 47.0 0.079 1901 44.3 3,987,000 36.6
Civil status
Married 694 64.3 676 65.8 684 63.3 707 63.9 2 2761 64.3 9,603,148 58.3
Other 385 35.7 351 34.2 396 36.7 400 36.1 0.667 1532 35.7 6,882,639 41.7
a
Based on Statistics Netherlands online.
Abs. = absolute numbers; Rel. = in relative terms.
813
Figure 3. Manipulations of store complexity and in-store vegetation.
subconstructs, mean scores are calculated, to be used conducted with in-store vegetation and store com-
in further analyses. plexity as fixed factors and evoked feelings of plea-
sure, stress, and excitement as dependent vari-
ables. In order to control for any confounding effects of
ANALYSES AND RESULTS the sociodemographic variables age and gender, these
were added to the analysis as covariates. Although the
So as to be able to formulate an answer to the research multivariate tests support the main effects of in-store
questions in this study, the analyses will be discussed vegetation (Wilks Lambda = 0.997, F (1, 4289) = 3.89,
in two stages. First, the hypotheses with regard to the p = 0.009) and complexity (Wilks Lambda = 0.971,
impact of in-store greenery on perceived pleasure, F (1, 4289) = 42.41, p < 0.001), the overall interac-
stress, and excitement in both lean and complex tion effect between the two factors appears to be only
store environments will be tested. Next, the mediat- marginally significant (Wilks Lambda = 0.998, F (1,
ing effects of these consumer emotions on subsequent 4289) = 2.17, p = 0.090). Nevertheless, a closer look
approach and avoidance responses will be addressed. at the univariate analyses provides clear evidence of
some pronounced interaction effects, as will be illus-
The Impact of In-Store Greenery on trated further. Both covariates, age (Wilks Lambda
Consumer Emotions, Considering the = 0.995, F (1, 4289) = 6.95, p < 0.001) as well as
gender (Wilks Lambda = 0.984, F (1, 4289) = 23.90,
Moderating Role of Store Complexity
p < 0.001), also appear to have a significant impact
In order to test Hypotheses H1 through H3, a two-way on the emotional reactions toward the store environ-
multivariate analysis of covariance (MANCOVA) was ments.
Pleasure Evoked by In-Store Greenery. Regarding to alleviate stress only in a complex retail setting (per-
the impact of in-store greenery on evoked pleasure (H1), ceived stress = 3.51(without vegetation) vs. 3.35(with vegetation) , t
a notable interaction effect of store complexity that was (2107.785) = 2.84, p = 0.005), which is in line with the
not anticipated can be observed (F (1, 4289) = 6.41, p = present studys hypotheses.
0.011).
Figure 4 clearly illustrates that adding greenery in
Excitement Evoked by in-Store Greenery. Although
the store interior can evoke higher feelings of pleasure,
it was tentatively hypothesized that the presence of
but that this effect is distinctly more pronounced in
in-store vegetation would lead to increased feelings of
complex store environments. As a matter of fact, an in-
excitement in consumers (H3a) and that this increase
dependent samples t-test reveals that the introduction
would be more pronounced in simple store environ-
of vegetation does not generate a significant effect in
ments than in complex retail settings (H3b), the find-
the lean store environment (t (2152.547) = 0.50, p =
ings do not reveal any effect of in-store greenery on
0.617). In a complex retail setting, on the other hand,
elicited excitement. A possible explanation is that veg-
the positive effect of foliage on perceived pleasure is
etation, basically plants, are not enough to surprise a
clearly evident (perceived pleasure = 3.86(without vegetation)
shopper. Not only may consumers be too familiar with
vs. 4.05(with vegetation) , t (2115.773) = 3.98; p < 0.001).
such natural elements to feel excited by them, previous
So the integration of greenery in the store environment
research also indicated that such elements are actually
apparently does not result in an equally beneficial effect
experienced as only moderately fascinating (as opposed
on the consumers perceived pleasure in complex versus
to, say, watching television; see Herzog, Black, Foun-
lean retail settings. This implies that the first hypothe-
taine, & Knotts, 1997).
sis is only partially supported, that is, in complex store
interior conditions only. As is observable, however, the
lean store interior was generally rated to be more
pleasurable than the complex store environment (F
(1, 4289) = 51.57, p < 0.001), which may have provided
too little room for improvement by integrating in-store
vegetation.
Approach Approach
Direct Effects Indirect Effects through
Pleasure Stress Excitement
Coefficient p Effect CIlow CIup Effect CIlow CIup Effect CIlow CIup
between greenery and store complexity on approach while Stress appears to have the least explanatory
and avoidance intentions, through evoked pleasure power (0.066, p < 0.001).
and/or stress. With regard to Avoidance, the multiple regres-
As a matter of fact, the multiple regression analysis sion analysis reveals that 62.1% of the variance (ad-
demonstrates that 66.7% of the variance in Approach justed R2 ) is explained by the variance in the explana-
(adjusted R2 ) is explained by the variance in the ex- tory variables Pleasure, Stress, and Excitement,
planatory variables Pleasure, Stress, and Excite- which all appear to contribute significantly. While the
ment. Each of these emotional dimensions appears to standardized beta coefficients reveal that pleasure is
contribute significantly to approach intentions. The again the most important explanatory factor (0.663,
standardized beta coefficients reveal that pleasure p < 0.001), stress (0.323, p < 0.001) comes second,
is the most important explanatory factor (0.721, p < and excitement appears to contribute least (0.084,
0.001), followed by Excitement (0.297, p < 0.001), p < 0.001).