Documente Academic
Documente Profesional
Documente Cultură
A R T I C L E
I N F O
Article history:
Received 19 November 2015
Accepted 18 January 2016
Available online 19 February 2016
Keywords:
Farmers markets
Lifestyle segmentation
Food
Consumption motivation
Consumer behaviour
A B S T R A C T
Lifestyle segmentation, used to explore consumer attitudes, interests, opinions and values, is applied to
a local farmers market (the Otago farmers market, in New Zealand) in order to build a prole of consumer motivations and behavioural drivers beyond simple demographic data. Prior research indicates a
growing need for studies that seek to understand the intangible aspects of food selection and consumption. Farmers market consumers are often treated as a homogenous group, with purchase and consumption
decision making assumed to rest with tangible aspects of produce on offer, such as its organic nature or
lower cost. The overall experience of the consumer within the food purchase environment is often overlooked. The current study nds that three distinctly different consumer segments exist within the farmers
market context, with those concerned with the tangible aspects of produce being the smallest group of
consumers. Two consumer segments driven by differing experience-related motivations give the greatest insight into consumer behaviour in a farmers market context, with committed, loyal consumers being
concerned with relationships with sellers and producers, and experience orientated consumers motivated by the opportunity to interact with other farmers market attendees.
2016 Australian and New Zealand Marketing Academy. Published by Elsevier Ltd. All rights reserved.
C H I N E S E
A B S T R A C T
Ot
ago
2016 Australian and New Zealand Marketing Academy. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Famers markets continue to rise in popularity, particularly in
urban settings, where the concept of buying food directly from the
producer has many perceived consumer benets (La Trobe, 2001).
This short supply chain model, with its implied positive implications for the environment and local economies, appears to have
achieved considerable consumer trust and acceptance in the food
industry (Migliore et al., 2014). Indeed, where current concerns in
the food industry are raised (such as genetic modication of food
http://dx.doi.org/10.1016/j.ausmj.2016.01.003
1441-3582/ 2016 Australian and New Zealand Marketing Academy. Published by Elsevier Ltd. All rights reserved.
136
on offer, such as price and product availability or variability and nutritional status of offerings (e.g. McGuirt et al., 2011; Lucan et al.,
2015). Fewer studies assess the beliefs consumers hold towards
famers markets and their reasons for both attending, and for purchasing food products within them. Where studies of consumer
perceptions do exist, consumers have tended to be aggregated together to provide an overview of reasons for attending these markets,
and consumers directed towards assessing the aforementioned tangible aspects of the good on offer (e.g. La Trobe, 2001; Migliore et
al., 2014). Overall, food based literature devoted to discussing the
farmers market context regards it as a mechanism for providing
increased access to fresh produce in areas where these products are
not normally as freely available (i.e. non-rural areas) (Lucan et al.,
2015; McKinnon et al., 2009). Tourism literature, however, provides an alternative perspective through examination of culinary
tourist motivations, suggesting that food, as a major element of any
tourist product, can provide opportunities for relaxation, learning,
communication and entertainment within the food procurement
experience (ela et al., 2007; Plummer et al., 2005). Where attendance at a farmers market can be treated as an experience, the
wider implications of culinary tourism attributes are likely to be
relevant in the overall consumer perceptions and evaluation of a
farmers market, irrespective of the local nature of most farmers
market visitors.
3. The farmers market lifestyle
Where the economic benets of farmers markets to regions are
well documented, these markets also offer unique social benets to
communities, as well as to individual consumers (Hofmann et al., 2008).
Prior research has addressed the role of farmers markets in satisfying moral issues that many consumers now associate with food
production and supply (Aslet, 2002; La Trobe, 2001), increased demand
for observable quality indicators in food (such as the ability to meet
the producer) (Guthrie et al., 2006; Youngs and Holden, 2002) and
the degree of social cachet associated with simply being seen to purchase in a farmers market setting (Fernandez-Armesto, 2001; Guthrie
et al., 2006). These factors, combined, relate to the use of famers
markets by consumers in the construction of a personal lifestyle and
identity associated with that (Schaefer and Crane, 2001).
4. Farmers markets and lifestyle segmentation
Lifestyle segmentation, when compared to traditional demographic segmentation (such as age or gender categories) offers the
opportunity to understand groups of consumers in terms of their
attitudes, interests, opinions, values and consumption preferences
(Vyncke, 2002). In this way, lifestyle segmentation allows the attitudes and motivations (and thus behavioural drivers) of specic
groups of consumers to be uncovered, rather than proling the
average consumer (Nie and Zepeda, 2011). In the context of farmers
markets, factors such as the relationship between the seller (often
the producer) and the consumer, as well as relevant situational
factors are suggested as likely to have a greater effect on consumer behaviour than socio-demographic factors such as age, gender
and education (Migliore et al., 2015). Recent studies suggest a need
for greater understanding of intangible factors in farmers market
purchasing, such as familiarity of the consumer with the shopping experience, cultural values and personal food shopping desires
(Lucan et al., 2015). Where farmers market food experiences can
be said to be inseparable from local culture and thus authentic and
unique to the host community (Plummer et al., 2005), it is clear that
consumer behaviour in a farmers market context should also be
examined in a region-specic fashion.
When groups of consumers are clustered into segments, differentiations between the segments are generated through the total
pling plan did not include stratication, as the population size was
undened. A sample of 200 respondents was collected, exceeding
the requirements for using the cluster analysis technique to achieve
valid results (Hair et al., 1998). Similar sample sizes have been used
throughout the farmers markets literature: e.g. Govindasamy et al.
(1998) conducted a consumer survey of 336 respondents; a slightly
larger sample size of 446 farmers market customers was obtained by Connell et al. (2008); and a sample size of 404 consumers
was used by Wolf (1997).
The survey was administered in person, with respondents given
the survey to ll out at the time of intercept. To qualify for participation, respondents were 18 years or over. In order to decrease
potential bias of respondents, the surveys were not administered
on a competing festival or event day in the city.
5.4. Analysis
Data collected from the survey were coded and analysed using
SPSS. Respondents rated different aspects of the Otago Farmers
Market on three different seven point Likert type scale questions,
and the data produced from these types of scales are assumed to
be interval, thus allowing mean scores to be compared through
cluster analysis. Factor analysis was performed on survey variables as preliminary analysis in order to reduce the original variable
set into a representative cluster variate for subsequent cluster analysis. Frequencies for each variable were rst examined to determine
their variances, and the variables to include for analysis decided upon.
A two-step clustering approach was used, where hierarchical
cluster analysis was used to determine the potential number of clusters, then non-hierarchical k-means clustering to form respective
clusters. A dendrogram was derived from the hierarchical cluster
analysis, and used to interpret the cluster solution in terms of the
number of clusters, using the Wards linkage method. The clusters
derived from cluster analysis were then subjected to proling using
cross-tabs in order to label the segments in terms of demographics. Nominal data such as gender, age, occupation and residence were
the variables used to prole the clusters.
6. Results
6.1. Respondent characteristics
A review of the general characteristics of the respondent sample
indicates that those 1825 years old make up 45% of the sample, followed by 26 to 35 years (18%) and 4655 years (17%). The majority
of the sample are female (59%). The vast majority of the sample are
New Zealanders (75%), and nearly all of the sample are local residents (90%).
6.2. Factor analysis
The rst stage of the data analysis involved conducting exploratory factor analysis in order to achieve data reduction for subsequent
analysis. Due to the large number of variables included in the segmentation data, it was deemed appropriate to determine whether
there was a smaller subset of factors that could represent the original variable set. Upon analysis of frequencies data for each variable,
it became apparent that very little variance in the negative variables
(such as I feel bad about my decision to shop at the Market) was
present. A preliminary factor analysis was run including these variables, which gave a six factor solution. All the variables related to
negative attitudes towards the market loaded on a single factor, and,
with all of them also exhibiting such low variance, the ve negatively oriented variables were removed from the analysis. Two other
variables were also removed from the analysis upon examination of
frequency data; these variables exhibited a very small level of vari-
137
Table 1
KMO and Bartletts Test of Sphericity.
KaiserMeyerOlkin Measure of Sampling Adequacy
Bartletts Test of Sphericity
Approx. chi-square
Df
Sig.
.866
2123.106
231
.000
Table 2
Factor labels.
Factor number
Factor label
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Appeal of produce
Variety of produce
Satisfaction of shopping experience
Value of produce
Likelihood to repeat purchase at the Market
138
Table 3
Final cluster centres.
Cluster
Appeal
Variety
Satisfaction
Value
Repeat
.67907
.64750
.66699
.52788
.40826
.16073
.63524
.07747
.47425
.14470
1.21384
.06585
1.37431
.15250
1.27632
Cluster 1 scored highest on all ve factors, indicating that this respondent group
rated all perceived aspects of the farmers market the most favourably compared
to other groups. Cluster 2, on the other hand, scored lower on all factors but was
comparable to cluster 3. In terms of Appeal of produce, cluster 2 scored slightly
higher than cluster 3, as was the case for Satisfaction of shopping experience and
Likelihood to repeat purchase at the Market. Cluster 3 scored higher than cluster
2 (but not cluster 1) in terms of Variety of produce and Value of produce.
factor solution (Table 4). A Tukey HSD post-hoc test was used to determine where these differences lie, which showed that there were
differences in cluster means for all factors other than Value of
Produce (for which clusters 2 and 3 did not differ signicantly).
These observations indicate that cluster 1 clearly represents those
consumers of the market that are highly loyal and hold extremely
positive views of the market offerings in comparison with other retail
outlets. This cluster is likely to be regular consumers of farmers
markets, who prefer the markets compared to other retail outlets.
Consumers in cluster 2 do not have such an attachment to the
market, but are relatively well satised with their experiences at
the market and rate the produce relatively highly, and like cluster
1 likely to attend the market to purchase produce on a regular basis.
Consumers within this cluster are likely to go to the markets from
time to time, and are primarily interested in the experiential aspect
rather than simply coming to buy produce. Cluster 3 consumers are
more attracted to the aspects of the market to do with the availability of produce at the markets and believe that it may be better
value than other retail outlets. Rather than going to the markets for
the experience, the consumers within this cluster may be more interested in coming simply to buy what they believe to be better
produce at better prices than supermarkets or other grocery outlets.
Clusters 1 and 2 were of similar sizes (n = 79 and 77, respectively),
and cluster 3 was relatively smaller (n = 34). A tentative description of the three clusters based on the characteristics described above
and the cluster membership is summarised in Table 5.
6.5. Proling the clusters
Demographics were used to prole the clusters and examine any
signicant differences in terms of age, gender, occupation, country
of origin and city of residence. Cross-tabulations were performed
on the data, which revealed that the three clusters only differed signicantly across the dimensions of gender and occupation. The
Pearson Chi-Square value for gender is 0.01, meaning that the differences in gender between clusters can be dened as signicant.
The Pearson Chi-Square values for the other three demographic variables used in the cross-tabs were found to be insignicant.
For this reason, only the cross-tabs for gender and occupation are
included for observation. Upon examination of the cross-tab for occupation, it was found that several cells had expected counts of less
than 5, which made the cross-tab inappropriate for interpreta-
Table 4
Benets sought and satisfaction levels at the farmers market.
Mean
Std. deviation
Analysis N
5.53
4.72
5.38
5.88
4.68
4.58
5.28
6.41
6.55
6.07
6.14
5.97
6.05
6.12
6.34
6.49
5.20
5.98
6.08
6.15
6.09
6.03
1.153
1.430
1.236
1.032
1.478
1.373
1.429
.986
.820
.836
.843
.920
.922
.840
.837
.871
1.240
.864
1.004
.904
.901
1.103
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
190
Table 5
Cluster membership and labels.
Committed loyals
Experience oriented
Product oriented
TOTAL
Missing
139
Frequency
Per cent
79
77
34
190
4
41.6
40.5
17.9
100.0
140