You are on page 1of 10

Food Research International 54 (2013) 21122121

Contents lists available at ScienceDirect

Food Research International


journal homepage: www.elsevier.com/locate/foodres

Evaluating bottles and labels versus tasting the oils blind: Effects of
packaging and labeling on consumer preferences, purchase intentions
and expectations for extra virgin olive oil
Claudia Delgado, Aurora Gmez-Rico, Jean-Xavier Guinard
a
University of California, Davis Department of Food Science and Technology, United States

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this research was to assess Northern California consumer liking, purchase intent and sensory and
Received 16 April 2013 nutritional expectations for commercial extra virgin olive oils based on their packaging and labeling, and to com-
Received in revised form 10 October 2013 pare those ratings to hedonic and purchase intent ratings for the same oils tasted blind. A set of 18 commercial
Accepted 16 October 2013
extra virgin olive oils was evaluated by two groups of Northern California consumers under different conditions.
One group (N = 102) visually assessed the bottles and labels and then indicated their liking and purchase intent
Keywords:
Extra-virgin olive oil
for the oils, among other variables. Another group (N = 110) tasted the oils blind, and then indicated their liking
Packaging and purchase intent for the oils. We uncovered two preference segments in the evaluation of the bottles and la-
Labeling bels (packaging study), both of which liked the California EVOOs, but differed in their liking of the imported oils.
Consumer expectations Consumers in segment 1 liked mostly California EVOO bottles, while consumers in segment 2 also liked some of
Preference mapping the imported oils, particularly those from Italy. The preference map generated from the hedonic ratings in the
packaging study was quite different from that generated from the blind hedonic ratings, with California oils faring
much better when fully dressed, likely because of the desirable features of their packaging, the possible familiar-
ity of the consumers with them, the desire of most consumers to buy local, and maybe most importantly because
many consumers in the blind tasting study disliked bitter and pungent oils, and/or actually liked oils that were
somewhat rancid. Region of origin was the variable that showed the biggest impact on the overall liking of the
bottles and labels in the packaging study. California EVOOs were signicantly preferred over most of the
imported oils, overall and for preference segment 1. Consumers were willing to pay more for the oils in the pack-
aging study than they expected to pay in the blind tasting study. Purchase intent followed similar trends to over-
all liking in the packaging study, and it was inversely related to the purchase intent recorded in the blind tasting
study.
2013 Elsevier Ltd. All rights reserved.

1. Introduction way to expand markets for EVOO and it facilitates the retention of
EVOO properties and quality. The material of the container (e.g. glass,
In emerging markets, extra virgin olive oil (EVOO) is seen as a premi- plastic, metal), the conditions and length of storage are critical fac-
um and relatively new product with consumer behavior differing from tors determining the shelf life of packaged olive oil (Kanavouras,
that in traditional producer countries (Martnez, Aragons, & Poole, Hernandez-Munoz, & Coutelieris, 2006). In many ways the olive oil
2002; Santosa & Guinard, 2011; Zampounis, 2006). As a consequence, industry in the U.S., and particularly in California, is mimicking the wine
the way to market and sell extra virgin olive oil is different; the packag- industry with many variables for consumers to consider, e.g., region of or-
ing, labeling and brand become more important and so are other non- igin, local or imported, mono-varietal or blend, bottle style and design,
sensory variables that inuence purchase (Enneking, Neumann, & and label. If one would expect the selection and purchase of EVOO to be
Henneberg, 2007; Jaeger, 2006; Sandalidou & Baourakis, 2002). This is inuenced mostly by non-sensory variables, then the repeat purchase
not so much the case in high consumption regions where oil may be should be affected by tasting the product. Sensory variables should
bought in bulk directly from the producer or in some cases home pro- play a more signicant role after the product was initially purchased for
duced (Fotopoulos & Krystallis, 2001; Mtimet, Zaibet, Zairi, & Hzami, consumers to keep buying the product. Previous research conducted
2013). The packaging of extra virgin olive oil also matters in that it is a among Northern California consumers showed that price and region
of origin are critical factors inuencing the purchase of EVOO
(Delgado & Guinard, 2011a; Santosa & Guinard, 2011).
Corresponding author at: University of California, Davis, Department of Food Science
and Technology, One Shields Avenue, Davis, CA 95616-8598, United States. Tel.: +1 530
Brand and labeling have an impact on consumer behavior as well;
754 8659; fax: +1 530 752 4759. these non-sensory variables may indeed inuence the overall liking of
E-mail address: jxguinard@ucdavis.edu (J.-X. Guinard). a product in blind tasting conditions if the information regarding the

0963-9969/$ see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.foodres.2013.10.021
C. Delgado et al. / Food Research International 54 (2013) 21122121 2113

product is provided. For example, in comparisons of blind tasting vs. in- (Peryam & Pilgrim, 1957) for the following factors: overall liking, bottle
formed tasting conducted with other food products, results have shown size, bottle shape, container material, label design, pictures or illustra-
that there is a disconnection or suppression of the hedonic scores when tions on the label and colors on the label. Consumers also indicated
the product is blinded, or an increase in the liking for a product blinded their sensory attribute (intensity) expectations for overall avor,
but a reduction of the scores if the brand is not well known (Ares & fruitiness, bitterness and pungency, using a 5-point scale (1 = low,
Deliza, 2010; Bower & Turner, 2001; G. Caporale & Monteleone, 2004; 3 = medium, 5 = high); nutritional property expectations com-
Deliza & MacFie, 1996; Deliza, Rosenthal, Abadio, Silva, & Castillo, pared to a typical EVOO (5-point scale: 1 = less nutritional properties,
2005; Guinard et al., 2000; Mueller & Szolnoki, 2010). In the case of 3=about the same, 5=more nutritional properties); the price they
olive oil, Caporale, Policastro, Carlucci, and Monteleone (2006) found expected to pay for each bottle (7-point scale: 1 = $15, 2 = $510,
an effect of information on the hedonic scores. For example information 3 = $1015, 4 = $1520, 5 = $2025, 6 = $2530 and 7 = more
regarding region of origin affected product acceptability, with some than $30) and their intent to purchase it (5-point scale: 1 = de-
negative disconrmation in some cases; furthermore, information nitely would not buy, 2 = probably would not buy, 3 = maybe/
about cultivar signicantly affected the expectation of bitterness maybe not, 4 = probably would buy and 5 = denitely would
and pungency for those oils. An informed (non-blind) tasting with buy). Upon completion of the test, consumers lled out a brief survey
consumers showed that consumers generate language associated with that assessed their attitudes, consumption and purchase habits about
sensory properties of the oils (Santosa, 2010). In this regard, labeling is olive oil and collected their demographic information ().
also important because it may show nutritional content or may suggest
a particular nutritional benet to attract consumers (Ares & Deliza, 2.3.2. Blind tasting of the oils
2010; Bialkova & van Trijp, 2010). For the blind tasting, 10-mL samples were served at room tempera-
The main objective of this research was to assess Northern California ture in souf plastic cups (Solo cup model B200) covered with a trans-
consumer liking, purchase intent, and sensory and nutritional expecta- parent plastic lid (Solo cup model PL2 2-ounce). White bread (Classical
tions for commercial extra virgin olive oils based on their packaging White, Wonder) was provided as a sample carrier. Samples were coded
and labeling. Other objectives were to compare those ratings to hedonic with three-digit numbers and poured at least 30 min before the tasting.
and purchase intent ratings for the same oils tasted blind; and to inves- Water, previously ltered in a Millipore Milli-Q water ltration sys-
tigate the price consumers were willing to pay for the packaged EVOOs. tem, slices of Granny Smith apple and unsalted crackers were provided
as palate cleansers.
2. Materials and methods For each EVOO sample, consumers indicated their overall liking of
the oil using the 9-point hedonic scale (Peryam & Pilgrim, 1957). They
A set of commercial extra virgin olive oils was evaluated by two also recorded their intent to purchase the oil (5-point scale: 1 = de-
groups of Northern California consumers under different conditions. nitely would not buy, 2 = probably would not buy, 3 = maybe/
One group (N = 102) visually assessed the bottles and labels and then maybe not, 4 = probably would buy and 5 = denitely would
indicated their liking and purchase intent for the oils, among other var- buy); and the price they would be willing to pay for a 375 mL bottle
iables. Another group (N=110) tasted the oils blind, and then indicated (8-point scale: 1 = $0, 2 = $15, 3 = $510, 4 = $1015, 5 = $1520,
their liking and purchase intent for the oils, in a traditional preference 6 = $2025, 7 = $2530 and 8 = more than $30). They indicated their
mapping exercise (Delgado & Guinard, 2011a). The packaging study willingness to consume the product a second time (5 point scale: 1 = I
consisted of one session, while the blind tasting consisted of two ses- would certainly not consume this EVOO again, 2 = I would probably
sions. The FIZZ software (Biosystmes, Couternon, France) was used not consume this EVOO again, 3 = Not sure or undecided, 4 = I would
for data collection in both instances. probably consume this EVOO again, 5 = I would certainly consume
this EVOO again). And they used a Likert scale (5 points: 1 = Strongly
2.1. Consumers disagree to 5 = Strongly agree) to record their impressions of the
oil's quality, color, taste, aroma, and texture. Finally, consumers were
All consumers were recruited from the same sources and popula- asked if they would recommend this product to a family member or
tion in local supermarkets, at farmer's markets, and through inter- friend using a 5-point Likert scale (1=Not likely, 2=Somewhat unlike-
net sites such as Craigslist and the email directory of the University of ly, 3 = Uncertain, 4 = Somewhat likely, 5 = Very likely).
California, Davis. Eligibility criteria were consumption frequency of The blind tasting study consisted of two sessions. During the rst
extra virgin olive oil, US citizenship or residency, and age older than session, consumers evaluated a total of 12 samples with a 15-min
20 years for both studies. There was no signicant difference in the de- break taken every 5 samples, and a 1-min break between samples.
mographics of the two groups (p N 0.05). In the second session, consumers evaluated 10 samples, with a 15-min
break after the rst 5, and completed a brief exit survey that assessed
2.2. Olive oils their attitudes and beliefs about olive oil and collected their demographic
information. A total of 22 oils were evaluated in the blind tasting study;
Eighteen commercial extra virgin olive oils (EVOOs) were used in eighteen of those were used for the packaging study. More details
this research. Eleven were produced in California and the rest were about the protocols and scales used in the blind tasting study can be
imported. With the exception of the generic brand oils (GBS and GBI), found in Delgado and Guinard (2011a).
which were bought in a local supermarket, the oils were donated by ei-
ther producers or distributors in California. Upon receipt, the oils were 2.4. Experimental design and data analysis
stored in dark and cool conditions. The country of origin and packaging
characteristics of the EVOOs are shown in Table 1. A Williams Latin square design was used in both studies, also provid-
ed by the FIZZ software. The sample presentation order was completely
2.3. Evaluation protocols randomized.
The majority of the statistical analyses were executed using XL-Stat
2.3.1. Visual evaluation of the bottles and labels Version 2009.3.02 (Addinsoft) and SAS version 9.1 (SAS Institute, Cary,
For the assessment of the EVOO bottles and labels, the bottles were NC). The level of condence of alpha was set at 0.05. Univariate analyses
presented standing on a white background, and labeled around the (correlation, one-way analysis of variance (factor: clusters), and Fisher's
neck with three-digit numbers. For each EVOO bottle and label, con- LSD multiple mean comparisons) and multivariate analyses (principal
sumers indicated their degree of liking on the 9-point hedonic scale component analysis and cluster analysis Euclidean distance, Ward's
2114 C. Delgado et al. / Food Research International 54 (2013) 21122121

Table 1
Extra virgin olive oil codication and packaging characteristics.

#ID Origin Variety Bottle Cap Notes on label Volume


(mL)

U1 USA (California) Arbequina Glass, opaque (black) Black cap wrapper Award winning, handpicked and 250
crushed immediately after harvest
U2 USA (California) Blend Glass, clear Dark red cap wrapper COOC label 250
U3 USA (California) Arbequina/Arbosana/Koroneiki Glass, opaque (black) Golden metal screw twist COOC label 500
U4 USA (California) Sevillano Glass, clear Golden cap wrapper Filtered fresh pressed 375
U5 USA (California) Frantoio Glass, clear (inside a light green Black cap wrapper 375
carton box)
U6 USA (California) Mission/Manzanillo/Sevillano/ Glass, opaque (black) Brownish olive green cap wrapper COOC label, milled within 24 h of 250
Barouni/Ascolano harvest
U7 USA (California) Manzanillo/Mission Glass, opaque (black) White metal screw twist First cold pressed, decanted, 500
unltered
U8 USA (California) Blend Glass, clear Black cap wrapper Los Angeles Olive Oil Competition 250
bronze award, handpicked and
immediately rst cold pressed
within 24 h
U9 USA (California) Arbequina Glass, opaque (black) Golden metal screw twist COOC label, handpicked, unltered, 500
obtained directly from olives and
solely by mechanical means
U10 USA (California) Frantoio/Leccino/Pendolino/ Glass, clear (inside carton box) Wood cork Certied Organic by COOC, cold 375
Coratina processed, unltered
U11 USA (California) Mission Glass, clear Dark green olive cap wrapper Cold pressed, unrened, contains 375
no imported olive oil, chemicals,
articial coloring or preservatives
S2 Spain Hojiblanca Glass, clear Olive green metal screw twist Obtained directly from olives and 500
solely by mechanical means
C1 Chile Picual Glass, opaque (black) Violet metal screw twist 500
I1 Italy Frantoio Glass, opaque (black) Dark green plastic screw twist Los Angeles Olive Oil Competition 500
gold award
I2 Italy Taggiasche (late harvest) Glass, wrapper: white paper 500
I3 Italy Taggiasche Glass, wrapper: golden paper 500
GBS Spain Generic brand (oils from several Glass, clear Olive green metal screw twist First cold pressed 250
countries)
GBI Italy Generic brand (oils from Italy) Plastic, clear Dark green plastic screw twist First cold pressed 2000

COOC California Olive Oil Council.

method) were applied to both data sets. The Chi-squared test was ap- the overall liking of the bottle may be associated with consumers
plied to compare the demographics, attitudes, and habits of the uncov- using olive oil for different purposes and selecting the size accordingly
ered segments in the preference and purchase intent mapping analyses. (Santosa & Guinard, 2011).
An internal preference map generated from the bottles' hedonic rat-
3. Results and discussion ings indicates that EVOO bottles from California were liked the most,
except U4, whereas imported EVOO bottles GBI, GBS and S2 were
3.1. Consumer characteristics liked the least (Fig. 1a, b). Furthermore, the gure shows some spread
of the consumers along the top right and bottom right quadrants of
A total of 102 consumers participated in the packaging study. Table 2 the biplot, suggesting preference segmentation that was then conrmed
shows the demographics of the consumers. Most participants were fe- by cluster analysis (Ward's method, Euclidean distance). Consumers in
male (72%), of Caucasian ethnicity (85%). Regarding their educational segment 1 (N = 59) and consumers in segment 2 (N = 43) are shown
level, 87% attended college (some college, 11%; bachelor degree, 35%; in Fig. 1a. The distribution of the oils is shown in Fig. 1b. Consumers in
master's degree, 31%; PhD, 10%). Marital status showed 31% as married segment 1 liked California EVOO bottles, except U4; the bottle and
or living with a partner with no children at home, 26% as married or image of U4 are closer to that of generic brands, which could explain
living with a partner with children at home, 21% living with roommates why consumers considered the design of this bottle as less appealing.
and 16% living alone. Income level was distributed evenly among the Consumers from segment 2, on the other hand, liked EVOO bottles
optional categories. I1, I2, I3 and C1 and disliked U4, S2, GBI and GBS. Region of origin is
Overall, the demographics of the consumers in the packaging study known to play a key role in consumer liking and association with high
were not different from those of the 110 consumers who participated quality (Dekhili & d'Hauteville, 2009; Dekhili, Sirieix, & Cohen, 2011;
in the blind tasting study (p N 0.05). Delgado & Guinard, 2011b; Krystallis & Ness, 2003; van der Lans, van
Itters, De Cicco, & Loseby, 2001); so given that consumers were rating
3.2. Hedonic ratings the overall package (bottle and label), it was not surprising to nd
that California products were liked the most. Buying local can be an im-
3.2.1. Degree of liking of the bottles and labels portant motivation, as shown for olive oil by Santosa and Guinard
Overall degree of liking of the EVOO bottles was signicantly corre- (2011), and more broadly by Chambers, Lobb, Butler, Harvey, and
lated to liking for the other packaginglabeling variables evaluated in Traill (2007), who found that there was a preference overall for pur-
the test (shape, container material, design of the labels, pictures on chasing local or national products.
the labels and colors of the bottle and labels), except for the size of the The mean overall liking ratings for both preference segments are
EVOO bottle. Liking of the shape of the bottle was signicantly correlated shown in Fig. 2a. Both segments seemed to like California EVOO bottles,
to liking of the material of the bottle, and liking for the label design was but they differed in their liking/disliking of the imported EVOO bottles,
signicantly correlated to that of the pictures in the label and to the with segment 2 liking the EVOO bottles from Italy (I1, I2, I3) and Chile
color of the label. The fact that liking for the size was not correlated to (C1) signicantly more than segment 1 (one-way ANOVA, p b 0.05).
C. Delgado et al. / Food Research International 54 (2013) 21122121 2115

Table 2 everyday vs. 37% for segment 2). There were no signicant differences
Demographics of the 102 consumers and of preference segments 1 and 2 in the packaging in their purchase frequency; though 44% of the consumers from seg-
study.
ment 1 purchased EVOO more than once a month vs. 27% for segment
Total Segment 1 Segment 2 2. Regarding their purchase habits, the majority of consumers bought
(N = 102) (N = 59) (N = 43) EVOO mainly at specialty stores (76% and 70% for segments 1 and 2, re-
Age spectively), traditional stores (54% and 51%) and farmers' markets (54%
2029 16 (16%) 7 (12%) 9 (21%) and 39%). Furthermore, after the price and EVOO category; the
3039 24 (23%) 12 (20%) 12 (28%)
size, container material and familiarity with the brand, besides
4049 16 (16%) 10 (17%) 6 (14%)
5059 22 (22%) 12 (20%) 10 (23%) country/region of origin and locally produced, were the most impor-
6069 20 (19%) 14 (24%) 6 (14%) tant factors affecting the olive oil purchased by consumers. These results
More than 70 4 (4%) 4 (7%) 0 (0%) agree with previous ndings in our consumer research with U.S. con-
CHISQ p = 0.280 NS sumers (Delgado & Guinard, 2011a; Santosa, Abdi, & Guinard, 2010;
Gender Santosa & Guinard, 2011). Moreover, the importance of the variables
Male 29 (28%) 15 (25%) 14 (33%) country/region of origin and locally produced to California con-
Female 73 (72%) 44 (75%) 29 (67%) sumers is clear in that 76% of people from segment 1 and 81% from
CHISQ p = 0.430 NS
segment 2 purchased EVOO from California. These numbers suggest
Ethnicity that the country or region of origin of the EVOO bottles used in this
African 1 (1%) 1 (2%) 0 (0%) study was the most important variable affecting the overall degree
Asian 8 (8%) 5 (8%) 3 (7%)
of liking of California consumers, and hence both segments reported a
Caucasian 87 (85%) 50 (85%) 37 (86%)
Hispanic 1 (1%) 0 (0%) 1 (2%) higher degree of liking for the California EVOO bottles (Fig. 1a,b). Further-
Native American 0 (0%) 0 (0%) 0 (0%) more, 72% of the consumers from segment 2 purchased EVOO from Italy
Pacic Islander 1 (1%) 1 (2%) 0 (0%) (vs. 49% from segment 1), and segment 2 accordingly showed a higher
Mixed 4 (4%) 2 (3%) 2 (5%)
degree of liking for EVOO bottles from Italy (Figs. 1a,b and 2a).
CHISQ p = 0.802 NS

Education level 3.2.2. Degree of liking of the oils tasted blind


High school diploma 2 (2%) 0 (0%) 2 (5%)
Fig. 1c and d shows the internal preference map and the distribution
Some college 11 (11%) 3 (5%) 8 (19%)
Bachelor 36 (35%) 25 (42%) 11 (26%)
of the 18 EVOOs derived from the blind tasting hedonic ratings. Three
Master 32 (31%) 16 (27%) 16 (37%) segments were identied after using cluster analysis (Ward's method,
PhD 10 (10%) 8 (14%) 2 (5%) Euclidean distance). Segment 1 (N = 47) showed higher preferences
Professional 12 (11%) 7 (12%) 4 (9%) for oils U4, U3, U9, U1, U7, I1 and GBI; this segment preferred mostly
CHISQ p = 0.036
California oils. Segment 2 (N = 25) preferred a few oils mostly the
Living with three from Italy, I2, I3, I5, U5 and GBI. From a descriptive analysis of
Alone 16 (16%) 9 (15%) 7 (16%) these oils (Delgado & Guinard, 2011b), we know that these 5 oils shared
Roommates 21 (21%) 8 (14%) 13 (30%)
a similar sensory prole, and were not very bitter or pungent. The pref-
Partner or spouse 32 (31%) 18 (30%) 14 (33%)
Children 4 (4%) 3 (5%) 1 (2%) erences of this segment may indicate that these were relatively new
Spouse and children 27 (26%) 20 (34%) 7 (17%) consumers of EVOO who had not previously been exposed to these
Extended family 2 (2%) 1 (2%) 1 (2%) attributes in EVOO and as a consequence, did not like the other oils in
CHISQ p = 0.231 NS the set (which they would have found too bitter or pungent). The
Income third segment showed a preference for both local and imported oils
Under $30.000 21 (20%) 9 (15%) 12 (28%) (GBS, I2, I3, GBI, I1, U11, U7, U1, U9, U3, and U4). These consumers
$30.000$60.000 23 (22%) 13 (22%) 10 (23%) also agreed with the consumers in segment 2 in their rejection of bitter
$60.000$100.000 21 (21%) 13 (22%) 8 (19%)
$100.000$150.000 16 (16%) 9 (15%) 7 (16%)
and pungent oils such as U10 and U8. In spite of being two of the least
More than $150.000 11 (11%) 6 (11%) 5 (12%) liked products in the blind tasting, probably because of their higher
Prefer not to report 10 (10%) 9 (15%) 1 (2%) bitterness and pungency (Delgado & Guinard, 2011b), U8 and U10
CHISQ p = 0.259 NS received some of the highest hedonic scores in the packaging study.
Occupation Overall the average scores for the EVOOs in the blind tasting study
Construction, extraction, 1 (1%) 0 (0%) 1 (2%) were lower than those from the packaging study. This conrms previ-
maintenance and repair ous ndings that consumers tend to give higher liking scores when
Farming, shing and forestry 1 (1%) 0 (0%) 1 (2%)
the brand is shown, and they are familiar with it, than in blind condi-
Sales and ofce 6 (6%) 2 (3%) 4 (9%)
Service-related 2 (2%) 2 (3%) 0 (0%) tions for the same product (Caporale et al., 2006; Guinard et al., 2000).
Management, professional 22 (22%) 13 (22%) 9 (21%) In the packaging study, consumers only rated the product without tast-
and related occupations ing and the values were higher for the packaged product than for the oil
Production, transportation 1 (1%) 1 (2%) 0 (0%) tasted alone. Fig. 2 conrms that consumers tended to be more critical
and related
Student 19 (18%) 8 (14%) 11 (26%)
in blind conditions than in the informed rating of the products.
Unemployed 4 (4%) 2 (3%) 2 (5%)
Other (education, research, art) 28 (27%) 18 (30%) 10 (23%) 3.3. Price expectations
Retired 18 (17%) 13 (22%) 5 (12%)
CHISQ p = 0.315 NS
3.3.1. Price expectations based on the bottle and label
NS, not signicant p N 0.05. Consumers in the packaging study were also asked to rate the price
Signicant difference at p b 0.05.
range they expected to pay for each EVOO bottle. Both segments
showed the same price expectation trends with no signicant difference
Demographics, consumption and purchasing habits for these two in the mean price they expected to pay. Overall, consumers were willing
clusters are summarized in Tables 2 and 3. Although consumers were to pay between $5 and $10 for a bottle of GBS (250 mL), $10 to $15 for a
screened based on their EVOO consumption before starting the study, bottle of GB1 (2L), S2 (500mL), U2 (250mL), U3(500 mL), U4 (375mL),
there were signicant differences in their frequency of consumption U6 (250mL), U8 (250mL), U11 (375mL), U7 (500mL) and U9 (500mL);
(p b 0.05), which was higher for segment 1 (i.e. 57% consumed EVOO $15 to $20 for I2 (500 mL), I3 (500 mL), I1 (500 mL), U1 (250 mL), C1
2116 C. Delgado et al. / Food Research International 54 (2013) 21122121

Fig. 1. Internal Preference Map (IPM) including preference segmentation, for the packaging study (a and b) and for the blind tasting study (c and d). a. Segment 1 in red/dark dots; segment
2 in blue/light dots. c. Segment 1 in red/dark dots; segment 2 in blue/light dots; segment 3 in black dots.

(500 mL), U10 (375 mL); and $20 to $25 for U5 (375 mL). U5 and U10 $10 range. This indicates that when all consumers are exposed to the oil,
were the most stylish bottles and both were packaged inside a carton they are not willing to pay much money for it (less than $10 a bottle),
box with logos on their designs. As expected, both generic brands re- but once they are exposed to the whole product in its package, they
ceived lesser price expectations (Fig. 3). The volume of each EVOO are willing to pay more. Packaging attractiveness is mentioned as one
may also have been considered by consumers when assigning price. critical factor when selecting olive oil (Krystallis & Ness, 2003). In
Mueller Loose and Szolnoki (2012) indicated an effect of price in the most cases, price is associated with quality; and the greater the percep-
market predictions of wine sales. They found the largest price differ- tion of the value, the greater the willingness of the consumer to pay
ences to relate to region of origin followed by label design type, grape more (Creyer & Ross, 1997).
variety, label color and label information.
3.4. Purchase intent
3.3.2. Price expectations generated in the blind tasting study
In the blind tasting of the oils, consumers were asked to indicate how 3.4.1. Purchase intent for the bottles and labels
much they would be willing to pay for a 375 mL bottle of the oil they Purchase intent for the EVOO bottles in the packaging study was
had just tasted. The scale used differed slightly as it offered a $0 option. only signicantly correlated with overall liking. Similarly, we have
The rest of the scale was the same as in the packaging study (see shown a positive correlation between overall liking and purchase intent
Section 2.3.1). Fig. 3 shows that, for the most part, the price consumers when olive oils were tasted blind (Delgado & Guinard, 2011a). On the
were willing to pay was relatively even across the oils, with oils U10, U8, contrary, no relationship was found between purchase intent and
U2, S2, and U6 in the $0 to $5 range, and the rest of the oils in the $5 to the price expected for the EVOO bottles, although around 80% of the
C. Delgado et al. / Food Research International 54 (2013) 21122121 2117

Fig. 2. Mean ratings of the overall liking for each EVOO by preference segments (a) packaging study, (b) blind tasting study. Within a segment, the overall liking means that share a com-
mon letter do not differ signicantly from one another (p b 0.05).

participants selected price as an important aspect affecting their extra Consumer purchase intentions for the EVOO bottles were relatively
virgin olive oil purchase. This suggests that packaging characteristics are homogeneous, following very similar trends to the overall liking of the
the most important determinants for any EVOO bottle's rst purchase. bottles (Fig. 4). Yet, a cluster analysis of the purchase intent ratings
2118 C. Delgado et al. / Food Research International 54 (2013) 21122121

Table 3 revealed a two-cluster solution, with 63 and 39 consumers in each clus-


Summary of olive oil consumption and purchasing habits for preference segments 1 and 2 ter, respectively. Purchase intentions for the California EVOO bottles
in the packaging study.
were similar in both consumer segments, although consumers in seg-
Segment 1 (N=59) Segment 2 (N=43) ment 2 showed a higher purchase intent for clear bottles, such as U2,
Kind of olive oila U8 and U11 (p b 0.05), whereas purchase intent in segment 1 was
EVOO 59 (100%) 40 (93%) greater for the dark bottles and those inside a box (especially U1, U6
Not EVOO (Virgin, light, pure) 11 (19%) 8 (19%) and U10, p b 0.05).
Flavored OO 15 (25%) 9 (21%)
The two consumer segments differed mainly in the purchase intent
Don't know, not sure 0 (0%) 1 (2%)
CHISQ p = 0.664 NS for imported EVOO bottles, with segment 1 showing a higher purchase
intent score for EVOO bottles from Italy (I1, I2 and I3) and Chile (C1)
EVOO consumption
(pb 0.05), all of them dark or wrapped bottles, while generic EVOO bot-
Everyday 34 (57%) 16 (37%)
23 times per week 14 (24%) 22 (51%) tles (GBS and GBI) received higher intents from segment 2 (p b 0.05).
Once a week 8 (14%) 2 (5%)
Twice a month 3 (5%) 0 (0%) 3.4.2. Purchase intent based on blind tasting
Once a month 0 (0%) 1 (2%) Fig. 5 shows the relationship between the purchase intent based on
CHISQ p = 0.011
the evaluation of the EVOO bottles and labels (no tasting) and that
EVOO purchasing based on the actual tasting of the oils in the blind tasting study. Because
Once a month or more 25 (44%) 11 (27%) purchase intent was signicantly correlated to overall liking (p b 0.05)
Once every 23 months 21 (37%) 15 (36%)
(Delgado & Guinard, 2011a), we chose to show the three different liking
Once every 4 months 6 (10%) 11 (27%)
Once every 6 months 2 (3%) 2 (5%) segments (from the blind tasting study) in this comparison. We found
Once a year or less 3 (5%) 2 (5%) that for all three segments, purchase intent for the bottles was inversely
CHISQ p = 0.227 NS related to purchase intent of the oils as tasted blind. However, only in
Place of purchasea the case of the rst segment was the correlation signicant.
Traditional store 32 (54%) 22 (51%)
Specialty store 45 (76%) 30 (70%) 3.5. Sensory and nutritional expectations generated by the bottles
Farmers' market 32 (54%) 17 (39%)
and labels
Wholesale 22 (37%) 14 (33%)
Discount retailer 6 (10%) 8 (19%)
Olive oil farmer 18 (30%) 9 (21%) Overall avor expectations for the EVOOs based on the evaluation of
Gourmet store 10 (17%) 6 (14%) the bottles and labels were signicantly correlated to the expected
Other (Winery, UCD bookstore) 6 (10%) 5 (12%) intensity of the sensory attributes included in the questionnaire,
CHISQ p = 0.888 NS
i.e., fruitiness, bitterness and pungency. This is logical and expected.
Origina With regard to what could have driven those expectations, Caporale
Italy 29 (49%) 31 (72%) et al. (2006) reported that information about the olive cultivar affected
Spain 15 (25%) 13 (30%)
consumer expectations for bitterness and pungency, and many of the
Greece 10 (17%) 5 (11%)
Portugal 4 (7%) 1 (2%) labels, particularly the California ones, named the cultivar(s).
France 2 (3%) 1 (2%) Preference segments 1 and 2 (that were derived from the packaging
Tunisia 2 (3%) 1 (2%) hedonic ratings) also differed in their avor expectations (Fig. 6). Both
Turkey 1 (2%) 0 (0%)
segments expected the same overall avor for the California EVOOs,
Morocco 3 (5%) 1 (2%)
Israel 5 (8%) 1 (2%)
and also gave the same scores for the EVOO from Chile (C1), Spain (S2)
Australia 0 (0%) 0 (0%) and the generic brands (GBS and GBI) (one-way ANOVA p N 0.05). The
South Africa 1 (2%) 0 (0%) preference segments differed in the overall avor they expected in the
Chile 0 (0%) 1 (2%) EVOO bottles from Italy (I1, I2 and I3). It was signicantly higher for seg-
California 45 (76%) 35 (81%)
ment 2 (one-way ANOVA p b 0.05). Furthermore, GBI, GBS, S2 and U11
USA (not California) 4 (7%) 2 (5%)
Don't know, not sure 8 (14%) 3 (7%) bottles were expected to be mild EVOOs for both segments, whereas
CHISQ p = 0.671 NS C1, U1, U5, U6 and U10 EVOOs were expected to possess a strong avor.
Consumers were asked to rate their nutritional expectations of the
Factors affect purchasea
Extra Virgin Olive Oil category 50 (85%) 33 (77%) EVOOs according to the packaging and labeling compared to a typical
Monovarietal vs. blend 5 (8%) 10 (23%) EVOO. Nutritional properties of an EVOO depend mainly on the anti-
Country/region of origin 34 (58%) 26 (60%) oxidant content (phenolic compounds and tocopherols) and fatty
Cooking suggestions 12 (20%) 4 (9%) acid composition (Bendini et al., 2007; Harwood & Aparicio, 2000;
Flavor description 29 (49%) 19 (44%)
Nutritional information 8 (14%) 3 (7%)
Servili & Montedoro, 2002), which cannot be estimated by extrinsic
Harvest date 14 (24%) 12 (28%) factors, except if this specic information is included in the nutri-
Best before date 23 (39%) 8 (19%) tional facts of the product. Only EVOO bottles C1, GBS, GBI, I3, U3,
Container material 40 (68%) 29 (67%) U9 and U11 showed some nutritional facts, including the fatty acid
Size 40 (68%) 29 (67%)
composition of the EVOO (saturated fats, monounsaturated fats and
Organic product 29 (49%) 20 (46%)
Locally produced 33 (56%) 20 (46%) polyunsaturated fats), which was similar in all those bottles. So in re-
Award winning oil 21 (36%) 17 (39%) ality, there were no useful extrinsic data to estimate and differentiate
Familiarity with the brand 33 (56%) 26 (60%) the nutritional properties of the EVOOs.
Relative or friends' recommendation 19 (32%) 12 (28%) Twenty-four consumers gave a rating of 3 (about the same nutri-
Price 47 (80%) 37 (86%)
CHISQ p = 0.736 NS
tional properties) for all the EVOO bottles, hence extrinsic factors
(packaging, labeling or available information in the labels) did not affect
NS, not signicant p N 0.05.
Signicant difference at p b 0.05.
their expectation of the EVOOs' nutritional properties. These consumers
a
Consumers were asked to check all that apply, so the percentages add up to more likely knew that it is not possible to establish the nutritional properties
than 100. of an EVOO just from the type of packaging (bottle or label). Neverthe-
less, about 77% of the consumers had the package and label affect their
nutritional expectations. The mean ratings for the nutritional properties
C. Delgado et al. / Food Research International 54 (2013) 21122121 2119

Fig. 3. Mean ratings of the price consumers were willing to pay for the EVOOs in the blind tasting study (N = 110) versus the expected purchase price for the EVOO bottles in the packaging
study (N = 102).

expected in each EVOO by those consumers are shown in Fig. 7. value for this product. There was no signicance difference between
Imported EVOO bottles GBI, GBS and S2 were expected to possess less the mean ratings of the two groups.
nutritional properties than EVOOs from California, Italy and Chile; the
same observation was made for the overall avor expected by con- 4. Conclusions
sumers. On the other hand, consumers gave the highest rating to bottle
U10 for expected nutritional properties. This extra virgin olive oil We uncovered two preference segments in the evaluation of the
was certied as organic by the California Certied Organic Farmers bottles and labels (packaging study), both of which liked the California
(CCOF) and the USDA a fact that was printed clearly on the label EVOOs, but differed in their liking of the imported oils. Consumers
and probably explains the greater score given by 77% of the consumers. in segment 1 liked mostly California EVOO bottles, while consumers
The reason behind this may be that some consumers have expectations in segment 2 also liked some of the imported oils, particularly those
of higher nutritional value for organic products. In terms of the sensory from Italy.
properties, this EVOO had strong bitterness and pungency. Since poly- There was some discrepancy between the overall liking of the oils as
phenols have been associated with pungency and bitterness (Bendini packaged (packaging, labeling, color, etc.) and the overall liking of the
et al., 2007), this would support the expectation of greater nutritional oils as tasted, in blind conditions. The overall appearance of the product

Fig. 4. Internal purchase intent map showing (a) individual consumers and purchase intent segmentation (N = 63 in segment 1, N = 39 in segment 2) and (b) the EVOOs, from the
purchase intent ratings in the packaging study.
2120 C. Delgado et al. / Food Research International 54 (2013) 21122121

Fig. 5. Mean ratings of purchase intention by segment in the blind tasting study (N = 110) versus the mean ratings of purchase intention in the packaging study (N = 102).

Fig. 6. Mean ratings of expected overall avor intensity for the packaging study. Within a segment, the expected overall means that share a common letter do not differ signicantly from
one another (p b 0.05).

plays an important role in creating consumer expectations. Some pro- if an EVOO is very bitter or pungent, it may be rejected for not meeting
ducers are attracting consumers with the packaging of their product. the expectations of consumers who are relatively new to olive oil, hence
However, when consumers taste the product in blind conditions those the need for further education of consumers and for exposing them to
factors are not there for the consumer to create any expectation; and the entire range of olive oils on the market. Region of origin was the var-
iable that showed the biggest impact on the overall liking of the bottles
and labels in the packaging study. California EVOOs were signicantly
preferred over most of the imported oils for this group of consumers.
Consumers expected to pay more (up to $30) when they were pre-
sented only with the packaging of the EVOO. In the blind tasting, however,
consumers were generally not willing to pay more than $10 for a 375-mL
bottle of oil, and their hedonic ratings were lower overall.
Purchase intent followed similar trends to overall liking in the pack-
aging study, with two uncovered clusters that differed in their inten-
tions to purchase the oils based on the bottle color (clear or dark), and
the origin of the oils (imported vs. domestic).
Future research should broaden the population of consumers sampled
to include representation from all regions of the US, and should evaluate
how the rapid growth in both production and consumption of EVOO in
the US, and the various campaigns to educate consumers about the prop-
erties and benets of EVOO are impacting consumer perceptions, habits
and attitudes regarding extra virgin olive oil.

5. Acknowledgments
Fig. 7. Mean ratings of expected nutritional properties in the EVOO bottles in the packaging
study (data obtained from 77% of consumers, who gave different scores across samples). The
expected nutritional properties means that share a common letter do not differ signicantly The authors thank Paul Vossen and Dr. Hildegarde Heymann for their
from one another (p b 0.05). helpful suggestions and assistance with the SAS codes for the multivariate
C. Delgado et al. / Food Research International 54 (2013) 21122121 2121

analyses, and gratefully acknowledge olive oil donations for these studies Fotopoulos, C., & Krystallis, A. (2001). Are quality labels a real marketing advantage? A
conjoint application on Greek PDO protected olive oil. Journal of International Food
by Corto Olive; UC Davis Olive Center; Corti Brothers; Olive Press; Enrique & Agribusiness Marketing, 12(1), 122.
Escudero, Hojiblanca; Golden Hill; Jovia Groves; Dos Colinas; McEvoy Guinard, J. X., Uotani, B., Mazzucchelli, R., Taguchi, A., Masuoka, S., & Fujino, S. (2000).
Olive Oil; Veronica Foods; Calolea; and Sciabica's Family. Consumer testing of commercial lager beers in blind versus informed conditions:
Relation with descriptive analysis and expert quality ratings. Journal of the Institute
of Brewing, 106(1), 1119.
Harwood, J., & Aparicio, R. (2000). Handbook on olive oil. Analysis and properties. Gaithers-
References burg, Maryland: Aspen Publishers, Inc.
Jaeger, S. R. (2006). Non-sensory factors in sensory science research. Food Quality and
Ares, G., & Deliza, R. (2010). Studying the inuence of package shape and colour on con- Preference, 17, 132144.
sumer expectations of milk desserts using word association and conjoint analysis. Kanavouras, A., Hernandez-Munoz, P., & Coutelieris, F. (2006). Packaging of olive
Food Quality and Preference, 21(8), 930937. oil: Quality issues and shelf life predictions. Food Reviews International, 22(4),
Bendini, A., Cerretani, L., Carrasco-Pancorbo, A., Gomez-Caravaca, A.M., Segura-Carretero, 381404.
A., Fernandez-Gutierrez, A., et al. (2007). Phenolic molecules in virgin olive oils: a sur- Krystallis, A., & Ness, M. (2003). Motivational and cognitive structures of Greek consumers
vey of their sensory properties, health effects, antioxidant activity and analytical in the purchase of quality food products. Journal of International Consumer Marketing,
methods. An overview of the last decade. Molecules, 12(8), 16791719. 16(2), 736.
Bialkova, S., & van Trijp, H. (2010). What determines consumer attention to nutrition labels? Martnez, M. G., Aragons, Z., & Poole, N. (2002). A repositioning strategy for olive oil in
Food Quality and Preference, 21(8), 10421051. the UK market. Agribusiness, 18(2), 163180.
Bower, J. A., & Turner, L. (2001). Effect of liking, brand name and price on purchase intention Mtimet, N., Zaibet, L., Zairi, C., & Hzami, H. (2013). Marketing olive oil products in the
for branded, own label and economy line crisp snack foods. Journal of Sensory Studies, Tunisian local market: The importance of quality attributes and consumers' behavior.
16(1), 95116. Journal of International Food & Agribusiness Marketing, 25(2), 134145.
Caporale, G., & Monteleone, E. (2004). Inuence of information about manufacturing pro- Mueller Loose, S., & Szolnoki, G. (2012). Market price differentials for food packaging
cess on beer acceptability. Food Quality and Preference, 15(3), 271278. characteristics. Food Quality and Preference, 25(2), 171182.
Caporale, G., Policastro, S., Carlucci, A., & Monteleone, E. (2006). Consumer expectations for Mueller, S., & Szolnoki, G. (2010). The relative inuence of packaging, labelling, branding
sensory properties in virgin olive oils. Food Quality and Preference, 17(12), 116125. and sensory attributes on liking and purchase intent: Consumers differ in their re-
Chambers, S., Lobb, A., Butler, L., Harvey, K., & Traill, W. B. (2007). Local, national and sponsiveness. Food Quality and Preference, 21(7), 774783.
imported foods: A qualitative study. Appetite, 49(1), 208213. Peryam, D. R., & Pilgrim, F. J. (1957). Hedonic scale method of measuring food preferences.
Creyer, E. H., & Ross, W. T. (1997). Tradeoffs between price and quality: How a value Food Technology, 11, 914.
index affects. Journal of Consumer Affairs, 31(2), 280. Sandalidou, E., & Baourakis, G. (2002). Customers' perspectives on the quality of organic
Dekhili, S., & d'Hauteville, F. (2009). Effect of the region of origin on the perceived quality olive oil in Greece. A satisfaction evaluation approach. British Food Journal, 104(3/4/5),
of olive oil: An experimental approach using a control group. Food Quality and 391406.
Preference, 20(7), 525532. Santosa, M. (2010). Analysis of sensory and non-sensory factors mitigating consumer behav-
Dekhili, S., Sirieix, L., & Cohen, E. (2011). How consumers choose olive oil: The importance ior: A case study with extra virgin olive oil. Unpublished Ph.D. Dissertation, University of
of origin cues. Food Quality and Preference, 22(8), 757762. California, Davis, Davis.
Delgado, C., & Guinard, J. -X. (2011a). How do consumer hedonic ratings for extra virgin Santosa, M., Abdi, H., & Guinard, J. -X. (2010). A modied sorting task to investigate con-
olive oil relate to quality ratings by experts and descriptive analysis ratings? Food sumer perceptions of extra virgin olive oils. Food Quality and Preference, 21(7),
Quality and Preference, 22(2), 213225. 881892.
Delgado, C., & Guinard, J. -X. (2011b). Sensory properties of Californian and imported Santosa, M., & Guinard, J. -X. (2011). Means-end chains analysis of extra virgin olive oil
extra virgin olive oils. Journal of Food Science, 76(3), S170S176. purchase and consumption behavior. Food Quality and Preference, 22(3), 304316.
Deliza, R., & MacFie, H. J. H. (1996). The generation of sensory expectation by external Servili, M., & Montedoro, G. (2002). Contribution of phenolic compounds to virgin
cues and its effect on sensory perception and hedonic ratings: A review. Journal of olive oil quality. European Journal of Lipid Science and Technology, 104(910),
Sensory Studies, 11(2), 103128. 602613.
Deliza, R., Rosenthal, A., Abadio, F. B.D., Silva, C. H. O., & Castillo, C. (2005). Application of van der Lans, I. A., van Itters, K., De Cicco, A., & Loseby, M. (2001). The role of the region of
high pressure technology in the fruit juice processing: benets perceived by con- origin and EU certicates of origin in consumer evaluation of food products. European
sumers. Journal of Food Engineering, 67(12), 241246. Review of Agricultural Economics, 28(4), 451477.
Enneking, U., Neumann, C., & Henneberg, S. (2007). How important intrinsic and extrinsic Zampounis, V. (2006). Olive oil in the world market. In D. Boskou (Ed.), Olive oil chemistry
product attributes affect purchase decision. Food Quality and Preference, 18(1), 133138. and technology (pp. 2139). AOCS Press.

You might also like