Consumer Attitudes and Preferences for Fresh

Market Tomatoes
A.E. Oltman, S.M. Jervis, and M.A. Drake

Abstract: This study established attractive attributes and consumer desires for fresh tomatoes. Three focus groups
(n = 28 participants) were conducted to explore how consumers perceived tomatoes, including how they purchased
and consumed them. Subsequently, an Adaptive Choice Based Conjoint (ACBC) survey was conducted to understand
consumer preferences toward traditional tomatoes. The ACBC survey with Kano questions (n = 1037 consumers in
Raleigh, NC) explored the importance of color, firmness, size, skin, texture, interior, seed presence, flavor, and health
benefits. The most important tomato attribute was color, then juice when sliced, followed by size, followed by seed
presence, which was at parity with firmness. An attractive tomato was red, firm, medium/small sized, crisp, meaty,
juicy, flavorful, and with few seeds. Deviations from these features resulted in a tomato that was rejected by consumers.
Segmentations of consumers were determined by patterns in utility scores. External attributes were the main drivers of
tomato liking, but different groups of tomato consumers exist with distinct preferences for juiciness, firmness, flavor, and
health benefits.

Keywords: conjoint, consumer preference, focus group Kano, tomato

Practical Application: Conjoint analysis is a research technique that collects a large amount of data from consumers in
a format designed to be reflective of a real life market setting and can be combined with qualitative insight from focus
groups to gain information on consumer consumption and purchase behaviors. This study established that the most
important fresh tomato attributes were color, amount of juice when sliced, and size. Distinct consumer clusters were
differentiated by preference for color/appearance, juiciness and firm texture. Tomato growers can utilize the results to
target attributes that drive consumer choice for fresh tomatoes.

Introduction preferred red color and sweet taste and one that preferred acidity

S: Sensory & Food

In 2011, tomatoes grown in the US had a gross production and firm texture (Pagliarini and others 2001).
value of 11.74 billion USD (FAOSTAT 2011). As a widely pro- Texture is also an important aspect of consumer perception

Quality
duced and consumed crop, sensory attributes of fresh tomatoes are of fresh tomatoes (Causse and others 2003; Serrano-Megias and
important, and lack of characteristic taste and flavor are frequent Lopez-Nicolas 2006). Texture traits previously established in-
complaints of fresh tomato consumers (Bruhn and others 1991). clude flesh firmness, mealiness, meltiness, crispness, and juiciness
Tomatoes (Solanum lycoperiscum) have been characterized as hav- (Harker and others 1997; Redgwell and Fischer 2002; Szczes-
ing a sweet–sour taste with a complex mix of aromatics such as niak 2002). Previous research has focused on creating tomatoes
fruity/floral and green notes (Baldwin and others 2008). Hong- with greater firmness in order to have greater disease resistance
soonern and Chambers (2008) identified a descriptive language for and longer shelf life (Hongsoongnern and Chambers 2008). The
fresh tomatoes and reported fruity, ripeness, sweetness, sourness, amount of gel and seeds within locules are unique aspects of
and bitterness as well as green/viney, musty/earthy and fermented tomato perceptions (Chaı̈b and others 2007). Finally, visual ap-
as attributes applicable to fresh tomatoes. In other studies with pearance of tomatoes is of utmost importance to consumer pur-
trained panelists, sweet taste and fruity flavor of tomatoes were chase decision, as other factors such as flavor and texture are not
correlated (Malundo and others 2007; Baldwin and others 2008). evident during purchase. Wolters and Gemert (1990) indicated
Additionally, a tomato perceived as having good overall flavor was that color and size were the most important visual attributes of
rated high in sweet taste, “tomato-like flavor” and fruity flavor tomatoes. Other extrinsic attributes such as health/nutrition and
with low intensities of sourness, bite and “green tomato flavor” growing conditions may also influence consumer purchase.
by a trained panel (Tandon 2006). Room ripened tomatoes had Consumer decisions for food purchase are influenced by various
lower flavor quality than vine ripened tomatoes (Bisogni and oth- factors, including familiarity with ingredients and manufacturers,
ers 1976; Stevens and others 1977), which could be due to lower taste, price, and the perceived product health benefits (Cardello
sugar content (Davies 1966; Kader and others 1977; Jones and and Schutz 2003). With many elements to weigh, choosing a fresh
Scott 1983). In a previous consumer study with cherry tomatoes, tomato becomes a complex task. Tomato varieties and consumer
internal preference mapping revealed 2 consumer clusters: one that preferences may vary vastly and by characterizing both tomato
attributes and consumer drivers for purchase, ideals can be pin-
pointed and optimized. The qualitative information gained in a
MS 20140375 Submitted 3/7/2014, Accepted 8/7/2014. The authors are with the
focus group allows the researcher to study consumer behavior and
Food Bioprocessing and Nutrition Sciences Dept., North Carolina State Univ., Raleigh,
NC 27695 U.S.A. Direct inquiries to author Drake (E-mail: [email protected]). product usage (Bellenger and others 1976). Focus groups have
been used on a number of foods including mungbean noodles

C 2014 Institute of Food Technologists

 R

doi: 10.1111/1750-3841.12638 Vol. 79, Nr. 10, 2014 r Journal of Food Science S2091
Further reproduction without permission is prohibited
 Fresh market tomatoes . . .

Table 1–Moderator’s guide for fresh Tomatoes.

Moderator Introduction: purpose of focus group

Respondent introductions: introduce yourself and share one dish you like to cook.
Focus area 1: tomatoes you and your family eat:
• How do you use fresh tomatoes?
• Is ease of cutting a tomato important to you? (probe frequency, seasonal)
• How important is shelf life?
• Do you process tomatoes yourself? (blending, juicing, peeling). Would you be more likely to buy a tomato product (such as sauce or paste) than
process yourself?
• What makes you want a fresh tomato over canned?
• How does buying fresh tomatoes make you feel? (probe emotions)
Focus area 2: visual/extrinsic factors (Allow participants to touch/feel whole tomatoes)
• What are the most important factors when you select a tomato from the store?
• What is the ideal tomato color? Show shades of tomato color
• Good skin: Is it thin? Thick? Smooth?
• Size? How important?
• Firmness: what do you want?
• Can you predict what is inside? How?
• Types. How do you pick? Do you only use one type primarily?
• Other than appearance, are there other things that you look for when you purchase? (looking for price, local, organic as well as smell/touch feel)
• What would steer you away?
Focus area 3: texture (prompt participants to touch and reference the tomatoes on the table as well as sliced tomatoes in their individual cups)
• Are there any descriptors that come to mind when describing your ideal tomato texture?
• Do you like tomatoes that are firm or soft? (Probe why)
• Juiciness: more or less?
• Flavors associated with juicy/nonjuicy
• Texture of the inside: what’s good and bad?
Focus area 4: flavor (prompt participants to taste their tomatoes)
• What are some descriptors that come to mind when you think “tomato flavor?”
• Which contributes more to a “good” tomato: flavor or texture? Why?
Focus area 5: nutrition and miscellaneous
• Nutrients in tomatoes
• What is attractive about nutrition?
• What nutritional factors are important for a tomato? Low calorie? Fiber? Antioxidants?
• Is organic important to you? Is locally grown important to you?
• How important are organic and locally grown compared to price?

(Galvez and Resurreccion 1992), steak (Grunert 1997), butter group was approximately 1.5 h and panelists were compensated
S: Sensory & Food

(Krausse and others 2007), protein beverages (Childs and others for their participation with $40 gift cards.
2010), and chocolate milk (Thompson and others 2007). Conjoint Diverse types of fresh tomatoes were purchased for use in fo-
Quality

analysis is a technique that decomposes a product into different cus groups. Some tomatoes were cut into slices and presented in
attributes to determine which of those attributes contribute to 473-mL styrofoam bowls with plastic wrap (for tasting and visual-
liking. Consumer preference is estimated by evaluating all possi- ization) and others remained whole to visualize extrinsic factors.
ble combinations of preselected, evaluated product attributes (van Focus groups were videotaped and tape-recorded for subsequent
Kleef and others 2005). Conjoint analysis has also been widely reference, and 2 observers took notes of key points mentioned by
applied to determine consumer preferences for extrinsic and in- participants. Key points (those issues mentioned by two-thirds or
trinsic product attributes (Childs and others 2009; Chung and more of the participants) from focus groups were recorded.
others 2011; Jervis and others 2012; Kim and others 2013). To
our knowledge, studies have not addressed the role that both ex-
trinsic and intrinsic properties contribute to consumer perception Conjoint survey and kano questions
of tomatoes. The objective of this study was to determine the at- An online survey was created using SSI Web (Sawtooth Soft-
tributes of fresh market tomatoes that influence consumer purchase ware version 7.0.22, Orem, Utah, U.S.A.). Prior to starting the
decisions. Focus groups and a conjoint analysis survey followed by survey, a series of knowledge and awareness questions about fresh
consumer segmentation were utilized. tomatoes were asked. These questions were aimed at understand-
ing how aware consumers were of terminology used in describing
Materials and Methods tomatoes. A short demographic section was included to eliminate
participants who did not consume tomatoes monthly. The at-
Focus groups tributes and levels for the conjoint analysis were developed based
Three focus groups were conducted to determine how con- on information collected in the focus groups (Table 2). These
sumers used tomatoes and to identify attractive tomato qualities. attributes and levels were developed with qualitative level defi-
Panelists (n = 28) (females, 19 to 45 yr) that consumed toma- nitions rather than quantitative definitions since the goal was to
toes at least once a week and were the primary shoppers of the present levels that would be understood by consumers. Adaptive
household participated. All subjects were recruited through email Choice Based Conjoint (ACBC) is a type of conjoint analysis that
listservs to an online database of more than 5000 consumers in the is a hybrid design of Choice Based Conjoint (CBC) and Adaptive
Raleigh/Durham, NC area maintained by the Sensory Service Choice Analysis (ACA) (Orme 2010; Rao 2010). The ACBC,
Center at North Carolina State Univ. A moderator facilitated the which incorporates the benefits of both ACA and CBC, has been
discussion using a planned discussion guide (Table 1). Each focus shown to engage respondents more (Jervis and others 2012). The

S2092 Journal of Food Science r Vol. 79, Nr. 10, 2014

Fresh market tomatoes . . .

Table 2–Attributes and levels used for conjoint analysis. server. Participants (n = 1037) were recruited through email list-
Attribute Levels servs to an online database of more than 5000 consumers in the
Raleigh/Durham, NC area maintained by North Carolina State
Color Red Univ. Participants were entered into a drawing for 1 of 20 $25
Dark red
Light red TargetR
gift cards drawings.
Orange
Yellow Statistical analysis
Green
Exterior Firmness Firm
Individual utility scores were extracted by hierarchical Bayesian
Soft (HB) estimation and rescaled using a zero-centered difference
Size Small (lemon sized) method (Childs and Drake 2009). The zero-centered difference
Medium (baseball sized) method was used to standardize utility scores for easy interpreta-
Large (grapefruit sized) tion. A one-way analysis of variance with Fisher’s least significant
Skin Thin skinned
Thick skinned difference was applied to utility scores. Cluster analysis of utility
Texture of slice Crisp scores was performed with XLSTAT version 2012.6.06 (Addin-
Soft soft, Paris, France) using Euclidean distances and Wards linkage to
Interior Juicy categorize similar respondents into groups, and chi-square analy-
Meaty
Seed presence No seeds
sis was applied to determine any demographic differences among
Few seeds clusters. Principal component analysis (PCA) (XLSTAT) was also
Many seeds conducted to see how attributes and clusters were characterized.
Flavor Bold tomato flavor Kano questions were evaluated according to the model proposed
Mild tomato flavor by Kano and others (1984).
Health benefits Lycopene
Vitamin C
Fiber Results
Vitamin A
None Focus groups
Juice Small amount of juice Participants indicated their various usages of fresh tomatoes,
Large amount of juice
No juice
among which, salads, sandwiches, home-made salsa, chili, and
soups were the most frequently mentioned. The majority of par-
ticipants used fresh tomatoes for salads and sandwiches. Participants
used both fresh and processed tomatoes. Almost all participants
ACBC survey was designed with 1 build-your-own (BYO) task indicated that they liked the concept of freshness and healthiness
followed by 10 choice tasks, with 3 product concepts per task with with fresh tomatoes and considered purchasing a fresh tomato as
the possible responses of “a possibility” or “won’t work for me” for an enjoyable feeling that would give them a higher quality of life.
each product concept. A minimum of 2 and a maximum of 3 at- More than half of participants used canned tomatoes more often

S: Sensory & Food

tributes varied from the BYO selections for each product concept. than fresh in preparing warm tomato recipes such as soups or
Each product concept was a random generation of levels within sauces. In all groups, at least 1 participant stated that they did not

Quality
each attribute, with each attribute represented in all 10 choice like tomatoes that were prepackaged on foam trays—it was impor-
tasks. Five unacceptable questions and 4 must-have questions were tant that they were able to handle the tomato and evaluate it for
built in through the survey. The screening task was followed by a themselves before purchasing. Participants preferred to touch the
10-question choice task tournament section. A maximum of 20 tomato before purchasing. Tomatoes on the vine were an indicator
product concepts were brought into the tournament section, with of freshness to some consumers. Many participants were generally
3 concepts per choice task. not willing to pay more for organic tomatoes. When asked why,
Kano questions were subsequently asked regarding the same at- some participants said that the organic tomato selection was not
tributes evaluated in the ACBC survey with the addition of places more visually appealing than conventional tomatoes.
where tomatoes were purchased (ex. Farmers’ market). Kano anal- Participants expected to buy different kinds of tomatoes in dif-
ysis is a technique where attributes are classified into quality-based ferent shopping environments. For example, because consumers
categories (Kano and others 1984; Erto and others 2011; Kim and can talk to the farmers and try samples in a farmers market, more
others 2013). These categories include: Attractive: unexpected than half of participants said it made them feel more comfortable
by the consumer; consumers are satisfied if this attribute is present. buying specialty types of tomatoes with unique colors and shapes
Indifferent: attributes that the consumer does not care about. in the farmers market. When participants were introduced to a
Must have: expected by the consumer; if unavailable, consumers dark green and brown tomato variety “Kumato,” many indicated
are dissatisfied. One dimensional: as the attribute increases, so they would never buy one of those in a grocery store without
does consumer liking. Reverse: leads to dissatisfaction. Attributes knowing any details about its flavor. Participants were only willing
were presented to participants in the form of a paired question. to buy unconventional tomato varieties such as “Costoluto Gen-
Each question was posed to the consumer as the attribute (func- ovese” which is large and has “ugly” ridges, when given positive
tional) and not fulfilled (dysfunctional). For example, tomatoes information about the produce, such as nutrient content or good
that are RED in color and tomatoes that are NOT RED in color. flavor. Even so, few participants indicated they would be willing
The response options for each question included “I will like it,” to purchase unconventional tomato varieties. In the grocery store,
“I must have it,” “I do not care,” “I can live with it,” and “I will participants expected to buy traditional types of tomatoes, such
dislike it.” as round and roma types. Local origin was an attractive feature.
The complete survey (awareness questions, demographics, con- When asked why, people agreed that it made them feel good to
joint, and Kano) was constructed and uploaded to an internet support the local economy by purchasing locally grown tomatoes.

Vol. 79, Nr. 10, 2014 r Journal of Food Science S2093

 Fresh market tomatoes . . .

Table 3–Importance scores for attributes evaluated in conjoint Table 4–Average utility scores for attributes evaluated in the
survey (n = 1037). conjoint survey (n = 1037).

Attribute Average importance Attribute Level Average utility

Color 22.9a Color Red 85.4a
Juice when sliced 14b Dark red 69.3b
Size 10.3c Light red 35.0c
Seed presence 9.6d Orange −25.6d
Firmness to touch 9.4de Yellow −62.3e
Health benefits 9.0e Green −101.9f
Texture 7.2f Firmness Firm 38.9a
Flavor 6.7f Soft −38.9b
Interior 5.6g Size Medium (baseball sized) 41.2a
Skin 5.3g Small (lemon sized) −4.1b
Large (grapefruit sized) −37.2c
Attribute importance scores for the total population. Letters (a–g) indicate significant
differences (P < 0.05). Skin Thin skinned 20.2a
Thick skinned −20.2b
Texture Crisp 16a
Soft −16b
Firmness to touch was considered an important attribute by the Interior Juicy 5.2a
majority of participants. Most participants would only buy toma- Meaty −5.2b
toes that were firm. Another highly cited attribute that determined Seed presence Few seeds 38.1a
No seeds 1.3b
tomato purchase was color. Most people liked dark-red colored
Many seeds −39.5c
tomatoes, with only a few people (<20%) mentioning they would Flavor Bold tomato flavor 21.3a
not mind a lighter red, or yellow, or orange color. Other at- Mild tomato Flavor −21.3b
tributes of interest, such as flavor, shape, size, juiciness, crispness, Health benefits Lycopene 26a
and skin types were mentioned by a small number of participants. Vitamin C 6.5b
Fiber 5.2b
For many consumers, good tomato flavor was not considered a key Vitamin A 1.4c
attribute for purchase decision, especially when tasting a tomato None −39.1d
was impossible in the grocery store. Most participants preferred Juice Small amount of juice 53a
perfectly round shaped tomatoes, but when compared with other Large amount of juice −6.1b
No juice −47c
factors such as firmness, color and price, shape was not as impor-
tant. Some unfavorable attributes were also identified by the focus Zero centered utility values for levels within attributes. Letters (a–f) indicate significant
differences (P > 0.05) within each attribute for the total population (n = 1037).
groups. The top 3 of these were mealy texture, seedy and full of
gel (too juicy).
For interior attributes, crisp texture was preferred over soft (P <
Conjoint analysis 0.05), a juicy interior was preferred over meaty (P < 0.05), few
S: Sensory & Food

Seventy-seven percent of those surveyed were female and 23% seeds were preferred to many or none (P < 0.05), bold flavor was
were male. Participants were mostly Caucasian (70%) followed by preferred over mild (P < 0.05) and a small amount of juice was
Quality

African American (15%). Most consumers had at least some col- preferred over a large amount or none (P < 0.05).
lege education (>95%). There was an even spread of age (18 to Four consumer clusters were identified from utility scores
65 y) and income (<$20000 to over $95000 per year). Sixty-five (Figure 1). Cluster 1 (n = 223) was characterized by consumers
percent of participants consumed tomatoes weekly and 28% con- who liked tomatoes that were juicy, red, soft, had bold flavor and
sumed tomatoes at least 2 to 3 times a month. Nearly half of par- a large amount of juice. Cluster 2 (n = 279) contained consumers
ticipants (48%) purchased tomatoes at least once a week. Most of who preferred firm, crisp tomatoes with no seeds. Cluster 3 con-
those surveyed (93%) purchased tomatoes at a local grocery store sumers (n = 219) were attracted to color, health benefits, bold
and 48% purchased tomatoes at a local farmer’s market. Color flavor and a small amount of juice and few seeds. Cluster 4 con-
was most important (P < 0.05), followed by juice when sliced sumers (n = 316) had the highest importance score for color and
(P < 0.05), followed by size (Table 3). Skin thickness and in- preferred tomatoes red in color, also rejecting soft tomatoes. Few
terior (juicy versus meaty) were the least important attributes. demographic differences were noted among consumer clusters
The importance of color was consistent with an internal prefer- (P > 0.05, results not shown). Cluster 4 consumers had a lower
ence mapping study with cherry tomatoes by Pagliarini and others incidence of purchasing tomatoes at a local farmers market com-
(2001). Wolters and Gemert (1990) also cited color as an impor- pared to other clusters (P < 0.05). This might explain why this
tant tomato attribute as well as size. Color was also an important cluster of consumers was solely color driven and not accepting of
attribute cited in the focus groups in the current study. atypical (not red) tomato color. There were no age, ethnicity or
An importance score differentiates the importance of attributes. income differences among the clusters (P > 0.05).
A utility score explains the attractiveness of levels within attributes,
higher utilities indicate more attractive attributes. Negative utility Kano questions
values do not necessarily indicate that an attribute is unattrac- All attributes were labeled “attractive” except for the colors
tive, rather it is less attractive compared to positive utility values yellow, orange, light red and green, areas of different color, soft
within the same attribute. For utility scores: red tomatoes were (exterior and texture), large size, thick skin, organic and the pres-
favored over other colors (P < 0.05) (Table 4). Among exterior ence of many seeds. Though attributes were classified as either
attributes, firm tomatoes were preferred over soft (P < 0.05), thin “attractive” or “reverse,” across all consumers, clusters were dis-
skinned tomatoes were preferred over thick skinned (P < 0.05) tinct (Table 5). Green color across the general survey population
and medium size was preferred over small and large (P < 0.05). was a negative attribute but for 2 clusters, green color had no effect.

S2094 Journal of Food Science r Vol. 79, Nr. 10, 2014

Fresh market tomatoes . . .

Cluster 1 (n = 223) was the only cluster to assign juicy interior as Discussion
a 1-dimensional performer, meaning the juicier the interior, the Focus group results were generally consistent with survey results.
higher consumer satisfaction would be. This result is consistent Focus groups revealed that external appearance (such as color) was
with utility scores for juiciness (Figure 1). This cluster assigned an indicator of the quality of a tomato to consumers. This was con-
“attractive” to features relating to health benefits, small/medium sistent with the conjoint survey: color was the attribute with the
size, and juiciness. This cluster was most dissatisfied with no juice, highest importance score. From the Kano questions, it was also ap-
no seeds and thick skin. This cluster was also most accepting of parent that appearance was important: colors other than red or dark
“nonred” color. red were negative attributes to the total population. However, just
Cluster 2 (n = 279) was the only cluster to classify firm tomatoes as a small proportion of people in focus groups indicated that they
as attractive, this attribute was indifferent to other consumers. Soft would be willing to try “unconventional appearing” tomatoes,
interior texture and softness of the whole tomato were dissatisfy- some consumer clusters in the survey (clusters 1, 2, and 3) were in-
ing attributes to these consumers. High satisfaction scores within different to “deviant” tomato colors, and conjoint and Kano ques-
cluster 3 (n = 219) were assigned to dark red and juicy. Cluster 4 tions allowed greater clarification of these attributes compared to
(n = 316) was the only cluster to assign “1-dimensional” to red, focus groups alone. According to the focus groups, traditional red
and “reverse” to yellow, orange and multiple areas of a different color was desired because of familiarity and consumers do not want
color. This cluster was also the only one to assign “indifferent” to to “risk” buying an unconventional tomato. Not surprisingly, con-
antioxidants, vitamins A, C, and D and bold flavor. This cluster joint and Kano results suggest that some consumers are more “ad-
had the highest number of indifferent attributes (n = 31) as well as venturous” and are willing to experiment with unexpected tomato
the lowest number of attractive attributes (n = 3, dark red color, colors.
flavor, health benefits), suggesting that cluster 4 consumers are less Juice when sliced received the second highest (P < 0.05) im-
likely to purchase fresh tomatoes for specific tomato attributes but portance score by conjoint analysis. Firmness, which relates to the
perhaps purchase them solely as condiments (for example, on a juice content, was important to focus group participants as well
sandwich) or for color in an application (salad), or have been dis- as to some in the conjoint survey—firmness was classified as an
appointed with previous purchases and have key expectations for attractive attribute in Kano analysis for the total population as well
color, flavor, and health benefits. as by cluster 2 consumers. Degree of firmness may also indicate

Biplot (axes F1 and F2: 92 %)

2

Large amount of
juice

S: Sensory & Food

Quality
Cluster 1
SoŌ SoŌ interior
Yellow Juicy Bold
Many seeds Vitamin C Thin skin Small amount of
Orange
Large Lycopene juice
F2 (7%)

0
Vitamin A Cluster 3
Increased Fiber Medium Dark red
Thick skin Small Few seeds
Green Red
Cluster 4
No seeds
No health benefits Light Red
Mild Meaty Crisp
Cluster 2 Firm

No juice

-2
-5 -3 -1 1 3
F1 (85%)

Figure 1–Principal component biplot of consumer clusters with respect to utility scores.

Vol. 79, Nr. 10, 2014 r Journal of Food Science S2095

 Fresh market tomatoes . . .

Table 5–Total Population and Clustered Kano Results.

Feature Total pop. (n = 1037) Cluster 1 (n = 223) Cluster 2 (n = 279) Cluster 3 (n = 219) Cluster 4 (n = 316)
Yellow Reverse Indifferent Indifferent Indifferent Reverse
Orange Reverse Indifferent Indifferent Indifferent Reverse
Light red Reverse Indifferent Indifferent Indifferent Indifferent
Red Attractive Attractive Attractive Attractive 1-dimensional
Dark red Attractive Attractive Attractive Attractive Attractive
Green Reverse Indifferent Indifferent Reverse Reverse
Even color Attractive Indifferent Indifferent Indifferent Indifferent
Small area of different color Reverse Indifferent Indifferent Indifferent Indifferent
Multiple areas of a different color Reverse Indifferent Indifferent Indifferent Reverse
Firm (firmness) Attractive Indifferent Attractive Indifferent Indifferent
Soft (firmness) Reverse Reverse Reverse Reverse Reverse
Small (lemon sized) Attractive Indifferent Indifferent Indifferent Indifferent
Medium (baseball sized) Attractive Indifferent Indifferent Indifferent Indifferent
Large (grapefruit sized) Reverse Indifferent Indifferent Indifferent Indifferent
Thick skin Reverse Indifferent Indifferent Indifferent Indifferent
Thin skin Attractive Indifferent Indifferent Indifferent Indifferent
Crisp (texture) Attractive Indifferent Indifferent Indifferent Indifferent
Soft (texture) Reverse Indifferent Reverse Indifferent Indifferent
Meaty Attractive Attractive Attractive Indifferent Indifferent
Juicy Attractive 1-dimensional Indifferent Attractive Indifferent
No seeds Attractive Indifferent Indifferent Indifferent Indifferent
Few seeds Attractive Indifferent Indifferent Indifferent Indifferent
Many seeds Reverse Indifferent Indifferent Indifferent Indifferent
Mild Attractive Indifferent Indifferent Indifferent Indifferent
Bold Attractive Attractive Attractive Attractive Indifferent
Health benefits Attractive Attractive Attractive Attractive Attractive
Antioxidants Attractive Attractive Attractive Attractive Indifferent
Lycopene Attractive Attractive Attractive Indifferent Indifferent
Vitamin A Attractive Attractive Attractive Attractive Indifferent
Vitamin C Attractive Attractive Attractive Attractive Indifferent
Vitamin D Attractive Attractive Attractive Attractive Indifferent
Fiber Attractive Attractive Indifferent Attractive Indifferent
No juice Reverse Reverse Indifferent Reverse Indifferent
Small amount of juice Attractive Indifferent Indifferent Indifferent Indifferent
Large amount of juice Reverse Indifferent Indifferent Indifferent Indifferent
Flavor Attractive Attractive Attractive Attractive Attractive
Organic Indifferent Indifferent Indifferent Indifferent Indifferent
Farmers market Attractive Attractive Indifferent Attractive Indifferent
S: Sensory & Food

Heirloom Attractive Attractive Indifferent Indifferent Indifferent

Vine Attractive Attractive Attractive Indifferent Indifferent
Quality

Kano classification was calculated by previously described methods (Kano and others 1984). The satisfaction and dissatisfaction questions were asked to consumers and the
contingency table of satisfaction and dissatisfaction answers was created for each feature.

the presence of gel and being soft and mealy; soft was undesir- which was color driven, was indifferent to individual health ben-
able by Kano questions to all consumers. Tomatoes that are not efits although they classified the general term “health benefits” as
firm are susceptible to chilling injury (Jackman and others 1990). attractive. The importance score for health benefits was low com-
Chilling injury (physical damage that occurs in plants because of pared to the importance of other attributes. In another study con-
exposure to low but not freezing, temperature) can be character- cerning dairy products, health benefits were attractive, but were
ized by nonuniform color, uneven surfaces and patches of green not as important as the flavor of the product (Bower and others
and yellow (Jackman and others 1990). These terms that Jackman 2003). Health benefits of tomatoes may be important to the con-
and others (1990) used to describe the effects of chilling injury sumer, but when faced with other attributes to choose from, health
on tomato skin are similar to attributes used in this survey: yel- takes a back seat to other attributes such as color, firmness, or
low, green, small area of a different color and multiple areas of a flavor.
different color. All of these attributes were reverse in Kano anal- Size, a top 3 attribute for the total population by conjoint anal-
ysis meaning they were undesirable to consumers. Firmness may ysis, was also defined as important by focus group participants.
be used to evaluate a tomato for the extent of its chilling injury, Half of focus group participants stated size was important in se-
which consumers may be able to visualize. lecting a tomato. Medium tomato size (baseball sized) was the
Among health benefits, lycopene received the highest utility ideal size for the total population in the conjoint survey. This
score in the conjoint survey (P < 0.05), greater than vitamins A was consistent with those in focus groups, most participants said
and C, and fiber. When asked about familiarity with lycopene that medium tomatoes were the ideal size. Those who preferred
in the demographic section, only approximately 25% of partic- smaller tomatoes said they would use smaller tomatoes for a salad.
ipants were not familiar with lycopene. While discussing health Some participants stated in the focus groups that large tomatoes
benefits in focus groups, most participants were aware that toma- were undesirable because they were more likely to be mealy. In
toes contained lycopene but very few mentioned other specific Kano analysis, tomato sizes (small, medium and large) were all in-
health benefits that they thought tomatoes possessed. Cluster 4, different attributes. Literature is sparse on consumer preference for

S2096 Journal of Food Science r Vol. 79, Nr. 10, 2014

Fresh market tomatoes . . .

tomato size, but our results suggest that a moderate size (baseball) Baldwin EA., Scott JW, Einstein MA, Malundo TMM, Carr BT, Shewfelt RL, Tandon KS.
1998. Relationship between sensory and instrumental analysis for tomato flavor. J Am Soc
is generally preferred, perhaps for its versatility. Hort Sci 123:906–15.
In a previous consumer study with cherry tomatoes, 2 con- Bellenger DN, Bernhardt KL, Godstucker JL. 1976. Qualitative research techniques: focus group
sumer clusters were revealed; preferring either red color or firm interviews. Qual Research Marketing 3:7–28
Bisogni CA., Armbuster G, Brecht PE. 1976. Quality comparisons of room ripened and field
texture (Pagliarini and others 2001). Clusters 1, 3, and 4 in this ripened tomato fruits. J Food Sci 41:333–8.
study were driven by color and cluster 2 was driven by firmness. Bower JA, Mohammad AS, Whitten C. 2003. Effect of liking, information and consumer
characteristics on purchase intention and willingness to pay more for a fat spread with a
In the current study with regular-sized tomatoes, the 3 clusters proven health benefit. Food Qual Pref. 14:65–74.
driven by color were further differentiated by preference for juice Bruhn CM, Feldman N, Garlitz JH, Ivans E, Marshall M, Riley A, Thurber D, Williamson
E. 1991. Consumer perception of quality: apricots, cantaloupes, peaches, pears, strawberries,
when sliced (cluster 1), health benefits (cluster 2), and rejection of and tomatoes. J Food Qual 14:187–95.
soft tomatoes (cluster 4). Both color and firmness can be used to Cardello AV, Schutz HG. 2003. The importance of taste and other product factors to consumer
interest in nutraceutical products: civilian and military comparisons. J Food Sci 68:1519–24.
judge the ripeness of a tomato. Ripe tomatoes are higher in sugar Causse M, Buret M, Robini K, Verschave P. 2003. Inheritance of nutritional and sensory quality
than unripe tomatoes, and sweet taste in tomatoes was previously traits in fresh market tomato and relation to consumer preferences. J Food Sci 68:2342–50.
Chaı̈b J, Devaux MF, Grotte MG, Robini K, Causse M, Lahaye M, Marty I. 2007. Physiological
correlated with overall tomato flavor liking (Malundo and others relationships among physical, sensory and morphological attributes of texture in tomato fruits.
2007; Baldwin and others 1998). Color is not a surprising purchase J Exp Bot 58:1915–25.
Childs JL, Drake MA. 2009. Consumer perception of fat reduction in cheese. J Sens Stud 24:
decision and has been established as a driver of liking (Pagliarini 902–21.
and others 2001). Firmness was also a driver of purchase deci- Childs JL, Drake MA. 2010. Consumer perception of astringency in clear acidic whey protein
beverages. J Food Sci 75:513–21.
sions but consumers differed in their preferred level of firmness. Chung HS, Hong H, Kim K, Cho CW, Moskowitz HR, Lee SY. 2011. Consumer attitudes
Although consumers varied in their preferred level of firmness, and expectations of ginseng food products assessed by focus groups and conjoint analysis. J
this attribute is one that can be perceived externally and is used Sens Stud 26:346–57.
Davies JN. 1966. Changes in the non-volatile organic acids of tomato fruit during ripening. J
to help consumers select a tomato in a situation where it cannot Sci Food and Agri 17:396–400.
directly be sampled. Flavor was an attractive attribute in Kano Erto P, Vanacore A, Staiano M. 2011. A service quality map based on Kano’s theory of attractive
quality. The TQM Jour 23:196–215.
analysis, but it was not a top attribute by conjoint. The majority Food and Agriculture Organization of the United Nations. 2011. FAOSTAT. Available from
of focus group participants indicated that flavor was important, http://faostat3.fao.org/faostat-gateway/go/to/download/Q/QV/E. Accessed 2013 March 3.
Galvez FCF, Resurreccion AVA. 1992. Reliability of the focus group technique in determining
and that they expected and desired tomatoes to have good flavor. the quality characteristics of mungbean noodles. J Sens Stud 7:315–26.
Other attributes may be indirectly used to judge the flavor of a Grunert KG. 1997. What’s in a steak? A cross-cultural study on the quality perception of beef.
Food Qual Pref 8:157–74.
tomato, since flavor cannot be evaluated before purchase. It is also Harker FR, Redgwell RJ, Hallett IC, Murray SH, Carter G. 1997. Texture of fresh fruit. Hort
important to note that while a large number of consumers were Reviews 20:121–224.
Hongsoongnern P, Chambers E. 2008. A lexicon for texture and flavor characteristics of fresh
included in this study, they were all from the Raleigh, NC area and processed tomatoes. J Sens Stud 23:583–99.
and it is possible that they may not represent all U.S. consumers. Jackman RL, Marangoni AG, Stanley DW. 1990. Measurement of tomato fruit firmness. Hort
Sci 25:781–3.
A survey with consumers from other U.S. regions would be nec- Jervis SM, Ennis JM, Drake MA. 2012. A comparison of adaptive choice based conjoint and
essary to confirm if all U.S. consumer preferences are represented. choice based conjoint to determine key choice attributes of sour cream with limited sample
size. J Sens Stud 27: 451–62.
The importance of tomato flavor as well as how the experience of Jones RA, Scott SJ. 1983. Improvement of tomato flavor by genetically increasing sugar and acid
consuming a tomato (visual, touch, slicing, and then eating) influ- content. Euphytica 32:845–55.
ence consumer acceptability should be evaluated in future studies. Kader AA, Stevens MA, Albright-Holton M, Morris LL, Algazi M. 1977. Effect of fruit ripeness

S: Sensory & Food

when picked on flavor and composition in fresh market tomatoes. J Amer Soc Hort Sci
This study established how consumers select a desirable tomato. 102:724–31.
Kano N, Seraku N, Takahashi F, Tsuji S. 1984. Attractive quality and must be quality. Quality

Quality
14:39–48.
Conclusion Kim M, Lopetcharat K, Drake MA. 2013. Influence of packaging information on consumer
The use of focus groups and a conjoint survey identified and liking of chocolate milk. J Dairy Sci 96:4843–56.
Krause AJ, Lopetcharat K, Drake MA. 2007. Identification of the characteristics that drive
characterized fresh tomato attributes that consumers desired. Con- consumer liking of butter. J Dairy Sci: 90:2091–2102.
sumers were driven by attributes that helped them judge ripeness Malundo TMM, Shewfelt RL, Scott JW. 1995. Flavor quality of fresh tomato (lycopersicon
esculentum mill.) as affected by sugar and acid levels. Postharvest Biol Tech 6:103–10.
which included color and firmness. The most important tomato Orme BK. 2010.Getting started with conjoint analysis: strategies for product design and pricing
attribute of the total population surveyed, as well as focus group research|. Madison, Wis. Research Publishers, p 39–50; 78–8.
Pagliarini E, Monteleone E, Ratti S. 2001. Sensory profile of eight tomato cultivars (lycopersicon
participants, was color. Color was indicative of tomato quality, esculentum) and its relationship to consumer preference. Italian J Food Sci 13:285–96.
and a traditional red color was most familiar and preferred. Firm- Rao VR. 2010. Conjoint analysis. Wiley international encyclopedia of marketing. Avail-
able from: http://onlinelibrary.wiley.com/book/10.1002/9781444316568. Accessed 2013
ness and juice when sliced were also desirable attributes and were March 3.
indicative of ripeness. Consumer clusters were identified and dif- Redgwell RJ, Fischer M. 2002. Fruit texture, cell wall metabolism and consumer perceptions.
In: Knee M, editor. Fruit quality and its biological basis. Oxford: Blackwell, p 46–88.
ferentiated by their preferences for color, juiciness, health benefits, Serrano-Megias M, Lopez-Nicolas JM. 2006. Application of agglomerative hierarchical clus-
and firmness. These results can be utilized by tomato producers tering to identify consumer tomato preferences: influence of physicochemical and sensory
characteristics on consumer response. J Sci Food Agric 86:493–9.
for breeding/selection of cultivars that meet these consumer pref- Stevens MA, Kader AA, Albright-Holton M, Algazi M. 1977. Genotypic variation for flavor
erences or by produce marketing for strategic advertising of key and composition in fresh market tomatoes. J Am Soc Hort Sci 102:680–9.
Szczesniak AS. 2002. Texture is a sensory property. Food Qual Pref 13:215–25.
consumer attributes. Tandon KS, Baldwin EA, Scott JW, Shewfelt RL. 2006. Linking sensory descriptors to volatile
and nonvolatile components of fresh tomato flavor. J Food Sci 68:2366–71.
Acknowledgment Thompson JL, Gerard PD, Drake MA. 2007. Chocolate milk and the Hispanic consumer. J
Food Sci 72:666–75.
The use of trade names does not imply endorsement nor lack Van kleef E, Van trijp, HCM, Luning P. 2005. Consumer research in the early stages of new
product development: a critical review of methods and techniques. Food Qual Pref 16:181–
of endorsement by those not mentioned. 201.
Wolters C J, Gemert L J van. 1990. Towards an integrated model of sensory attributes, in-
References strumental data and consumer perception of tomatoes. Part 1. Relation between consumer
perception and sensory attributes. Acta Hort (ISHS) 259:91–106.
Baldwin EA, Goodner K, Plotto A. 2008. Interaction of volatiles, sugars and acids on perception
of tomato aroma and flavor descriptors. J Food Sci 73:294–307.

Vol. 79, Nr. 10, 2014 r Journal of Food Science S2097