Sensory benefits, emotions and usage patterns for olives: using Internet-based conjoint analysis and...

www.elsevier.com/locate/foodqual

Food Quality and Preference 16 (2005) 369–382

Sensory benefits, emotions and usage patterns for olives:using Internet-based conjoint analysis and segmentation

to understand patterns of response

Howard Moskowitz a,*, Matthias Silcher a, Jacqueline Beckley b,Dorothy Minkus-McKenna c,*, Teri Mascuch d

a Moskowitz Jacobs Inc., 1025 Westchester Ave., White Plains, NY 10604, USAb The Understanding & Insight Group, 3 Rosewood Lane, Suite 103, Denville, NJ 07834, USA

c Berkeley College, 3 East 43rd Street, New York, NY 10017, USAd McCormick and Company, 204 Wight Avenue, Hunt Valley, MD 21031, USA

Received 12 June 2002; revised 28 January 2004; accepted 28 January 2004

Available online 10 August 2004

Abstract

Conjoint analysis was used in an Internet-enabled format in order to understand the drivers of interest at the concept level for olives,

a common but not well-researched snack food. The study comprised 36 descriptive phrases covering sensory properties, uses, product

origins and emotions. The utility values revealed differences in interest level when respondents rated the test concepts on an intensity of

craving scale. Although there were some differences in response pattern due to other variables such as hunger, the concept–response

segmentation showed the largest differences, and provides an analytic strategy to uncover groups of individuals with interest in main-

stream or more narrow food categories who demonstrate heightened response to specific messages and product features.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Choice; Consumer motivation; Product attributes; Conjoint; Snack foods; Olives

1. Introduction

American consumers are spending an increased

amount of money on snack foods, 1 according to US

0950-3293/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.foodqual.2004.01.003

* Corresponding authors. Tel.: +1 914 421 7400; fax: +1 914 428

8364.

E-mail addresses: [email protected] (H. Moskowitz), dmm@

berkeleycollege.edu, [email protected] (D. Minkus-McKenna).1 The term ‘‘snack’’ is difficult to define as all eating patterns are

based on culture. According to Oltersdorf et al. ‘‘To date, there is no

general accruement on definitions to categorize the different eating

occasions, like meal or snack.’’ (p. 4). Webster�s Dictionary defines

snack as ‘‘a light meal: food eaten between meals.’’ The Snack Food

Association membership includes, but is not limited to, manufacturers

of potato chips, tortilla chips, cereal snacks, pretzels, popcorn, cheese

snacks, snack crackers, meat snacks, pork rinds, snack nuts, party mix,

corn snacks, pellet snacks, fruit snacks, snack bars, granola, snack

cakes, cookies and various other snacks (www.sfa.org).

Market Trends and Forecasts (deJong, 1999). Snackingfoods overall is a $40 billion a year market (Mintel

International Group Ltd., 2003a). This trend is relevant

to health and wellness because in many cases consump-

tion of snack foods is driven by factors other than basic

nutritional needs. Two words that continue to emerge

regarding snack foods are �want� or �desire�. The desire

by consumers for a particular snack food can be de-

scribed by them as a craving. The exact attributes of thisterm are elusive. A simple, universally accepted defini-

tion of craveability does not exist. According to Web-

ster�s Dictionary, crave is defined as ‘‘to want greatly

or to yearn for’’, and is a word used to describe non-clin-

ical, highly desired foods. Pelchat and Schaefer (2000)

use ‘‘an intense desire or longing to eat a particular

food’’ and note that food cravings are considered to

be highly ‘‘salient psychological phenomena.’’ This

mailto:[email protected]

mailto:dmm@

mailto:[email protected]

http://www.unf.edu/groups/snc/articles/health

370 H. Moskowitz et al. / Food Quality and Preference 16 (2005) 369–382

paper will use the concept of craving as what it meant to

the consumers who decided to participate in this study

and reacted to the scale of craving intensity.

A recurrent issue in both science and business for

food products is the absence of a significant body of

data about what drives consumers to have deep and sus-tained desire for certain foods. One can find a large body

of research regarding liking and food acceptance; how-

ever, there is a lack of substantial research in the area

of sustained specific desire for specific foods with a large

base of consumers.

Olives serve as the basis of this study on craveability

because of their long history and global usage, but a

dearth of formalized knowledge on subjective character-istics exists. Little appears to have been published about

the olive per se. Olives are not particularly popular

among the mainstream American population, although

they have been a staple in middle-eastern diets for thou-

sands of years. Furthermore, olives are commonly used

as an ingredient in food preparations. This under-utili-

zation of a well-known food, unique properties of the

olive, and the possibility to identify new positioningstatements to understand consumers� desire for particu-

lar foods provided rationale for utilizing olives for fur-

ther study.

In addition, this study utilizes olives, as a model for

the snack food market. Following the logic of increasing

value of a data set when it is linked to a larger �macro-

study�, the study on olives comes from a larger study

on 30 foods which may be consumed as snacks, con-ducted in two waves (20 wave 1, 10 wave 2), which used

the same frame of reference described above.

This paper deals with olives in two ways: the first sub-

stantive and the second methodological.

Substantive. Explore consumer responses to an

under-studied, yet culturally important food, olives,

often used as a non-carbohydrate snack. This particular

information is relevant for a food industry looking for�good for you� foods, especially snack items.

Methodological. The objective is to illustrate how

Internet-enabled conjoint measurement provides a

source of information and insight about a food, and par-

ticularly about a snack food. This particular approach

and the ensuing information are important because it

provides a template for research studies in a variety of

food and non-food areas.The use of conjoint analysis is uniquely capable of

helping understand the impact of food desirability due

to its ability to incorporate minor modifications of

words and phrases to determine whether there is a

change in response from a person or not. Conjoint anal-

ysis systematically varies the composition of small,

descriptive vignettes about a product. These vignettes

comprise components or elements dealing with differentaspects of the product. Some of these vignettes or test

ideas will be rated more highly for the generation of

more intense craveability response while other vignettes

will generate less intensity of response. The statistical

analysis of the ratings reveals which specific elements

drive this intensity of �craveability� for an individual.

The learning about consumer based reactions of

craveable intensity increases dramatically when the dataset is constructed from a set of linked conjoint studies.

These studies are designed to be structurally similar,

but each study pertains to a separate food. The research-

er is afforded greater knowledge about the types of com-

munications that drive craveability, and can discover

overarching patterns that might be elusive with only

one study on one product (Beckley & Moskowitz,

2002). With the notion of intensity of craving, as deter-mined by the consumer taking the study, as the rating

attribute, and with linked studies using conjoint analy-

sis, a general pattern of three segments has emerged

when the respondents rated intensity of craving. These

three segments are ‘‘Elaborates’’ (motivated by novel

product features), ‘‘Classics’’ (responding to the classical

product form), and ‘‘Imaginers’’ (features beyond the

food as a driver of heightened response).

2. Snacks and snacking behavior

2.1. Snack foods

An exact consumer definition of snack food is elusive

(see footnote 1). Synonymous with snack foods are sav-ory snacks, potato chips, snack nuts, and popcorn (de-

Jong, 1999). Savory snacks comprise pretzels, tortilla

chips, cheese curls/puffs, and meat snacks. Other foods

eaten as snacks may be sweet and generally referred to

as candies. Chocolate, alone or in combination with

other ingredients, is one of the most popular sweet

snacks.

The challenge for snack developers is to find a newproduct that will be successful in a field where nearly

three-quarters of all launches fail (Roberts, 2001). The

issue of finding this snack is made more difficult and

yet paradoxically perhaps easier with the increasing

exploration of new flavors by consumers. With almost

everything that touches our lives going global these

days, it should be no wonder that consumers� taste budsalso want to expand their global travels with snackfoods. Spice and flavor producers state that the US

snack food market continues to use ingredients that re-

flect the growing demographic diversity of the country,

as well as the desire of consumers to have snacks with

rich, robust and exotic tastes (Hoch, 2000). Yet, with

all of the efforts on product development, innovation,

globalization, and marketing, the identification of a suc-

cessful, good-for-you snack, with the potential to be-come popular (i.e. purchased and repeat purchased) in

a crowded market continues to remain elusive.

H. Moskowitz et al. / Food Quality and Preference 16 (2005) 369–382 371

2.2. Snacking behavior and health issues

One rationale for the consumption of snacks is that

adults and especially young children need to eat snacks

because it is difficult to obtain enough energy and nutri-

ents in the three main meals. In general, a snack shouldbe consumed in small amounts at frequent intervals.

Mid-morning, mid-afternoon (especially after school

for older children) and after-dinner snacks can become

part of a daily routine. Carefully planned, nutritious

snack foods should make a valuable contribution to

the daily nutrient and energy intake. A snack should

not take the place of a meal, 2 but can be used as a

�mini-meal� that supplies protein, vitamins, mineralsand energy (Baker & Henry, 1987).

The foregoing, however, is an �idealized� descriptionof snacks. It is not the way reality is, however. The qual-

ity and quantity of snacks in individuals� diet can con-

tribute to a number of health problems, such as

obesity when they become part of the cycle of height-

ened energy consumption. Over 300,000 deaths each

year in the United States are related to obesity. Ameri-cans consume 4.72 billion pounds of snack foods a year

(Abersold, 1998). In a 1994 nutritional survey of snack-

ing patterns among 1800 adults and children, the Amer-

ican Dietetic Association found that all age groups

snacked primarily at one time during the day. Asked

the top reasons for choosing a snack, the top response

was taste, followed by health or nutrition. The most

popular and typical snacks include soft drinks, potatoand corn chips, cookies, candy bars, and ice cream

(Cross, Babicz, & Cushman, 1994; Sullivan et al., 2002).

Growth of healthier snacks has been driven by the

natural chip/popcorn/pretzel segment; yet the most

growth was in the smallest category––energy and snack

bars (Mintel International Group Limited, 2003b).

Healthy snacks and the search for items that can become

exemplars, thus take on increased importance in light ofnational health awareness regarding issues like obesity.

From 1998 to 2003, healthy snacking in the US has

established a steady growth pattern, influenced greatly

by diet trends and an increased awareness of incidence

of overweight and obesity issues. Ramona Bennett, mar-

keting manager with Cargill Salt concludes, ‘‘The thing

we are seeing and will continue to see is more variety

with regards to organic and natural snacks. GMO-freeand healthiness is going to be a major concern. The rea-

son is that 50–55% of Americans are overweight’’ (Ro-

berts, 2001).

2 In Oltersdorf et al. (p. 4), the criteria for a meal consists of time of

consumptions (eating events in the morning, at midday and in the

evening); energy content (consumption of more than 375 kcal); social

interaction (presence of fellow eaters); food quantity (more than a

single food); and energy content (Oltersdorf, Schlettwein-Gsell, &

Winkler, 1999).

As Americans have become more interested in taking

control of their health and as obesity in the United

States has reached near epidemic levels, the Food Indus-

try has made health and nutrition a major part of its

growth strategy. This focus continues to be reaffirmed

by corporations in the popular press, but is often chal-lenged by researchers and dietary specialists. For exam-

ple, PepsiCo Chief Executive Officer Steve Reinemund

says that as a food and beverage company PepsiCo

has a �responsibility to offer nutritional guidance to con-

sumers�. Marion Nestle, chairwoman of New York Uni-

versity�s Department of Nutrition and Food Studies, is

skeptical about these types of motivations and corporate

policies directed by those motivations. ‘‘The argumentthat all foods can be part of a healthy diet is the favorite

argument of the food industry,’’ Nestle says. She sees

low fat alternatives of regular snacks as just a way to sell

more food. ‘‘It�s still junk food,’’ she says (Klingbeil,

2003).

2.3. Understanding consumer responses to foods through

a large-scale �Foundation Study�

One strategy to identify the features of a good-for-

you snack food is to survey responses to concepts about

many different foods, as well as asking consumers about

the occasions when they eat the food. Rather than deter-

mining the food as a snack food ahead of time, one can

identify when consumers eat different foods, and thereby

understand more fully the types of foods eaten duringsnack occasions as compared to meal occasions. Fur-

thermore, if the same study deals with different aspects

of the food, then it becomes possible to identify the fea-

tures of the food that drive this type of usage occasion.

This �two-pronged� approach, working with appropri-

ateness of a food for various occasions and intensity

of desire (craving) for the food at the level of �concept�may help to identify the most compelling attributes fornew good-for-you snacks. This approach of working

with appropriateness and then relevant communications

extends the pioneering work of Schutz (Schutz & Mar-

tens, 2001) on appropriateness as a key dimension for

foods.

A �Foundation Study� merges questionnaires about

one�s interaction with products or ideas with responses

to communications about the same products or ideas(Beckley & Moskowitz, 2002). The Crave It! Founda-

tion Study, from which the olive data are taken,

comprises studies using conjoint measurement (Green,

Carroll, & Goldberg, 1981). The underlying ration-

ale for Foundation Studies is to create an inte-

grated database, in which results from one study

could illuminate results from another, and wherein

trends across categories might emerge. Such integratedstudies enable the researcher and the marketer to

understand the individual food in detail, and the food


in relation to the performance of other foods. The ele-

ments studied in the conjoint analysis deal with product

features, emotional benefits, origin, product use, etc.

The larger study comprised a set of 30 different, but

linked studies, all having the same structure for

the stimuli (conjoint measurement), and all havingthe same classification questionnaire. This foundation

study, specifically the study dealing with olives, may

provide insights about a potential good-for-you, health-

ful snack.

2.4. Olives

Olives, biologically classified as a fruit, have been cul-tivated for over 5000 years. Approximately 10% of olive

production is for consumption with the remaining 90%

converted into olive oil. Depending on the processing

method, olives are identified as black olives, or green

olives, also referred to as Spanish olives.

In the United States, olives are primarily sold in cans

or jars in the condiment section of grocery stores. In eth-

nic neighborhoods and delis, olives are sold in bulk ineither the black or green variety, and prepared with

various flavorings. Given the ethnic heritage, the growth

of snacks in the United States, and the endless search

for new food products, a rationale for further explora-

tion of attitudes towards olives as a snack food by the

general consumer exists. Olives can be used as is

(straight from the can or jar) or as an ingredient in rec-

ipes. If used as is, then olives may be eaten as a snackfood and thus fall into the category of a snack food that

might conceivably be �craved�, at least as describe in

Section 1.

3. Method

3.1. Conjoint analysis

Conjoint analysis typifies an approach that goes be-

yond conventional questionnaires to understand the

consumer�s response to concepts, and to identify �driversof behavior� (Green et al., 1981; Moskowitz, 1994,

Chap. 1; Moskowitz & Martin, 1993; Wittnk & Cattin,

1989). The conjoint analysis approach can be described

as �stimulus–response�. The stimuli or the raw materialfor a conjoint measurement study comprise single,

stand-alone elements. These elements or phrases are

combined with other phrases to form a concept. The

combination uses principles of experimental design

(Box, Hunter, & Hunter, 1978), which ensures that these

elements appear statistically independently, albeit in the

body of a test concept. The specific stimuli comprised

36 elements. The elements for olives and their ration-ale in the Crave It!� Foundation Study appears in

Appendix A.

The elements were created by a team of research pro-

fessionals with extensive knowledge in the areas of psy-

chology, food product testing, concept development,

and general consumer communications, respectively. A

pilot study was developed to understand the appropriate

structure of the phrase and the categories. With resultsfrom the pilot (and over 300 consumer responses), the

Crave It! categories were created. The general phrases

developed in the pilot study were then customized for

each specific category (Olives being one of the catego-

ries). The phrases were developed from words and ideas

used by manufacturers of the products, off of websites,

in the popular media, from journals and other scholarly

publications, and using terms that the general publicassociates with food.

3.2. Experimental design

The basic experimental design comprised 60 combi-

nations, with 2–4 elements in each combination or vign-

ette describing the olive. The experimental design

structure was permuted in order to create 20 differentvariations. By permuting the design structure, but keep-

ing the design structure itself unchanged, the researcher

ensures that no single combination of concept elements

could ever dramatically influence the results. Each

respondent evaluated the 60 elements in the design in

a unique randomized order, which further eliminated

the possibility of order bias.

The respondent rated the concepts on a single attri-bute scale, anchored at both ends. The question read

�How intense is your craving for this olive?; 1 = Not

craveable at all, 9 = Very intense�. The scale was presentat the bottom of each concept. The respondent read the

concept, and rated the concept by clicking on the appro-

priate scale button which was identified with a numerical

value from 1 to 9.

After completing the evaluation of the 60 systemati-cally varied concepts, all rated by all consumers against

the intensity of craving scale, the respondent completed

a classification questionnaire, dealing with geo-demo-

graphics and attitudes towards olives. The classification

questionnaire comprised 16 questions, and was materi-

ally the same across the different foods in the Crave It!

Foundation Study, enabling a cross-food analysis. The

questions focused on

1. geo-demographics (age, gender, income, market, time

of day),

2. body state (self-reported hunger level, oral health

such as dentures),

3. drivers of craveability (product features and situa-

tions suggested in the literature as being believed to

be associated with food cravings),4. liking of the food (FACT Scale, Schutz, 1964),

5. location the item is purchased.


3.3. Respondents and invitations

The respondents were members of an opt-in e-mail

service (Open Venue Ltd., Toronto), who agreed to par-

ticipate in surveys. The respondents were sent an e-mail

invitation which asked them to participate in a study oncravings. They were directed towards a �wall� of 20 stud-

ies. One of these 20 was the olive study. A total of 44,000

invitations were sent out over a two-week period. From

this number more than 4500 respondents participated in

the full set of studies. Altogether 168 respondents com-

pleted the olive study and generated the database used

for this paper.

3.4. Modeling responses at the individual respondent

level

The study was set up so that a model could be cre-

ated for each respondent. The ratings for each respond-

ent were analyzed twice by regression analysis, after

two separate data transformations. The first data

transformation consisted simply of multiplying the rat-ing (1–9) by 11 to generate an 11–99 point scale. The

rationale was to make the coefficients easier to read.

This is called the persuasion scale. The persuasion scale

is used to determine whether or not each respondent val-

idly rated the concepts, in terms of consistency of

response.

The second transformation comprises a simple binary

transformation. If the rating for the concept was 1–6,then the rating was transformed to the value ‘‘0’’. In

contrast, if the rating for the concept was 7–9, then

the rating was transformed to the value ‘‘100’’. This

transformation generated a binary response, called the

�top 3 interests� (or Top 3 for short). The Top 3 repre-

sents the type of data used by marketers, who are inter-

ested in the proportion of respondents interested in a

concept. The marketer does not typically focus on theintensity of the interest but is more interested in the

number of individuals who are positive towards the con-

cept. This difference in the focus, a result of different

intellectual heritages enjoyed by researchers in psychol-

ogy versus sociology, generates differences in the way

that data is analyzed.

After the data were transformed, the individual data

was analyzed by dummy variable regression analysis, atthe individual level. Dummy variable regression is a ver-

sion of ordinary least squares analysis in which the pre-

dictor, independent variables, taken on one of two

values; �0� if the element is absent from the concept, or

�1� if the element is present in the concept. The data from

each respondent generated a simple additive model of

the form:

Rating ¼ k0 þ k1 ðElement 1Þ; . . . ; k36 ðElement 36Þ

3.5. Interpreting the results

The additive constant, k0, can be interpreted as the

conditional probability that an olive will be craved

(viz. the concept about olives rated 7–9) if there are no

elements present in the concept. Clearly this additiveconstant represents only a theoretical value because all

concepts evaluated by respondents comprised at least

two and at most four elements. The regression approach

requires the use of the additive constant as a correction

factor. The individual coefficients, k1 to k36, each show

the conditional probability that the concept describing

the olive will drive the rating of �crave� (e.g., higher

scaled intensity) if the element is added to the concept.Other words for the coefficient are �utility� and �impact�.

From many previous studies, where the concept ele-

ments were evaluated using experimental design, the fol-

lowing norms emerge, that allow interpretation of the

results from this particular study. These norms are not

precise numbers, but rather suggest ranges correspond-

ing to different behaviors in the market.

a. Utility value <0
Element detracts from
strength of concept

b. Utility value 0–5
Element adds to strengthof concept, but only slightly
c. Utility value 6–10
Element adds to strength of
concept, should be in concept

d. Utility value 11–15
Element is important
(often seen in segment results)

e. Utility value >16
Element is extremely important
For this paper, and for the Foundation Studies on

foods in general, the operational definition of �intensityof craving� is the sum of two parts; the additive constant,and the average utility value for the first nine concept

elements that deal with the features of the specific food.

This definition represents the re-composition of the util-

ity function into a single estimated value. One particular

benefit of conjoint analysis is the ability to re-create new

combinations and estimate their total utility value by

summing the utilities of the components.

3.6. Consistency and validity of individual data

The issue of validity of data cannot be addressed eas-

ily in consumer research because there are no external

criteria by which to assess validity. Validity often then

devolves down to the demonstration that the panelist

was actually a validly recruited consumer. Fortunately

for studies in which the stimulus is systematically varied,the researcher does have a way to establish validity. One

can relate the ratings assigned by the panelist to the

presence/absence of the concept elements. To the degree

that the panelist is consistent in the ratings of the full

0.5 0.6 0.7 0.8 0.9 1.0RSQUARE

Fig. 1. Distribution of the R2 statistic for olives.


concepts, one will obtain a high value for the R2, which

measures the goodness of fit of the model to the actualdata. R2 goes beyond reliability to validity, because

the statistic measures how well the independent varia-

bles account for the variation in the data. The R2 statis-

tic is computed on the persuasion data, which was

derived from a simple multiplication of the rating scale

by the value 11. A simple multiplicative transformation

does not affect the R2 statistic.

Table 1

Percent of respondents identifying sensory inputs as being key to craving th

Food Taste Appea

Average 90% 43%

Taste primarily

Potato chips 93% 24%

Cheese 93% 43%

Cola 92% 10%

Nuts 90% 39%

Chocolate candy 89% 29%

Pretzels 86% 46%

Tortilla chips 83% 41%

Taste>appearance

Olives 89% 52%

Taste>aroma

Coffee 91% 15%

Taco 91% 46%

French fries 95% 46%

Taste> texture

Peanut butter 91% 21%

Ice cream 95% 46%

Taste>appearance>aroma

Pizza 91% 61%

Steak 83% 59%

Taste>aroma>appearance

Chicken 89% 54%

BBQ ribs 89% 54%

Cinnamon rolls 84% 62%

Hamburger 93% 50%

Taste>appearance> texture

Cheesecake 87% 61%

Fig. 1 shows the distribution of the R2 statistic for the

respondents in the olive study. As one might expect

from these types of studies, there is a distribution of

goodness of fit statistics, with many of them at the high

end. A value of R2>0.66 corresponds to a goodness of

fit that would occur by chance only 5 times in a 100 tri-als for 36 predictor variables embedded in 60 cases

or concepts. More than 75% of the individuals show

this high R2 statistic. The results suggest that the quality

of the data is fairly high at the individual respondent

level.

4. Results

4.1. What sensory attributes drive snack food

�craveability� by �self-profiling�, and where do olives

fit in this pattern?

The data come from the larger-scale Crave It! Study,

where olives comprised one of 20 studies in the first

wave of the study. One can compare the proportions

e food

rance Aroma Texture

45% 33%

24% 40%

41% 40%

19% 7%

47% 19%

23% 35%

30% 42%

25% 39%

29% 35%

92% 11%

54% 17%

50% 39%

49% 61%

7% 50%

53% 21%

52% 38%

72% 26%

72% 26%

69% 17%

57% 18%

17% 58%


of respondents across foods and across sensory attri-

butes to better understand the uniqueness of olives.

The results are detailed in Table 1. Taste was always

most important. This study confirms the survey of

snacking patterns of the American Dietetic Association

that taste was the top reasons for choosing a snack(Cross et al., 1994). However, the three remaining sen-

sory inputs (appearance, aroma, texture) vary in their

importance. For the ‘‘classic’’ snack-foods like potato

chips, nuts or chocolate candy, taste was the primary

sensory input considered to be important. For ice cream,

the combination of taste and texture were the primary

sensory inputs considered to be important. For a more

complex product such as chicken, taste, aroma andappearance were all considered to be important. Only

coffee aroma is as important for coffee as taste is

for other foods and beverages. For olives, only appear-

ance was the most important sensory attribute. It is

important to note that the aspect of appearance as the

Table 2

Parameters of the utility model for olives, averaged across the total sample

Constant Constant

E03 Meaty ripened olives ready to stick on your finger or pop

E01 Whole black olives without the pits

E16 Extra large and nicely salted . . .with the stuffing you want

E12 Any size you want . . . small, medium, large, jumbo . . .you

E02 Green fleshy olives with that salty taste

E13 Premium quality . . . that great classic taste . . . like it used to

E05 Fleshy, jumbo olives with the bitter, salty note

E30 From Italy

E28 From your favorite grocery store

E35 Simply the best olives in the whole wide world

E18 So good . . .you practically have to lick your fingers & lips

E25 An outrageous experience . . . shared with family and friend

E34 Freshly prepared . . .especially for you

E24 A joy for your senses . . . seeing, smelling, tasting

E20 When you think about it, you have to have it . . .and after

E09 All sizes of green and black olives in olive oil with lots of

E10 Olives are a party pleaser

E17 You can imagine the taste as you walk in the door

E23 Now you can escape the routine . . .a way to celebrate spec

E27 It feeds THE HUNGER

E29 From California

E04 Medium size glistening olives with the briny taste of salt a

E19 Quick and fun . . .eating alone doesn�t have to be ordinary

E31 From Greece

E36 With the safety, care and cleanliness that makes you trust

E15 100% natural . . .and new choices every month to keep you

E21 Fills that empty spot in you . . . just when you want it

E26 Pure ecstasy

E14 You can just savor it when you think about it during wor

E22 When you�re sad, it makes you glad

E33 From Dean & Deluca

E32 Select from the Deli

E06 Cracked olives, a delicacy flavored with herbs, spices, and

E11 With a chilled glass of water . . . or carbonated beverage

E08 Olive paste with lots of chopped olives

E07 Small dark wrinkled olives marinated with hot pepper flak

�secondary key driver� is unique to olives across these

20 foods.

4.2. Positive and negative element utilities for

statements about olives based on the conjoint study

Dummy variable regression analysis generated one

additive constant and 36 utility values, one per element,

for each respondent. Across the total sample, the addi-

tive constant is 36, which means that 36% of the

respondents say that they �crave� the olive product

(i.e., they assign a rating 7–9), when they evaluate the

concept. Table 2 becomes more interesting when we

look at the ratings of positive and negative elements.The first thing to notice is that the strongest performing

element only scores a +9. The best-scoring statement

�Meaty ripened olives ready to stick on your finger or

pop in your mouth� will attract only 9% more respond-

ents if that strongest performing element is inserted into

Total sample

36

in your mouth 9

7

. . .garlic, pimentos, almonds . . .whatever 7

name it! 7

7

be 4

4

4

3

3

twice after each bite 3

s 2

2

2

you have it, you can�t stop eating it 2

herbs and spices 1

1

1

ial occasions 1

0

0

nd garlic 0

0

0

it & love it all the more �1

tantalized �1

�1

�1

k and school �2

�2

�2

�2

a tasty marinade �5

�6

�15

es �19


the concept about olives. In contrast, the worst perform-

ing element, �Small dark wrinkled olives marinated with

hot pepper flakes� will alienate 19% of the respondents.

In the case of olives, the larger negative utility for the

worst scoring element suggests that being wrong is far

worse than being right.The most positive elements are: �Meaty ripened olives

ready to stick on your finger or pop on your mouth� (util-ity = +9); �Whole black olives without the pits� (util-

ity = +7); �Extra large and nicely salted . . .with the

stuffing you want . . .garlic, pimento, almonds, . . .what-ever� (utility = +7); �Any size you want . . . small, medium,

large, jumbo (7); and Green fleshy olives with that salty

taste� (utility = +7). All of these elements deal with thesensory aspect of the product.

The most negative elements in descending order are:

�Small dark wrinkled olives marinated with hot pepper

flakes� (utility = �19); �Olive paste with lots of chopped

olives� (utility = �15); �With a chilled glass of water . . .orcarbonated beverage� (utility = �6); and �Cracked olives,

a delicacy flavored with herbs, spices and a tasty mari-

nade� (utility = �5). These poorly performing elementsalso speak to the product as well, except the statement

about water. It is important to note that although olive

paste scores poorly among consumers, it is used by chefs

for flavoring.

4.3. The winning elements from respondents who buy

olives in a specialty store

Table 3 illustrates how this group of consumers

shows especially strong responses to the concept ele-

ments. One element in particular, stating ‘‘Extra large

and nicely salted . . .with the stuffing you want . . .garlic,pimentos, almonds . . .whatever’’ does extraordinarily

well among this group (utility = 26). This particular ele-

ment paints a very strong word picture for this group of

respondents, as well as for those who say they crave ol-ives in a social situation (utility = 33). In contrast, this

element performs only modestly among the total sample

(utility = 7, Table 2). Those who buy olives at a specialty

store prefer ‘‘the Deli’’ instead of the ‘‘favorite grocery

store’’. These people buy the olives at the special olive-

store where they have a bigger choice of different ‘‘stuff-

Table 3

Utility value of winning elements respondents who buy at a specialty store

Buy olives at a specialty store

E16 Extra large and nicely salted . . .with the stuffing you w


E09 All sizes of green and black olives in olive oil with lots

E06 Cracked olives, a delicacy flavored with herbs, spices, a

E20 When you think about it, you have to have it . . .and a

E12 Any size you want . . . small, medium, large, jumbo . . .y

E32 Select from the Deli


ing you want’’ (E16) and ‘‘all sizes of green and black

olives’’ (E09).

4.4. Impact of hunger

The respondents rated their own level of hunger on afour-point hunger scale (from no hunger to high hun-

ger). The objective was to determine in subsequent anal-

ysis whether hunger potentiated the response to the

concepts as a whole, based upon the additive constant,

and/or whether there were specific concept elements that

showed greater sensitivity to hunger. Sixty percent of

respondents rated themselves as having either no or

low hunger whereas the remaining 40% rated themselvesas having medium or high hunger. These two groups

(low and high) are separately analyzed, with the results

appearing in Table 4. The additive constant for olives

is far higher for respondents who said that they were

hungry (constant = 45), as compared to respondents

who said they were not hungry (constant = 29).

The array of elements that do well in each segment is

quite diverse. For the low hunger group (rating of noneor low hunger), only two elements show high utilities. In

contrast, for the higher hunger group (self-rating of

medium or high hunger), far more elements show high

utilities. These elements comprise both statements of

product features as well as emotional promises. These

data suggest, therefore, that hunger has an impact on

desire for olives, and that the effect is not limited to re-

sponses to product features alone. Hunger level canmodify the response to other types of messages about

olives.

Some comparisons between utilities for the low and

high hunger groups suggest three findings worth pursing

in more detail through systematic experimentation (Ta-

ble 4).

1. Hunger increases the magnitude of the additive con-stant from 29 to 45. The conditional probability of

a person saying that he or she craves an olive

increases from 29% if the person is not hungry to

45%, if the person is hungry. Clearly, the greater

the perceived state of hunger, the greater the

�craving�.

ant . . .garlic, pimentos, almonds . . .whatever 26

16

of herbs and spices 16

nd a tasty marinade 13

fter you have it, you can�t stop eating it 12

ou name it! 12

10

10

Table 4

Effect of hunger on utility of concept elements for olives

Total No or

low hunger

Medium or

high hunger

Base size 101 67

Constant 36 29 45

Low hunger (hunger rating 0, 1)

E03 Meaty ripened olives ready to stick on your finger or pop in your mouth 9 10 7

E01 Whole black olives without the pits 7 7 7

High hunger (hunger rating 2, 3)

E12 Any size you want . . . small, medium, large, jumbo . . .you name it! 7 4 11

E16 Extra large and nicely salted . . .with the stuffing you

want . . .garlic, pimentos, almonds . . .whatever

7 5 10

E02 Green fleshy olives with that salty taste 6 4 9

E30 From Italy 3 �1 9

E18 So good . . .you practically have to lick your fingers & lips twice after each bite 4 1 8

E03 Meaty ripened olives ready to stick on your finger or pop in your mouth 9 10 7

E17 You can imagine the taste as you walk in the door 2 �2 7

E01 Whole black olives without the pits 7 7 7

E10 Olives are a party pleaser 3 0 7

E13 Premium quality . . . that great classic taste . . . like it used to be 4 3 6


2. Hunger increases the number of concept elements

that show a significant utility value (6 or higher).

Respondents reporting no hunger or low hunger

show only two elements reaching this significant util-

ity value. Respondents reporting medium hunger or

high hunger show 10 of the 36 elements reaching this

significant utility value.

3. The utility values for elements evaluated under hun-ger condition do not increase beyond the utility

values for elements evaluated under non-hunger con-

ditions. What changes is the number of strongly per-

forming elements, not the strength of performance.

4.5. Segmentation based on response patterns to concept

elements

Concept–response segmentation has been developed

for consumer-packaged goods and for services, respec-

tively (Moskowitz, 1996). The organizing principle is

the notion that there exists a limited set of segments or

basic groups in the population, defined by their pattern

of responses to concept elements. From an individual�spattern of responses one can put the individual intoone of a limited set of clusters, with the property that

individuals in a cluster or segment resemble each other

more than they resemble individuals in other clusters.

The allocation of individuals to a group is made strictly

on the basis of the pattern of their responses to the

concept elements. Subsequent allocation of individuals

to these segments is often made on the basis of an

assignment rule that relates segment membership toother, non-concept related variables, such as age, in-

come, etc.

The algorithm for clustering respondents uses a meas-

ure of distance or dissimilarity between pairs of respond-

ents. This measure is empirically defined, and chosen by

the researcher. The dissimilarity measure must be inde-

pendent of the types of rating scales used by the

respondents. For conjoint measurement, the researchers

chose the variable [1�R], where R is defined as the Pear-

son correlation coefficient between the 36 element utili-ties of one respondent and the same 36 element

utilities of another respondent. The Pearson R can vary

from a high of +1 meaning that the two respondents

show identical patterns of utilities, different by an addi-

tive constant or a common multiplier down through 0

meaning no relation, to a low of �1 (meaning that the

two respondents show opposite patterns). The corre-

sponding dissimilarity variable, [1�R], thus ranges froma high of 2 corresponding to an inverse relation, down to

a low of 0 corresponding to a perfect linear relation (Sy-

stat, 1997).

With this operational definition of the distance meas-

ure, respondents can then be clustered by k-means clus-

tering in order to create a limited set of groups. An

analysis of the different foods in the Crave It! Founda-

tion study suggested approximately three basic segmentsthat were easy to interpret (Beckley & Moskowitz,

2002). These segments are not, and need not, be of equal

size. The segments are best named and assessed by

inspection of the utility values. For the segments to be

meaningful, the positive elements must logically be asso-

ciated suggesting a single overriding theme.

The results show that three segments emerge among

olive respondents. The segments appear to be quiteindependent of each other, rather different from the ten-

dency toward linearity shown by the more conventional

SEGC1SEGC1

SEGC2 SEGC3

SEGC1

SEGC2 SEG

C2

SEGC1

SEGC3

SEGC2 SEGC3

SEGC3

Fig. 2. Scatterplot for utility values of three olive segments obtained

by clustering 168 respondents based upon the patterns of their utilities.


pairs of subgroups. Fig. 2 shows the scatterplot, andclearly reveals that there is no appreciable relation be-

tween the utility values obtained by any pair of the

segments.

The three segments identified show rather different

positive elements, and the utility values of the positive

elements are very high (see Table 5). Segment #1, with

Table 5

Strongest performing elements for three olive segments, clustered by the pat

Base size

Constant

Segment #1 (‘‘Classics’’)


E03 Meaty ripened olives ready to stick on your finger or pop in your m

Segment #2 (‘‘Elaborates’’)

E09 All sizes of green and black olives in olive oil with lots of herbs and


E04 Medium size glistening olives with the briny taste of salt and garlic

E03 Meaty ripened olives ready to stick on your finger or pop in your m


E16 Extra large and nicely salted . . .with the stuffing you

want . . .garlic, pimentos, almonds . . .whatever

E06 Cracked olives, a delicacy flavored with herbs, spices, and a tasty ma


E13 Premium quality . . . that great classic taste . . . like it used to be

E12 Any size you want . . . small, medium, large, jumbo . . .you name it!

E18 So good . . .you practically have to lick your fingers & lips twice afte

E25 An outrageous experience . . . shared with family and friends

Segment #3 (‘‘Imaginers’’)


E28 From your favorite grocery store

E16 Extra large and nicely salted . . .with the stuffing you want . . .garlic,pimentos, almonds . . .whatever

E18 So good . . .you practically have to lick your fingers & lips twice afte

E13 Premium quality . . . that great classic taste . . . like it used to be

42% of respondents, responds only to a simple message;

�whole black olives without the pits�. This group appears

to be the �Classics’’ segment, using the nomenclature

of the larger Crave It! Foundation Study. What is

important about the classic segment is its interest in

the simple, traditional messages. These individuals nei-ther respond to unusual or novel flavored products

nor do they react to alternative product forms. Segment

2 may be the ‘‘Elaborates’’ segment. ‘‘Elaborates’’ want

lots of flavors, including stuff that was added to the

product. These individuals want added flavors and fea-

tures of the olive. ‘‘Elaborates’’ also constitute a very

large segment, comprising 42% of respondents. Segment

3 appears to be the ‘‘Imaginers’’. ‘‘Imaginers’’ are distin-guished by their reactivity to elements other than prod-

uct features. ‘‘Imaginers’’ like to dream about the

product, and appear to be swayed by messages about

venue and history as well as being swayed by product

features. They constitute the smallest segment, 16% of

respondents. The three segments identified in this study

correspond to the three overarching segments for the

other food categories in the Foundation Study (Beckley& Moskowitz, 2002).

Table 5 shows clearly different winning elements for

the different consumer groups. Classics select the ‘‘clas-

sic’’ black olives, pure without pits or other ingredients

tern of their utility values

Total Classics Elaborates Imaginers

168 70 71 27

36 44 22 50

7 18 15 �42

outh 9 8 18 �12

spices 2 �12 23 �20

6 �13 22 15

�1 �17 18 �9

outh 9 8 18 �12

4 �11 18 5

7 �5 17 11

rinade �5 �24 17 �13

7 18 15 �42

4 �2 10 6

7 5 9 4

r each bite 4 �1 7 6

3 3 6 �5

6 �13 22 15

3 0 3 12

7 �5 17 11

r each bite 4 �1 7 6

4 �2 10 6


like salt (E02, E04, and E05) or flavored spices (E06). It

is not intuitively clear why the classics do not select with

positive intensity green olives. The classics do not care

about the place where they buy the olives nor is the sim-

ple statement of quality particularly appealing; �premium

quality with classic taste� (Utility = �2).The Elaborates positive selection indicates the ‘‘more

is more’’ framework relating to sizes (E05, E016) and

spices (E09, E06), while the product itself does not gen-

erate this intensity (black or green olives, E09, E02,

E01).

The Imaginers gather the most intensity of desire

from a location, a typical favorite grocery store, and

intensity is decreased by the idea of the classic black ol-ive without pits (E01).

5. Discussion

5.1. The uniqueness of olives in terms of sensory drivers

Each respondent had to identify up to three aspectsof the product that drive their intensity of craving.

Among these aspects are the sensory impacts of taste,

appearance, aroma, and texture, but there are others,

such as mood, situation, etc. This paper has only consid-

ered the sensory-based inputs. For foods it is the sensory

attribute of �taste� (including flavor) that drives intensity

of craveability. The secondary sensory attributes of

appearance, aroma, and texture are differentiators ofthe different foods. The definition of �importance� in

the analysis is that an attribute is important as a driver

of craving intensity if 50% or more of the respondents

select it from among the three aspects. For olives,

appearance, but not aroma or texture, is of secondary

importance. Olives are the only food out of the large

scale Crave It! Foundation Study for which taste

and then appearance are key, but aroma and textureare not.

5.2. Olives as the �healthy�, �good-for-you� snack

In the introduction to this paper the point was made

that manufacturers are looking for new products that

can be positioned as healthy snacks. The olive may be

ripe for this positioning, perhaps based upon its uniquedriver: appearance. In the self-profiling classification

questionnaire the unique combination of taste and

appearance as drivers appeared quite clearly. This

means, that for the olive, consumption may be driven

by �seeing the olive� either physically or in one�s mind.

Whether the respondent profiles himself as being hungry

or full, the driving element is appearance. Two appear-

ance elements do best among the hungrier respondents:�Any size you want . . . small, medium, large, jumbo . . .you

name it!�, and �Extra large and nicely salted . . .with the

stuffing you want . . .garlic, pimentos, almonds . . .what-ever�. An appearance element does well among the less

hungry respondents: �Whole black olives without the pits�.Another element, a call to fun and consumption,

also does well among the less hungry respondents:�Meaty ripened olives ready to stick on your finger or

pop in your mouth�. The combination of appearance ele-

ments and the call to consumption among the less hun-

gry respondents suggest that the olive could be the next

healthful snack, whose consumption need not be tied to

a meal.

5.3. Practical guidance for product development of the

healthy snack, and communication from these

consumer insights

A key benefit of conjoint analysis is the concreteness

of the messages. The segmentation is particularly useful

because it pulls apart the different groups, and suggests

different directions for development and communica-

tion. With this in mind it should become easier toachieve a degree of market success because through re-

search the marketer and developer have already ob-

tained some of the keys to the consumer mind

(Hoban, 1998; Trout & Rivkin, 2000). For example,

‘‘Elaborates’’ want a word picture of an exciting taste

to eat. Elaborates want a more spicy olive-snack. This

is the exact trend snack food producers see in the U.S.

snack food market. The consumer wants to have snackswith rich, robust and exotic tastes. It might well be that

the olive, properly flavored to appeal to the �Elaborates�represents that long sought after highly desired, healthy

snack. It is important to keep in mind, however, that the

features that drive response for the Elaborates may de-

creases desirability among the other segments. ‘‘Clas-

sics’’ are not particularly responsive to product

attributes except the promise of a good olive. ‘‘Imagin-ers’’ focus on the experience of eating olives. It may be

impossible to satisfy all of these segments with one prod-

uct, suggesting that perhaps several olive products

would be appropriate for a line of desirable healthful

snacks.

Although the macro research objective was to learn

about the map of the consumer�s mind, in business,

one of the key micro objectives is to discover what spe-cifically to say in marketing communications in order to

promote purchase and maintain repeat purchase of ol-

ives and snack foods. One of the lessons from the study

is that the key messages promoting purchase, on aver-

age, are those that deal with the product features (Hol-

lingsworth, 1996). Most American consumers need to

hear primarily about the product itself. There are how-

ever groups of people who will be motivated by very dif-ferent attributes.


5.4. The overriding importance of concept-response

segmentation to provide a set of �ordering� or�organizing� principles

Segmentation constitutes a recurrent theme in the

Foundation Study and in olives in particular. The exist-ence of segments is well known in the marketing field, so

that should cause no surprise (Hammond, Ehrenberg, &

Goodhardt, 1996; Lappin, Figoni, & Sloan, 1994). Seg-

mentation divides consumers into distinct groups with

similar needs and characteristics (Boyd, Westfall, &

Stasch, 1989). Segmentation procedures can also divide

people into different groups, based upon one or another

criterion (e.g. Customer needs, Lappin et al., 1994; Prod-uct use, Wellner, 2000; Demographics, Wellner, 2002;

and Ethnic origin, Gardynhen, 2001).

What is very interesting, however, is the continuing

emergence of the three radically different mind-sets.

Whether these segments (‘‘Classics’’, ‘‘Elaborates’’, and

‘‘Imaginers’’, respectively) represent basic groups or just

emerge as convenient statistical artifacts needs to be fur-

ther explored. In the case of foods as different as ham-burger and cheesecake, this segmentation appears to

work, and to drive different directions of product devel-

opment. Vis-a-vis other snack foods, olives appear to

possess this type of segmentation as well.

Despite all of the advantages of segmentation, the re-

searcher has to be aware of the fact that the assigning of

consumers to segments is fundamentally a statistical

tool, and not a statement of the way the world mayactually be constituted. According to Szmigin (2002),

people of similar geo-demographics do not necessarily

show the same behaviour. Geo-demographic segmenta-

tion is realistic because this is the way people are di-

vided. The criteria are eternal. Concept response

segmentation works at a different level, one that is more

theoretical. One cannot easily point to the characteris-

tics of a respondent and predict membership in one ofthe three segments. What is important is whether the

segmentation makes sense, whether it is general, whether

it leads to increased scientific understanding, and

whether it leads to improved product development and

marketing.

The segmentation into the three groups appears to

satisfy the four above-mentioned criteria. It is possible,

therefore, that the intermediate or even low-levelperformance seen with olives for the total panel may re-

sult from the intermixing of respondents from the three

segments. What one respondent finds acceptable in

terms of the product statements or benefits another

respondent from a different segment may find objection-

able. This difference is clearly seen by looking at winning

elements from one segment, to see how the other seg-

ments respond. Quite often respondents from one oreven both of the other segments dislike that positive

element.

5.5. Conjoint analysis as a new opportunity to

understand the consumer in the sensory and product

testing worlds

Conjoint analysis is often used in the early concept

development stages, where there are many options fromwhich to choose, and where the objective of the research

is to understand the �lay of the land�. Especially for this

initial stage, the Internet offers a particularly attractive

environment for product testing and market research.

First, the cost of creating and testing virtual prototypes

is considerably lower than that for physical prototypes

and many product concepts can be tested within the

same market research budget. Second, the Internet al-lows efficient and expedient access to respondents. Third,

if it is needed, the new technologies, such as VRML (Vir-

tual Reality Markup Language), streaming video, and

interactive sensory peripherals, allow visual, auditory,

and tactile information to be disseminated and retrieved

in rather powerful ways (Mosley-Matchett, 1998).

5.6. The usefulness of Internet research to better

understand products

These data suggest that the Internet can provide a

cost-effective and efficient venue to collect data and

therefore conduct high level consumer research. The

speed at which one can run studies allows the researcher

to move from expensive, difficult to execute tests, to-

wards more iterative tests to explore the consumer per-ception of products. The Foundation Study in general

and the olive study in particular, were created fairly rap-

idly. Most of the effort in Internet based research is ex-

pended on the up-front design of the study. The actual

field execution is fairly straightforward, as described

here. The high degree of consistency in the data is also

very encouraging.

The change in efforts allocated to design versus execu-tion, and the reduction of difficulty in execution means

researchers should be able to conduct a series of studies

to better understand a product. The Foundation Study

comprised a large-scale, cross-sectional view of many

product categories, from which the olive data represent

one study. There is every reason to believe that future

studies conducted in the same way, on the Internet, with

pictures and text, rather than text alone, can further in-crease understanding of products. Thus, one might look

at the Foundation Study as simply the first step in a ser-

ies of studies to understand a product in depth. This

form of semi-syndicated research has been suggested

as a desirable and needed outcome of this type of re-

search (Struse, 2000).

In conclusion, the sensory benefits, emotions, and

usage patterns ascertained for olives through the useof Internet-based conjoint analysis can serve as a model

for the study of other snack foods, enhance product seg-


mentation and contribute to the understanding of crave-

ability as a drive of consumer response.

Appendix A. The 36 concept elements for olives, and their

rationales

Category Rationale Olives

E01 Primary Basic physical attributes Whole Black olives without the pits

E20 Primary ** (continuum: basic to complex/detailed

physical attributes) in some cases . . . �healthy�Green fleshy olives with that salty taste


physical attributes)

Meaty ripened olives ready to stick on your finger

or pop in your mouth


physical attributes) in some cases . . . �real�Medium size glistening olives with the briny taste

of salt and garlic



Fleshy, jumbo olives with the bitter, salty note



Cracked olives, an Italian delicacy flavored with herbs,

spices, and a tasty marinade



Small dark wrinkled olives marinated with hot pepper flakes



Olive paste with lots of chopped olives

E09 Primary Complex physical attributes; details All sizes of green and black olives in olive oil

with lots of herbs and spices

E10 Mood Secondary Party pleaser/inviting Olives are a party pleaser

E11 Mood Secondary Beverages With a chilled glass of water . . . or carbonated beverage

E12 Mood Secondary with . . . Any size you want . . . small, medium, large, jumbo . . .you name it!

E13 Mood Secondary Premium quality/classic taste Premium quality . . . that great classic taste, like it used to be

E14 Mood Secondary Savor it . . . You can just savor it when you think about it during work and school

E15 Mood Secondary All natural/changing flavors 100% natural . . .and new choices every month to keep you tantalized

E16 Mood Secondary With all the extras you want . . . Extra large and nicely salted . . .with the stuffing you want.

Garlic, pimentos, almonds . . .whateverE17 Mood Secondary Imagine the taste . . . You can imagine the taste as you walk in the door

E18 Mood Secondary Lick your lips twice . . . So good . . .you practically have to lick your fingers

&lips twice after each bite

E19 Emotional Quick/fun/alone Quick and fun . . .eating alone doesn�t have to be ordinary

E20 Emotional Have to have it . . .can�t stop When you think about it, you have to have it . . .and

after you have it, you can�t stop eating it

E21 Emotional Fills that empty spot . . . Fills that empty spot in you . . . just when you want it

E22 Emotional Cheers you up . . . When you�re sad, it makes you glad

E23 Emotional Escape routine/celebrations Now you can escape the routine . . .a way to celebrate special occasions

E24 Emotional Multi-dimensional sensory experience A joy for your senses: seeing, smelling, tasting

E25 Emotional with family & friends An outrageous experience . . . shared with family and friends

E26 Emotional Ecstasy. . . Pure ecstasy

E27 Emotional Satisfies hunger . . . It feeds THE HUNGER

E28 Brand or Benefit Basic brands/experiences From your favorite grocery store

E29 Brand or Benefit Basic to premium brands From California

E30 Brand or Benefit Basic to premium brands From Italy

E31 Brand or Benefit Basic to premium brands From Greece

E32 Brand or Benefit Basic to premium brands Select from the Deli

E33 Brand or Benefit Premium brands/experiences From Dean & Deluca

E34 Brand or Benefit Fresh . . . for you . . .by you Freshly prepared . . .especially for you

E35 Brand or Benefit Best in world . . . Simply the best olives in the whole wide world

E36 Brand or Benefit Safety . . . With the safety, care and cleanliness that makes

you trust it & love it all the more

References

Abersold. (1998). Internet Healthy Eating Habits. http://www.unf.edu/

groups/snc/articles/health.

Baker, S., & Henry, R. R. (1987). Parents� guide to nutrition: healthy

eating from birth through adolescence. Reading, MA: Addison-

Wesley Publishing Company, Inc.

Beckley, J., & Moskowitz, H. R. (2002). Databasing the consumer

mind: the crave it!, drink it!, buy it! & healthy you! databases.

Anaheim, CA: Institute Of Food Technologists.

Boyd, H. W., Westfall, R., & Stasch, S. (1989).Marketing research text

and cases. Homewood, IL: Richard D Irwin.

Box, G. E. P., Hunter, J., & Hunter, S. (1978). Statistics for

experimenters. New York, NY: John Wiley.




Cross, A. T., Babicz, D., & Cushman, L. F. (1994). Snacking patterns

among 1,800 adults and children. Journal of American Dietetic

Association, 94, 1398–1403.

deJong, A. L., (Ed.). (1999). US market trends and forecasts.

Farmington Hills, MI: The Gale Group.

Gardynhen, R. (2001). Habla English? American Demographics.

Available: www.demographics.com.

Green, P. E., Carroll, D. J., & Goldberg, S. M. (1981). A general

approach to product design optimization via conjoint analysis.

Journal of Marketing, 45, 17–37.

Hammond, K. A., Ehrenberg, A., & Goodhardt, G. J. (1996). Market

segmentation for competitive brands. European Journal of Mar-

keting, 30(12), 39–49.

Hoban, T. J. (1998). Improving the success of new product develop-

ment. Food Technology, 52(1), 46–49.

Hoch, G. J. (2000). Snacks get hotter too––snack foods continue trend

toward hotter, more ethnically diverse products. Available: http://

www.foodprocessing.com.

Hollingsworth, P. (1996). Sensory testing and the language of the

consumer. Food Technology, 50(2), 65–69.

Klingbeil, A. (2003). PepsiCo keeps its plate full. The Journal News.

July 28, 2003.

Lappin, J., Figoni, P., & Sloan, S. (1994). A primer on consumer

marketing research: procedures, methods, and tools. Cambridge,

MA: National Transportation Systems Center.

Mintel International Group Ltd. (2003a). Kid�s snacking, US. Avail-

able: http://www.marketresearch.com.

Mintel International Group Limited (2003b). Healthy snacking, US,

July 2003. Available: http://www.preparedfoods.com.

Moskowitz, H. R. (1994). Concepts: their systematic design, develop-

ment and optimization. In Food concepts and products: just in

time development (pp. 1–69). Westport, CT: Food and Nutrition

Press.

Moskowitz, H. R. (1996). Segmenting consumers worldwide: an

application of multiple media conjoint methods. In Proceedings

of the 49th ESOMAR congress, Istanbul, pp. 535–552.

Moskowitz, H. R., & Martin, D. G. (1993). How computer aided

design and presentation of concepts speeds up the product

development process. Proceedings of the 46th ESOMAR congress,

Copenhagen, pp. 405–424.

Mosley-Matchett, J. D. (1998). Leverage the Internet�s research

capabilities. Marketing News, 32.

Oltersdorf, U., Schlettwein-Gsell, D., & Winkler, G. (1999).

Assessing eating patterns––an emerging research topic in nutri-

tional sciences: introduction to the symposium. Appetite, 32, 1–7.

Pelchat, M. L., & Schaefer, S. (2000). Dietary monotony and food crav-

ing in young and elderly adults. Physiology & Behavior, 68, 353–359.

Roberts, Jr., W. A. (2001). Category analysis: salty foods––worth its

salt. Available: http://www.preparedfoods.com.

Schutz, H. G. (1964). A food action rating scale for measuring food

acceptance. Journal of Food Science, 30, 202–213.

Schutz, H. G., & Martens, M. (2001). Appropriateness as a cognitive–

contextual measure food attitudes. In People, food and society a

European perspective of consumer food choices. New York, NY:

Springer-Verlag.

Struse, D. (2000). Researcher�s Top 25 influence. Marketing Research,

11(4), 5–9.

Sullivan, D. K., Legowski, P. A., Jacobsen, D. J., Synder Heelan, K.

A., Johnson, S. L., & Donnelly, J. E. (2002). A low-fat afterschool

snack improves the nutritional quality of elementary school

children�s diets. Journal of the American Dietetic Association,

102(5), 707–710.

Systat (1997). Systat, the system for statistics. Evanston, IL: SPSS

Corporation.

Szmigin, I. (2002). Consumer behaviour. Available: http://bss2.bha-

m.ac.uk/business/teaching/szmigin/introduction.ppt.

Trout, J., & Rivkin, S. (2000). Differentiate or die: survival in our era of

killer competition (pp. 2–3, 65–73). New York, NY: John Wiley &

Sons, Inc.

Wellner, A. (2000). Brand strategies: profit in loss. American Demo-

graphics 2000. Available: http://www.demographics.com.

Wellner, A. (2002). The female persuasion brand strategies: profit in

Loss, American Demographics 2002. Available: http://

www.demographics.com.

Wittnk, D. R., & Cattin, P. (1989). Commercial use of conjoint

analysis: an update. Journal of Marketing, 53, 91–96.

http://www.demographics.com

http://www.foodprocessing.com

http://www.foodprocessing.com

http://www.marketresearch.com

http://www.preparedfoods.com

http://www.preparedfoods.com

http://bss2.bham.ac.uk/business/teaching/szmigin/introduction.ppt

http://bss2.bham.ac.uk/business/teaching/szmigin/introduction.ppt




Sensory benefits, emotions and usage patterns for olives: using Internet-based conjoint analysis and...

Documents

Transcript of Sensory benefits, emotions and usage patterns for olives: using Internet-based conjoint analysis and...