• No results found

Willingness to pay for nutritional value in spread food.

N/A
N/A
Protected

Academic year: 2021

Share "Willingness to pay for nutritional value in spread food."

Copied!
37
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Willingness to pay for nutritional value in

spread food.

Choice-Based Conjoint Analysis of Peanut Butter with the aim

of finding consumer differences

.

(2)

Willingness to pay for nutritional value in spread food.

Choice-Based Conjoint Analysis of Peanut Butter with the aim

of finding consumer differences.

Author: Nicole Clemens

Department: Faculty of Economics and Business Qualification: Master thesis

Completion date: 20th June 2014

Poelestraat 40A, 9712KC Groningen

0617330289 - n.j.m.clemens@student.rug.nl Student ID: S2582902

(3)

Abstract

This study looks at the situation in the food market, in which companies struggle to find new ways to sell products. Functional foods is a way to attract new customers, but because of limited budgets, companies must know what to focus their attention on. With the health trend in mind, the research performed a choice-based conjoint analysis online, letting respondents choose peanut butter based on fat, protein, carbohydrates, sugar and price. Based on this, different willingness-to-pay levels were found for the nutritional values. These were even significantly different over five segments found with Latent Gold. The segments can be described as (1) International Sporty People, (2) International Community, (3) Dutch Sugar Avoiders, (4) Pleasure Seekers and (5) Female Carb Counters. Overall this provides evidence that different people look at food in different ways. Especially those people on a diet, or who are engaged in sports, will look at the nutritional value. Those people who are price-focused however, will only continue to look at price.

Preface

This thesis was written to conclude my MSc Marketing studies at the University of Groningen. The reasoning for a nutritional topic was based on personal interests in the field, together with trends seen in marketing. Seeing many companies only promote unhealthy food, it was of big interest to me how to change this situation. Consumers do not always know what is best for them, and with the expansion of processed foods, I fear the worst. Marketing is a big influencer in food choices, so surely it can also be used ‘for good’. I wanted to contribute to this potential future movement by finding out whether people would be open-minded about eating healthier, and what their knowledge of food is so far. Based on that, and differences between groups, people can be educated and hopefully companies would also be commercially interested in producing nutritional foods.

(4)

Table of Contents

1 Introduction: ... 5

2 Relevant Literature & Hypotheses ... 6

2.1 Nutritional value ... 6

2.1.1 Fat ... 7

2.1.2 Protein ... 7

2.1.3 Carbohydrates ... 8

2.1.4 Sugar ... 8

2.2 Control variables: consumer characteristics ... 9

2.2.1 Health interest ... 10

2.2.2 Pleasure ... 10

2.2.3 Physical Activity & diet ... 11

2.2.4 Demographics: ... 11

3 Methodology ... 12

3.1 Procedure: ... 13

3.1.1 Sample and collection: ... 14

3.1.2 Sample descriptive ... 15 4 Results ... 16 4.1 Reliability ... 16 4.2 Conjoint ... 17 4.2.1 Willingness to pay ... 18 4.3 Segment descriptions: ... 20 4.4 Hypothesis rejection ... 23 5 Managerial implications ... 24

6 Limitations and future research... 25

7 Conclusion ... 26

8 References ... 28

(5)

1 Introduction:

Both policy makers and companies are urged to deal with the dramatic rise in obesity levels and poor diets in large parts of the world. Policies are changing and created to reverse these developments, by funding healthy initiatives, informing consumers and tightening of legislation with respect to media and labeling (PHEIAC, 2013). It can be expected that these types of policies will only increase over the years, as the severity of obesity consequences, and the damage to societies will as well. The private sector would therefore be wise to comply to regulations, in order to avoid fines and restrictions, and avoid risking legal claims from consumers, which are linked to negative publicity and thus negative effects on sales (Henson & Hooker, 2001).

However, even if companies do not care about these developments, another one is taking place: the one of changing consumer perspectives. Consumers increasingly link food intake directly to their health, as it can be used to influence illness and mental and physical wellbeing. This has led to the trend of food producers using technology to create healthier foods, which started in the early 1980s in Japan, with the introduction of functional foods, those foods with health claims. Other regions such as the US and EU also found out there was a need and demand for these types of products, and thus the food category spread out worldwide (Siro, Kapolna, Kapolna & Lugasi, 2008). The biggest benefit is that functional foods can attract new customers and revitalize mature segments (Heasman & Mellentin, 2001). So the changes for companies are not only mandatory, but also necessary to stay sustainably competitive in a global world with varying needs.

However, there are also downsides; higher margins are necessary to get a return on investment from all the R&D expenses, marketing and diseconomies of scope (Bonanno, 2013). This increases prices, which most consumers see as a disincentive to buy the product. Luckily, these price increases could also be a sign of a great quality product, which genuinely is healthy (Bonanno, 2013; Wedel & Leeflang, 1997). Considering the limited resources companies have, it is essential to make the best decision possible, which also diminishes all the risks of product introductions and changes.

(6)

include the macro nutritional values, which are described on each product. Thus, this paper aims to find out which of the differences in nutrition are most valued by consumers, based on

their willingness-to-pay.

Firstly the literary background of food improvements and consumer behavior is described, which leads to the hypotheses and conceptual model of the research. Secondly, the methodology is looked at, by explaining the rationale behind it based on literature and interests. Thirdly an in depth analysis is made of the results which consequently will be discussed, together with the research’s limitations and directions for the future.

2 Relevant Literature & Hypotheses

2.1 Nutritional value

Consumers are interested in what they eat. They associate food with their health and consequently, there is a widespread interest in nutritional information on food packages. In the absence of nutritional labels, a lack of efficiency is apparent for shoppers. Most of them do not have the knowledge of nutritional content of products and cannot guess their worth. Since the introduction of standardized labels, comparison has become easier (Berning, Chouinard, Manning, McCluskey & Sprott, 2009). Consumers are willing to pay 11 per cent more for a box of cookies with a nutritional label than for one without (Loureiro, Gracia & Nayga, 2006), meaning they truly value the information, mainly because of a guarantee of food quality and safety.

Interest in nutritional information on products varies across situations and products (Grunert & Willis, 2007). It was already known that local and organic products generally generate positive willingness to pay (Hu, Woods & Bastin, 2009), but there is still not a lot of insight on how nutritional labels are being used and how it affects consumer behavior (Grunert & Willis, 2007). These are less obvious and require some degree of knowledge, whereas ‘organic’ and ‘local’ are considered positive.

(7)

nutritional information. They found out that willingness to pay is negative for those nutrients which a product is already rich of. Also, consumers often make a trade-off between taste and nutrition. Nutrition matters when breakfast cereal has a lot of fat or sugar and when potatoes consist of a lot of salt. Taste seems to matter more when breakfast cereal and bread have a lot of salt. Overall, nutritional information tends to increase the intake of fiber, decrease total calories, saturated fat, cholesterol and sodium. Loreiro, Gracia & Ngaya (2006) find it is linked to what consumers actually look at. Based on a TNT survey (CCFN, 2009, p.31), it was stated that “most Canadians consider the amount of whole grains (82%), fiber (81%), protein (76%), total fat (75%), calories (74%) and sugar (73%)”. These were taken from qualitative studies, but in order to truly get to find out the effects and their strength, quantitative research is needed.

2.1.1 Fat

Fats are a fuel source and are mainly used as a storage form of energy. They have become one of the most avoided nutrients. As consumers mainly get their information from health authorities, which have communicated strong links between fat intake and several chronic diseases, fat is considered unpopular. More and more reduced-fat foods are available on the market, which have increased the importance people place on low-fat alternatives (Roininen, Lähteenmäki & Tuorila, 1999). Additionally, according to Burton and Biswas (1993) and

Russo, Staelin, Nolas, Russell & Metcalf (1986), fat is undesirable and therefore decreases the purchasing likelihood of consumers. Consumers find out about these through the nutritional labels and thus it is expected consumers want less fat in their food. Based on Thunström and Rausser (2008)’s research that richness in a nutritional value lead to negative effects on price: As fat numbers are high in most foods, and because of general ideas that low-fat foods are healthier, it is expected that further increases, will lead to a negative effect on price.

H1: Increase in fat will have a negative effect on willingness-to-pay.

2.1.2 Protein

(8)

protein intake (NPD, 2014) Protein thus is considered good. Based on Thunström and Rausser (2008)’s research that richness in a nutritional value lead to negative effect on WTP: As protein numbers are relatively low in most foods, as well as the positive reputation of protein, it is expected that further increases, will lead to a positive willingness-to-pay.

H2: Increase in protein will have a positive effect on willingness-to-pay.

2.1.3 Carbohydrates

Carbs are a fast source of energy. Decreased carbohydrate intake has been linked to weight loss benefits and reduction of metabolic risk factors (Hu, Mills, Demanelis, Eloustaz, Yancy, Nelly, He & Bazzano, 2012). The Food and Drug Administration of the U.S. (FDA, 2009) found that consumers perceive products which are low in carbohydrates as beneficial to their health, based on weight management. Many health authorities advise a healthy diet of which 45-65% of daily calories come from carbohydrates, however, the trend towards low-carb diets creates an opposing and even slightly confusing situation. Considering the slight negative reputation of carbs, it is expected low-carbs are generally preferred, also seen by reduced carb intakes in Canada (CCFN, 2009). Based on Thunström and Rausser (2008)’s research that richness in a nutritional value lead to negative effect on price: As carbohydrate numbers are high in most foods, it is expected that further increases, will lead to a negative effect on willingness-to-pay.

H3: Increase in carbohydrates will have a negative effect on willingness-to-pay.

2.1.4 Sugar

Sugar has become a concern for consumers over the years. With the introduction of sugar-substitutes, this realization has even become more visible (CCFN, 2009). Weaver and Finke (2003): looked at the sugar consumption of people after checking the labels, however, they found no large difference. This could be because the sugar in absolute values is often considerably lower compared to the macronutrients. Based on Thunström and Rausser (2008)’s research that richness in a nutritional value lead to negative effect on WTP: As sugar numbers are relatively high in most foods, it is expected that further increases, will lead to a negative effect on willingness-to-pay.

(9)

2.2 Control variables: consumer characteristics

Consumers differ in their attitudes towards food. Identifying segments consequently would aid in both product marketing and a healthier society. Simultaneously, willingness to pay thus differs for various consumers groups. To start off, food choice is influenced by attitudes towards nutrition. Demographics, health-consciousness and trust in health-related information influence consumption of foods (Krystallis, Maglaras & Mamalis, 2008). Food choice in turn is influenced by the quality of the food, habit/routine of people, price, what family/spouse will eat, trying to eat a balanced diet, taste of food, convenience preparation, presentation/packaging, slimming effect of product, vegetarian/other diet, prescribed diet, additives/coloring and cultural/religious/ethnic background (Lappalainen, Kearney, Gibney, 1998). Overall, it means many factors influence food decisions, and consumers can be tried to be grouped. Hu, Woods & Bastin (2009), although looking at willingness to pay for nonconventional attributes instead of the nutritional value, found out that there was a general heterogeneity in consumer preferences for blueberry products. This could indicate that in general, food characteristics other than the brand and taste can have different effects among different customers. Additionally, some people have conflicting preferences for nutritional labels, as some prefer ease of use, but other want full information, or simply feel pressured to behave in a certain way due to the labels (Grunert & Willis, 2007). Regarding preferences for labels of food characteristics, Grunert & Willis (2007) suggest ease of use is important. A lot of information about the various benefits of the product and nutrition in different ways would oppose this finding. Thus, a single food benefit, stated in a clear manner, would most likely be most appreciated. This because it makes interpretation easier and gets the point of the single benefit across easily. Additionally, research has found out that consumer mainly make their food choices based on simple to interpret claims. In turn, ‘fat-free’ and ‘low-fat’ claims are often misunderstood (Tarabella & Burchi, 2012).

(10)

care about certain food attributes, or felt unaware (CCFN, 2009). As a consequence of that, it can be expected that consumers differ in preferences.

The question remains, based on which characteristics should these consumers differ. Previous research has found a multiple of interesting segmentation methods.

2.2.1 Health interest

Consumers’ interest in health has been seen as an indicator for segmentation. Contento, Michela & Goldberg (1988) found “five health orientations in adolescents: hedonistic, social/environmental, personal, peer-supported and parent-supported health orientations”. (Roininen et al, 1999, p2.).Bower, Saadat, Whitten (2003) already found that buying spreads with proven health benefits is often done because of the health factor for some groups. Moreover, some consumers truly aim for a feeling of variety when it comes to food. A clear majority of EU citizens consider their diets healthy enough and have no plan of changing it. Lack of knowledge of nutrition is not commonly cited as a barrier, yet health interest is. Considering the results of Roininen et al (1999), general health interest is part of healthy eating decisions, and a good indicator for possible segments.

2.2.2 Pleasure

(11)

2.2.3 Physical Activity & diet

Lifestyles determine attitudes and motivations. It is also related to how much importance consumers place on food. Those people whose aim is to be healthy, will not only try to eat in that manner, but also engage in physical activities. Those who are more physically active will have a more positive attitude towards eating healthy (Hearty, McCarthy, Kearney & Gibney 2007). People who are very engaged in sports also have different food needs, including more energy and protein to maintain muscle.

A second option are those people who are aiming to lose weight. They will exercise more, and are on a weight-loss diet to get to their goals. Because of that reasoning, they may have specific needs or opinions on foods, which could influence the results.

2.2.4 Demographics:

Demographics play an important role not only in consumption choices, but health choices as well. The two main demographics this study will look at are age and gender. Previous research indicated differences in health consciousness between men and women. Not only are women more interested in health, and show healthier behavior patterns then men, they avoid more fatty foods. However, males engage more in physical activity than females (Roininen et al, 1999). Moreover, they accept dietary changed more willingly and are more aware of health issues.

(12)

Figure 1 Conceptual Model

3 Methodology

This paper’s objective is to calculate the value consumers place on nutritional compounds in order to both adhere to consumer trends as well as policy changes. This is done so by calculating the willingness to pay for nutritional values for different segments by using conjoint analysis. Willingness to pay can be defined as the price at which a consumer is indifferent between purchasing and not purchasing (Moorthy et al, 1997). It is especially very useful when deciding on product line or assortment extensions or changes, as consumers drive sales (Gensler, Hinz, Skiera & Theysohn, 2012).

In order to calculate willingness to pay, different techniques are possible. Choice-based conjoint analyses have gained popularity in the recent years. One of the biggest benefits of it is the option to consider possibilities of extreme behavior of always- and never purchasers. These often bias the willingness to pay estimates quite strongly and thus are better to be accounted for (Gensler, Hinz, Skiera & Theysohn, 2012).

The following equation represents the final WTP model:

(13)

3.1 Procedure:

In order to create a realistic experiment for participants, the choice of a product had to be made. It had to be a product which could be processed, would last for a longer amount of time to avoid last-minute purchase decisions, and which was generally accepted across multiple consumer and health groups. Peanut butter is eaten by both health conscious consumers, as well as consumers looking for a great taste. Also, the attributes made sense for the product. With the chosen attributes in place, the amount and choice of levels had to be chosen consistently. Of importance is a similar amount of levels for all attributes, and creating a realistic range, while still showing clear differences. Similarly, for price, to be able to calculate willingness to pay, prices must reflect actual choices (Gensler, et al, 2012). The chosen levels are visible in table 1. The decision to go for three levels was to avoid too many choices which would lead to confusion, as the decision making on nutritional values is considered relatively intense. The base levels were taken from a regular peanut butter jar, and the adjusted numbers were taken by comparing to other brands and making it slightly more extreme, but still within an acceptable possibility range.

The survey starts off with the choice-based conjoint analysis part. Respondents are shown 8 choice-sets with 3 choices of peanut butter. The participant is asked which option they would choose, or whether they would go for the non-option. As brand and texture preferences can be key to choices, while it is not the aim of the research, these must be held constant. It is therefore mentioned that brand and crunchiness/creaminess is the same for all alternatives.

(14)

Table 1 Attributes and attribute levels

Regular ‘Low’ ‘High’

Protein 21 15 27

Carbs 16 10 22

Sugar 6 4 8

Fat 54 42 66

Price €1.85 €1.60 €2.10

Figure 2 Questions on general health interest

1.Reversed: The healthiness of food has little impact of my food choices.

2. I am very particular abou the healthiness of food I eat.

3. I always follow a healthy and balanced diet.

Questions on pleasure

1. Reversed: I finish my meal even when I do not like the taste of a food.

2. When I eat, I concentrate on enjoying the taste of food.

Question on price:

1. When buying food, I buy the cheapest item.

3.1.1 Sample and collection:

The survey was distributed online through social media to ensure participants being able to do the survey at their own convenience. This was necessary as a limited budget required people to voluntarily participate, and with the cognitive loading associated with the survey, this would be hard. Out of interest, both a Dutch and English survey were made. Fellow students as well as people actively interacting with a sports club were the main respondents. A total of 139 people participated in the Dutch survey, while 36 filled out the English version. After checking for biased responses, 2 people were deleted from the Dutch survey.

Considering Johnson and Orme (1996)’s rule-of-thumb with regard to sample size in Choice-Based Conjoint Analysis:

𝑛𝑡𝑎

𝑐 ≥ 500

(15)

the largest amount of levels for any one attribute. For this study, it would lead to the following calculation:

173 ∗ 8 ∗ 3

3 = 1384

As 1384 is larger than 500, the sample size used is sufficient.

3.1.2 Sample descriptive

With 173 respondents gathered over the period of a few days, a pattern is visible in who has responded. The majority (70.8%) being Dutch, with some foreigners who mainly have lived in the Netherlands, the sample is slightly representative of the Netherlands.

Agewise, the majority of the respondents are quite young. Up to 20% is 22 years of age or younger, and 50% of the respondents are not older than 25. The highest quartile are represented by people who are 30 years or older. Consequently, this is a young sample, which is understandable as the survey has been distributed through social media, which has a younger crowd of people.

Graph 1: Age Graph 2: Sports

(16)

and 25% actively engages in sports for 8 or more hours per week. This is not representative of the Dutch population, in which people are less active.

Surprisingly, when asked whether the people were on a weight loss diet, 87% of the respondents indicated they were. This may have been related to the sample in which many sport enthusiastic people have participated. These people are probably more likely to be on a specific diet, and considering the timing of the survey (April), many of them would be trying to lose weight for the summer.

Graph 3 % Respondents on a diet

4 Results

4.1 Reliability

(17)

which would describe the negative link. For that reason, it was decided to simply leave out the reversed question from further analysis.

When testing the internal consistency of the two pleasure questions, an even lower cronbach’s alpha appeared. With a mere 0.235 (table b in Appendix), the use of combining the two was low. Explanations for this could be that the items measured separate things after all. People are able to both want to focus on the taste of food when eating, but will still finish their meal even if they do not like it. Therefore the items were kept in, but separate.

4.2 Conjoint

The online survey tool used, provided direct conjoint results, visible in figure 3. It shows a comparison between the Dutch and English survey, with clear differences. In the Dutch survey, which is mainly represented by a group of physically active people, protein and sugars are the main drivers of choices. In the English survey, Fat takes an extremely strong lead. However, these results, no matter how interesting they are, are biased due to a different international understanding of ‘peanut butter’. While in the Netherlands peanut butter is considered relatively healthy, comments from the English survey have proven the fierce rejection of the product as it is considered extremely unhealthy and unnatural.

Figure 3 Summary Basic Conjoint

(18)

In the model statistics, fat, protein, sugar and price are all significant at the 1% level, while carbohydrates is significant at the 95% confidence interval, meaning there are clear differences in the way respondents reacted to the nutritional values (visible in appendix C)

4.2.1 Willingness to pay

Looking at the results of the conjoint analysis, the output of utilities can be used to measure the willingness to pay per segment. This is possible in two ways. The first way involved a parth-worth estimation of the WTP per segment, per attribute and per level. This is especially useful as in a few cases the utilities do not seem linear. In the graphs below it is visible that for class 1, fat is slightly linear, for class 2 carbs are linear, for class 4 fat is non-linear, for class 5 protein is non-linear and sugar is slightly non-linear. These are all based on table 5. However, most interesting would be to also provide a WTP formula for all segments. Table 6 shows the linear amounts, although what must be taken into account is that for Segment 1 and 3 price is positive, while for the other three segments, price is negative. It means that the WTP will be interpreted after multiplying by -1 in those three cases..

LL BIC(LL) AIC(LL) AIC3(LL) CAIC(LL) Class.Err. R²(0)

(19)

FAT PROTEIN

CARBOHYDRATES SUGARS

Graph 4: Part-worth WTP per nutritional value

Table 5: Part-worth WTP

Fat Class1 Class2 Class3 Class4 Class5

42 0.012 -1.298 -0.841 -0.018 -0.892 54 -0.037 -0.192 0.235 -0.037 -0.132 66 0.025 1.490 0.606 0.055 1.024 Protein 15 -1.572 0.348 -1.285 0.043 1.031 21 -0.161 -0.171 -0.360 -0.018 -0.681 27 1.734 -0.176 1.645 -0.025 -0.351 Carbohydrates 10 0.837 0.000 0.483 -0.017 -1.169 16 -0.064 -0.251 -0.135 0.004 -0.119 22 -0.773 0.251 -0.348 0.013 1.288 Sugars 4 0.334 -0.421 3.926 -0.042 -0.510 6 0.222 -0.065 0.726 0.006 0.345 8 -0.556 0.486 -4.653 0.036 0.164 -0,002 -0,001 0,000 0,001 0,002 15 21 27

Class1 Class2 Class3

Class4 Class5 -0,002 -0,001 0,000 0,001 0,002 15 21 27

Class1 Class2 Class3

Class4 Class5 -0,002 -0,001 0,000 0,001 0,002 10 16 22

Class1 Class2 Class3 Class4 Class5

-0,010 -0,005 0,000 0,005

4 6 8

Class1 Class2 Class3

(20)

Table 6: Linear WTP

Class 1 Class 2 Class 3 Class 4 Class 5

Fat 0.003 0.116 0.060 0.004 0.008 Importance % 1.0 59.0 10.5 34.3 27.6 Protein 0.276 -0.044 0.244 -0.006 -0.143 Importance % 56.3 11.1 21.3 25.3 24.7 Carbohydrates -0.134 0.042 -0.069 0.003 0.205 Importance % 27.4 10.6 6.0 11.2 35.4 Sugars -0.223 0.227 -2.145 0.019 0.214 Importance % 15.2 19.2 62.2 29.2 12.3 4.3 Segment descriptions:

(21)

Segment 1: International Sporty People

The first segment is quite non-Dutch and has mainly males in it which are on a diet and engage in more sports. What stands out is that even they don’t always finish their meals when they dislike it, they are not solely focused on the taste of food. When they decided upon the buying of peanut butter, their main motivator was protein, followed by carbohydrates. They are less interested in price, and even willing to pay a bit extra for the right product. They also have hardly any interest in fat.

With the use of the price function, the linear willingness to pay formula can be established for this segment. Based on the output the formula would be:

Table 7: Segmentation Information International Sporty People International Community Dutch Sugar Avoiders

Pleasure seekers Female carb

(22)

WTP = 0.003*Fat + 0.276*Protein – 0.134 Carbohydrates – 0.233*Sugar.

Segment 2: International Community

The second segment is a typical representation of the international community, in which peanut butter is considered quite unhealthy. Quite some are on a diet even though they exercise less than most other segments. They are not very focused on the taste of food, and will finish their meals. They also consider themselves quite healthy. They do get influenced by price, but their main predictor was fat. They hate high fat, and some have even stated that they think peanut butter is unhealthy because of the “butter” and fatty part of it. They also look at sugar and prefer some lower sugar, but their main aim is to get the lowest fat amount.

Considering they have a negative price function, they WTP will be interpreted the other way around as compared to what table 6 states:

WTP = -0.116*Fat + 0.044*Protein -0.042*Carbohydrates -0.227*Sugars

Segment 3: Dutch Sugar Avoiders

This segment stands out a lot. Extremely Dutch, slightly older, a bit sporty and some diet interest, they have their preferences clear. They consider themselves very health oriented and not focused on the taste of food too much. They even have no real interest in cheapest product and seem to care about nutrition. Their choice for food has an extreme bias towards sugar. It is the strongest of all effects within the sample, which shows their aversion towards high sugar amounts. Additionally they have a preference for higher protein, but it cannot be compared to the choice of sugar. They are even willing to pay a little extra just to avoid the sugar. Considering this segment is overwhelmingly Dutch, it could mean that Dutch people often consider sugar as the worst nutritional value out there.

Leading to a WTP formula of:

WTP = 0.06*Fat + 0.244*Protein – 0.069*Carbs – 2.145*Sugars

Segment 4: Pleasure seekers

(23)

looking at price, they made the majority of their choices. They have even stated to not be interested in health too much, and extremely focus on the taste of the food that they eat. They however also want to go for the cheapest item most of the time, and are not very likely to be on a diet. Additionally they don’t really do sports that much and are quite the younger segment. These could be considered the relatively ‘normal’, young Dutch persons, most likely students with a budget, who don’t do a lot of sports. Their aim is to enjoy life and some calories more or less do not matter.

WTP= -0.004*Fat + 0.006*Protein – 0.003* Carbohydrates - 0.019* Sugars

Segment 5: Female carb counters

Carbohydrates have been big in the media for a while. When wanting to lose weight, many women’s magazines have encouraged readers to look at the carbohydrates. In this segment, which is mainly the slightly older female segment, which do not like sports a lot, diet seems unpopular. Therefore they may have been influenced by older media quite a bit, as they do not want or feel the need to do their own research. Their main choice in deciding on food is carbohydrates, although fat also seems quite important. However, in the other groups carbohydrates were of smaller importance than in this case, so it shows a pattern for this group. Additionally they say not to really be interested in health, although taste is not everything for them.

WTP = -0.008*Fat + 0.143*Protein - 0.205*Carbohydrates – 0.214*Sugars

4.4 Hypothesis rejection

(24)

negative effect on WTP. For none of the segments, this could be rejected. However, it was not of high importance to most. For the Female Carb Counters it accounted for 35% of their nutritional choices, and it has the strongest WTP. H4 expected increases in sugar to have a negative effect on WTP. Again, in all segments this is valid. This means overall none of the hypotheses can be rejected, except for fat in two cases.

Table ? below provides an overview of the hypotheses rejections:

International Sporty People International Community Dutch Sugar Avoiders Pleasure Seekers Female Carbs Counters H1: Rejected Not Rejected Rejected Not Rejected Not Rejected H2: Not Rejected Not Rejected Not Rejected Not Rejected Not Rejected H3: Not Rejected Not Rejected Not Rejected Not Rejected Not Rejected H4: Not Rejected Not Rejected Not Rejected Not Rejected Not Rejected

5 Managerial implications

With policies changing to promote healthier eating habits, and consumers reacting to it, companies would be wise to react to these changes. Even better, they should proactively promote these changes themselves, as it gives the food industry opportunities to attract new customers and revitalize mature segments (Heasman & Mellentin, 2001). Thunström and Rausser (2008) researched how willingness to pay is lower for those nutrients a product is already rich of. However, this research found out that the effect of the WTP differs for consumer groups. First of all, some consumers are very price sensitive, and will not even look at the nutrients. They know their basic health information, but probably would need to be forced to look at nutrition. And even then, they will still mainly look at price. This means that companies who want to only be low-cost competitors, don’t need to participate in this health trend at the moment. They should highlight their low prices, and target at younger males. These are also the ones who have a stronger focus on the taste of food.

If a manager decides to respond to the health trend, they need to find out who their target market is. If they want to aim at sporty people, they should highlight the protein amount. Additionally, sugar seems to be important to those people who care about protein as well, especially when comparing WTP. Therefore a combination of protein and sugar might stimulate sales even more.

(25)

market consists of less physically active women who do not like diets, the choice of advertising for lowered carbs will be good.

Interestingly, fat does not have a very strong effect on WTP. It might be product dependent, but it seems that people tolerate fats. Managers should analyze whether this is also the case for their product. Perhaps in some cases, advertising for more fats could even be beneficial.

6 Limitations and future research

Even though this research has given interesting results, there have been limitations. Initially the issue is that no product consists merely of only fat, protein, carbs and sugar. Nutritional labels give information on salt, fiber, saturated fats, and most importantly, total calorie count. Some respondents have even stated in the survey that they usually look at the calorie amount if they have to decide on a product based on nutrition. Others are very brand-focused and would consider the lack of this information as a limitation. In the future, the conjoint can be extended to include more nutritional information. Perhaps some people will still look at, for example, sugar, instead of total calorie count. These results can be very interesting. However, great care needs to be taken to avoid cognitive overload of information. Thus, perhaps a smaller sample would be more fitting, or different research methods.

Another limitation is the fact that this research only looked at peanut butter. This makes it harder to generalize the results over all product categories. They might still be applicable to long-lasting spread foods, but even this can be questioned. Future research should see whether the same segments are visible in other types of spread foods or processed foods. Also, some people may not like peanut butter and may have been influenced. Their likes or dislikes of the food have not been asked.

The sample mainly consisted of people actively engaged in sports, on a diet, and a relatively younger sample. This is a selection bias as due to lack of resources, a convenience bias had to be taken. This strong influence of people on a diet and involved with sports can influence the results significantly. Thus, future research should try to get a more representative sample of the general population in order to make the results applicable to all consumers.

(26)

could be due to the sample which involved people more interested in healthy foods. Also, the fact that brand was missing, and the price differences may have been seen as small, could have influenced the relative price importance as well.

An interesting future research would be to see if people’s perception of taste differs when they hear of ‘more protein’ or ‘less fat’. With the importance of taste in mind, this could definitely bring useful information to the table for the food industry.

7 Conclusion

This study aimed to research which of the differences in nutrition are most valued by

consumers, based on their willingness-to-pay. All hypotheses were not rejected, although

their impacts on WTP differed. Fats are not always considered bad, and only in the international community it is something considered very bad. Within the Netherlands, people accept fats in peanut butter. This is interesting as Thunström and Rausser (2008) stated that those nutrients a product is rich of, will have a negative effect on WTP. In the peanut butter case, there are a lot of fats presents, yet most people do not seem that affected by it. Protein seems to be more important, and all segments see the benefit of additional protein. Although it could be sample biased, this means that protein cannot be overlooked. Most products or advertisements do not mention protein, but this could be a reason why they should. Even the people less involved in sports seemed to appreciate protein. It also seems that those people who have more of an interest in health, are more influenced by the ‘health standards’ per nutrients. This is interesting as it does indicate that some people, like the female carb

counters, are less informed about nutrition. Carbs in general are not the main priority of the

groups. Sugar however, seems to be a popular reason to not buy a product. This is also linked to protein, as the two segments who care most about protein, also are influenced by sugar.

Next, price seems to only be of focus in one segment, which are the price conscious customers who claim to be after taste more than healthiness. It means there is still a market for people who prefer cheap, unhealthy items, and this will probably not change soon. However, the question is how big this segment in reality is, and what their spending power is. These are the younger people, who might change segments over time. It might be good to find out whether these people are more taste, or price focused, in order to capture this segment.

(27)
(28)

8 References

Bech-Larsen & Scholderer (2007). Functional Foods in Europe: consumer research, market experiences and regulatory aspects. Trends in Food Science & Technology. 18: 231-234

Berning, J.P., Chouinard, H.H., Manning, K.C., McCluskey, J.J. & Sprott, D.E. (2009). Identifying Consumer Preferences for Nutritional Information on Grocery Store Shelf Labels.

Food Marketing Policy Center. Connecticut.

Bonanno, A. (2013). Functional foods as differentiated products: the Italian yoghurt market.

European Review of Agricultural Economics 40:45-71.

Burton, S. & Biswas, A. (1993). A Preliminary Assessment of the Effects of Changes in Nutrition Labels Required by the Nutrition Labeling and Education Act of 1990. Journal of

Consumer Affairs. 27(1):127-144

CCFN (2009) Tracking Nutrition Trends: a 20-year History. Canadian Council of Food and

Nutrition. Mississauga.

Contento, I. R., Michela, J. L. & Goldberg, C. J. (1988). Food choice among adolescents: population segmentation by motivations. Journal of Nutrition Education, 20, 289–298.

FDA (2009). Experimental Study of Carbohydrate Claims on Food Packages Analysis Report.

Food and Drug Administration. Silver Spring.

Gensler, S. Hinz, O., Skiera, B & Theysohn, S. (2012). Willingness-to-pay estimation with choice-based conjoint analysis: Addressing extreme response behavior with individually adapted designs.

Grunert, K.G. & Willis, J.M. (2007). A review of European Research on consumer response to nutrition information on food labels. Public Health. 15: 385-399

Hearthy A.P., McCarthy, S.N. , Kearney, J.M. & Gibney, M.J. (2007). Relationship between attitudes towards healthy eating and dietary behavior, lifestyle and demographic factors in a representative sample of Irish adults. Appetite 48:1-11

Heasman, M. & Meletain, J. (2001). The Functional Foods Revolution: Healthy People, Healthy Profits? Routledge.

Hensen, S. & Hooker, N.H. (2001). Private sector management of food safety: public regulation and the role of private controls. International Food and Agribusiness Management

Review 4:7-17.

Hu, T., Mills, K.T., Yao, L., Demanelis, K, Eloustaz, M., Yancy, W.S., Kelly, T.N., He, J. & Bazzano, L.A. (2012). Effect of Low-Carbohydrate Diets versus low-fat diets on metabolic risk factors: A meta-analysis of Randomized Controlled Clinical Trials. American Journal of

(29)

Hu, W., Woods, T. & Bastin, S. (2009). Consumer Acceptance and Willingness to Pay for Blueberry Products with Nonventional Attributes. Journal of Agricultural and Applied

Economics 41:47-60

Johnson, R.M. & Orme, B.K. (1996). How many questions should you ask in choice-based conjoint studies? Sawtooth Software, Inc. 360/681-2300

Koivisto, U.K., Sjöden P.-O. (1996).Reasons for rejection of food items in Swedish families with children aged 2-17. Appetite, 26:89-103

Krystallis, A., Maglaras, G & Mamalis, S. (2008). Motivations and cognitive structures of consumers in their purchasing of functional foods. Food Quality and Preference 19: 525-538 Loureiro, M.L., Gracia, A. & Nayga, R.M. (2006) Do consumers value nutritional labels?

European Review of Agricultural Economics 33(2): 249-268

Menrad, K. (2003). Market and Marketing of functional foods in Europe. Journal of Food

Engineering 56:181-199

Millward, D.J., Layman, D.K., Tomé, D. & Schaafsma G. (2008). Protein quality assessment: impact of expanding understanding of protein and amino acid needs for optimal health. The

American Journal of Clinical Nutrition. 87(5):15765-15815

Moorthy, S., Ratchford, B., Talukdar, D., 1997. Consumer information search revisited: Theory and empirical analysis. Journal of Consumer Research 23:263–277.

Nayga, R.M. (2000). Nutritional knowledge, gender and food label use. Journal of Consumer

Affairs 34: 97-112

NDP Group (214). Eggs, chicken, and yoghurt top Growing Protein sources.

https://www.npd.com/wps/portal/npd/us/news/press-releases/us-consumers-want-more-protein-in-their-diets-and-look-to-a-range-of-sources-for-it/

PHEIAC (2013). Evaluation of the implementation of the Strategy for Europe on Nutrition, Overweight and Obesity related health issues. European Commission Public Health Reports. Roininen, K., Lähteenmäki, L & Tuorila, H. (1999) Quantification of Consumer Attitudes to Health and Hedonic Characteristics of Foods. Appiteite. 33:71-88

Russo, J. E., Staelin, R, Nolas, C.A., Russell, G.J. & Metcalf, B. L. (1986) Nutrition Information in the Supermarket. Journal of Consumer Research. 13:48-70

Siro, I., Kapolna, E., Kapolna, B. & Lugasi, A. (2008). Functional food. Product development, marketing and consumer acceptance – A review. Appetite. 51:456-467

Tarabella, A. & Burchi, B. (2012). The role of nutrition and health claims in consumers’ perception. Creating educational paths to resolve information asymmetries caused by promotion and marketing activies regarding foodstuffs. Procedia – Social and Behavioral

(30)

Thunström, L. & Rausser, G. (2008). The Marginal Willingness to Pay for Health-Related Food Characteristics. Food Economics. 5, 194-206.

Verbeke, W (2006). Functional Foods: Consumer willingness to compromise on taste for health? Food Quality and Preference. 17:126-131

Weaver, D. & Finke, M. (2003). The relationship between use of sugar content information on nutrition labels and the consumption of added sugars. Food Policy 28:213-219

Wedel, M. & Leeflang, P.S.H. (1997). A model for the effects of psychological pricing in Gabor-Granger price studies. Journal of Economic Psychology. 19, 237-260.

9 Appendix

a. Reliability analysis health: Cronbach’s alpa, health1 representing the reversed

formulated question. Reliability Statistics Cronbach's Alpha N of Items ,536 3 Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted health2 7,0231 7,642 ,502 ,199 health3 7,0809 7,741 ,523 ,179 health1 7,1098 9,186 ,115 ,848

b. Reliability analysis taste: Cronbach’s alpa, pleasure_taste2 representing the

reversed formulated question on whether or people finish their meals even when they do not like the taste

Reliability Statistics Cronbach's

Alpha N of Items

(31)

Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Cronbach's Alpha if Item Deleted pleasure_taste 3,2081 3,310 ,138 . pleasure_taste2 3,8382 1,893 ,138 .

(32)
(33)

e. Histogram health 3

f. Histogram: I always finish my meal

(34)
(35)
(36)
(37)

Referenties

GERELATEERDE DOCUMENTEN

Choudhury (2002) found that education is highly correlated with long-term financial well-being and that homeownership patterns based on education are similar to homeownership

This study further investigates the field of customization by testing the effect of a personalized direct mail, that fits customer preferences, on the drivers of customer equity

Various energy transfer investigations where the energy is transferred from Bi 3+ to different luminescent ions (Bi 3+ is utilized as a sensitizer for the emission)

Including the effect of answering more choice sets to the image condition will lead to depletion of resources sooner and therefore lead to a higher price

The aim of this research is to investigate the role of awe, a discrete positive emotion, on individuals’ levels of message reception and willingness to pay for consumer goods that

Need for Cognitive Closure (Webster & Kruglanski, 1994; Roets & Van Hiel, 2011) 15-items scale; 6-item Likert ranging from strongly disagree to strongly

Looking at the estimated utilities for the type of appeal that is on the package, a higher coefficient for a biospheric appeal indicates that consumers prefer to pay more for a