The effect of adding variation suggestions to in-store meal kits on the purchase intention

Master Thesis

The effect of adding variation suggestions to in-store

meal kits on the purchase intention

R.J. (Rob) Loman

The effect of adding variation suggestions

to in-store meal kits on the purchase

intention

Master Thesis

University of Groningen

Faculty of Economics and Business

MSc Marketing Management

June 14, 2020

Rob Loman

Verzetsstrijderslaan 176

9727CJ Groningen (NL)

robloman9999@gmail.com

(+31)625332965

Student Number: S3812103

1

_{supervisor: M.T. van der Heide}

2

_{supervisor: A. Schumacher}

2

Abstract

The increasing need for convenience and awareness of healthy eating, in combination with lack of time and inspiration may have inspired supermarkets to sell in-store meal kits. In-store meal kits are fresh, pre-portioned ingredients with an accompanying recipe for consumers to cook a healthy homemade meal. However, consumers do not always want the same food and are looking for variety in their food choices.

In this research, there is investigated if people have a higher purchase intention for a meal kit when the meal kit has different variation suggestions. This is the first study in examining the influential role of variety within products whereby variations are determined by the company and it is not

personalized by the customer. It is proposed that by offering more variety in a meal kit, consumers have the freedom to choose and do not experience satiation due to eating the same meal kit every time. Furthermore, there is investigated if this effect on the purchase intention is more pronounced for people with high cooking skills because people display more variety-seeking behavior if they are more knowledgeable and find more personal relevance with the object. In addition, there is

investigated whether this effect is less pronounced for convenience-oriented people because with additional variation suggestions, planning and shopping will take more time, effort and risk. An experimental study with two different conditions was conducted, a meal kit without variation suggestions and a meal kit with variation suggestions. The study involved 230 participants, and it was found that people do not have a significantly higher purchase intention for meal kits with variation suggestions. Furthermore, there was no interaction effect between variation suggestions and cooking skills or between variation suggestions and convenience orientation. However, further analysis proved fruitful as some significant relationships were established. The willingness to pay for the meal kit with variation suggestions was positive and significant. The participants were willing to spend €0.18 more on the meal kit with variation suggestions. Therefore, it is interesting for managers to add some variation suggestions on a meal kit. Furthermore, if people perceive the meal kits as convenient or they perceive the meal kit as giving a high degree of freedom, is has a significant and positive effect on the purchase intention and willingness to pay. Marketers should research how to ensure that the product gives a higher degree of freedom, whilst retaining convenience Therefore, further research using field studies and greater sample sizes is recommend to investigate the

variation suggestions within a meal kit. This may make it possible to find an optimal balance between variety-seeking behavior and the convenience of buying the same meal kit each time.

Key words: meal kits, variety, variety-seeking behavior, purchase intention, cooking skills,

3

Table of contents

2.2 Variation suggestions in meal kits ... 10

2.3 Cooking skills ... 13 2.4 Convenience orientation ... 15 2.5 Conceptual framework ... 18 3. Methodology ... 19 3.1 Participants ... 20 3.2 Procedure ... 20 3.3 Measures ... 21 4. Results ... 23 4.1 Data analysis ... 23 4.2 Descriptive analysis ... 23

4.3 Main and moderating effects ... 24

4.4 Follow up analysis ... 26

5. Discussion ... 30

5.1 Main findings ... 30

5.2 Theoretical contributions and managerial implications ... 32

5.3 Limitations and future research ... 34

5.4 Conclusion ... 36

6. References... 37

7. Appendix ... 43

Appendix A: Questionnaire ... 43

Appendix B: Original Convenience Orientation scale ... 51

Appendix C: Reliability statistics ... 51

Appendix D: Distribution cooking skills and convenience orientation scores ... 52

Appendix E: Demographic characteristics ... 53

Appendix F: PROCESS Macro output ... 54

Appendix G: PROCESS Macro output with control variables ... 54

Appendix H: Multiple regression predictors ... 54

Appendix I: PROCESS Macro willingness to pay ... 55

Appendix J: PROCESS Macro willingness to pay – manipulation check ... 55

Appendix K: PROCESS Macro willingness to pay – manipulation – control variables ... 56

4

1. Introduction

Worldwide, home cooking forms part of public health strategies to combat obesity and poor

nutritional intake (Mills, 2017). There is growing evidence of an association between increasing levels of obesity and poor-quality diets (Lobato et al, 2009). Furthermore, consumers have become

accustomed to convenience and speed, implicit in the fast-food industry. This has changed food habits, family life, and consumption habits. People spend less time cooking and rely more on prepared convenience foods, fast foods, and other foods made away from home (Wolfson & Bleich, 2015). The decrease in cooking and the increased consumption of convenience foods is leading to increasing levels of obesity in the world. (Mills, 2017; Flegal et al, 2010). Because of this, health authorities have invested in public health initiatives to minimize fast food intake and to encourage food preparation and cooking at home (Smith et al., 2013).

Many studies show that home cooking is healthier. The Johns Hopkins Bloomberg School of Public Health (2016) concluded in a recent study that people who often cook their meals at home eat more healthily and consume fewer calories than those who cook less at home. They consume less sugar, less fat and fewer carbohydrates (Wolfson & Bleich, 2015). If people prepare their own meals, they have a high level of control over the ingredients and techniques. A study by Dohle et al. (2016), shows that food preparation improves ingredient awareness. They know better of ingredients are healthy or unhealthy. Such findings reinforce recent proposals for new initiatives that allow people to cook and eat at home more often.

Warde (1999) attributes the increasing consumption of convenience foods to time pressure more than to food preference. The feeling that people do not have enough time to do everything they need or want to do has been identified as a contributor to a decline in home cooking (Storfer-Isser, 2017). People have less time than ever before due to socio-economic revolutions such as the increased female participation in the work force and longer office hours (Costa et al., 2007). People with less time prefer convenience when it comes to food consumption. Home cooking is more than just the time spent in the kitchen. It includes time and effort (physical and mental) to buy, plan, prepare, and eat food. Industry and service sectors have reacted to the convenience trend by

increasing product development and considering expanding their offers of convenience products and services (Costa et al., 2007). The increasing need for convenience, consumers’ increasing awareness of the importance of healthy eating, and the lack of time and inspiration may have inspired

supermarkets to sell in-store meal kits.

5

are convenient because they relieve pressure on meal planning. Meal kits also reduce shopping time and provide socially acceptable solutions to the demands of “feeding the family and cooking from scratch”(Hertz & Halkier, 2017). They differ from typical convenience foods, in that they are often more nutritious, inspire people with novel foods and recipes and are more environmentally friendly (less food waste and packaging). Therefore, they have the added benefit of avoiding the negative associations linked with convenience foods (Hertz & Halkier, 2017).

Historically, suppliers of convenience foods have promised the consumer speed, domestic efficiency, and freedom (Smith, 2001), and this is no less the case with meal kits. All convenience foods rely on the promise of accelerating consumer food provisioning in some way, from reducing the need to shop, to simplifying the cooking process. Meal kits are presented as a solution by removing the work of planning and shopping for all the products one by one so that consumers can concentrate just on cooking. Marketers promote meal kits as a bridge between convenience and from-scratch cooking and as a healthy option due to their fresh ingredients (Mintel, 2016). These kits enable people to cook from scratch more easily while escaping the normative accusation of serving inappropriate meals (Hertz & Halkier, 2017). The convenience of meal kits provides consumers with a way to save time by outsourcing planning-related obligations. Subsequently, the reduced grocery store shopping and meal planning time can then be invested in other tasks which the consumer may believe to be more valuable.

In-store meal kits are popular with consumers and demand for them could grow. Many Dutch supermarkets offer in-store meal kits with ingredients for meals such as Thai green curry, Lasagna and Burritos. By 2019, more than a third of Dutch households had bought a meal kit at some point, compared to only 10% in 2018 (Motivaction, 2019). Convenience, price, taste, and the idea of being healthy are the most important factors contributing to meal kits’ success. The most significant deterrents to buying meal kits are the price and consumers’ appetite for cooking (Motivaction, 2019). Meal kits make it easier for consumers to eat enough fresh vegetables and make healthy choices. According to the GroentenFruit Huis (2020), Dutch people bought 2% more vegetables and fruit in 2019 than in 2018. This was partly due to the growing popularity of meal kits.

6

kits. Consumers do not always want the same food and seek variety. Consumers’ variety-seeking is an important characteristic that affects the behavior of consumers in food choices (van Trijp, 1992). Even consumers are satisfied with current products, they have the desire to search for alternative products. Boredom and satiation reflect the fact that consumers become satiated with some of the attributes they are consuming, resulting in decreasing marginal value for incremental units of those products (van Trijp, 1992). A viable marketing strategy might be to add more variation suggestions to offer a portfolio of food products within which consumers can satisfy their need for variety.

Some studies have researched the variety-seeking tendency of consumers with regards to food. However, these papers have not considered whether adding variation suggestions to food packages has a positive effect on the purchase intention. The intensity of variety-seeking will decide the potential market share of the meal kits and the marketing strategies of suppliers and distributors (Feinberg et al., 1992). There is also little evidence and few existing literature concerning which type of people are more influenced by variety. Van Trijp (1995) indicates that variety-seeking behavior is more likely to arise among more knowledgeable consumers. Gursoy and Gavcar (2003)

demonstrated that as consumers find more personal relevance with the object, their decision-making behavior varies. Therefore, it can be assumed that people with high cooking skills are looking for more variation suggestions in a meal kit.

Little literature exists about which type of people want less variety. However, previous research by Candel (2001) has shown that convenience orientation was negatively correlated with the

importance of variation within and between meals. Convenience-oriented people want to save time and effort and do not try new recipes. The convenience orientation of the consumers toward meal planning and shopping is relevant to understand consumer behavior about food (Candel, 2001). With the results of this study, marketers and practitioners will have additional information about

convenience-orientated people and convenience products. Anderson (1972) claims that convenience orientation is a basis for market segmentation and more efficient marketing efforts. Therefore, it is an interesting construct for marketers to investigate if they do not want variation

suggestions(Frewer,& van Trijp, 2007) or if it makes little difference for them.

7

customer, and therefore attracting customers (Chang et al., 2006). Customers are able to customize their products to their specific wishes, which suggests that this is an effective strategy to satisfy customer’s needs. The difference between variation suggestions in meal kits and the customized products, is the ability for customers to incorporate their own ideas in the customized products (Zhen et al., 2017). The offering of variation suggestions in meal kits are determined by the company. It is interesting to conduct this research for a product that is highly convenient because it is

important to ensure that the product still offers enough convenience, whilst ensuring that customers do not experience satiation. This enables the identification of gaps and opportunities to improve knowledge of how consumers try and should try to maximize the enjoyment of the same meal kit over time.

Therefore, this paper builds on past research, investigating whether consumers want variation suggestion on meal kits. To fill this knowledge gap, the following research questions has been developed: “What is the effect of variation suggestions in meal kits on the purchase intention?” Identification of the degree to which consumers prefer variation suggestions in a meal kit may be useful in understanding consumer behavior toward the meal kits. Based on the research gap, two moderators are expected to influence the effect of meal kit variation suggestions on the purchase intention: cooking skills and convenience orientation. By investigating the impact of these

moderators, managers can determine the optimal variation in meal kits. Therefore the second research question is: “How is this effect influenced by cooking skills and convenience orientation?” To answer these questions, this research will follow an experimental research strategy. The outline of the thesis is as follows. Chapter 2 starts with a literature review of the already existing literature on this topic, followed by the derived hypotheses and conceptual model. The methodology of this research is then elaborated upon (Chapter 3), after which the hypotheses are tested in the results section (Chapter 4). This is followed by the discussion and conclusion that provide a summary of the most important findings and put them into a broader context (Chapter 5).

8

2. Literature review

This study focuses on the variation suggestions in meal kits. Therefore, relevant articles about home cooking, variety-seeking, cooking skills and convenience orientation are analyzed and discussed. This is followed by the research hypothesis regarding meal kit variation suggestions and a conceptual model.

2.1 Home cooking

As described in the introduction, most nutritional approaches assume a positive influence of home cooking on diet, health and social effects (Mills, 2017). Food prepared at home is associated with lower calorie intake, richer nutritional quality and weight loss (Wolfson & Bleich, 2015). Cooking for oneself can provide a high level of control over the ingredients. However, when the focus is on getting food on the table as quickly as possible, nutrition becomes less important (Brunner, 2010). Therefore, in recent decades home cooking has received greater attention and has been marketed as a strategy for improving nutritional intake and preventing obesity (Ducrot et al, 2017).

Main obstacles to home cooking include affordability, lack of time, and lack of enjoyment (Mills, 2017). Besides price, sensory appeal and health-related concerns, convenience is an important determinant of food choice (McIntosh, 1996). Convenience has gained importance because of increased demand for an easy, tasty, and ready-made solution (Närvanen, 2013). Previous studies found that convenience is, besides saving time, also dependent on minimizing physical and mental effort with the preparation and planning of the meal(Man & Fullerton, 1990). In this paper, I use Brunner et al.’s definition of convenience. This defines convenience broadly, describing convenient food products as “those that help consumers minimize time as well as the physical and mental effort required for food planning, shopping, preparation, consumption, and clean up”(Brunner et al, 2010). Meal kits ensure that people spent less time in the planning and shopping stage, and are therefore considered a type of convenience food.

9

To save time and energy in the cooking process, consumers follow several coping strategies, based on their degree of motivation on meal preparation and cooking skills (Alm & Olsen, 2017). The most common strategy for food coping was to speed up cooking process by cooking convenience foods because that can be prepared quickly (Mandemakers & Roetes, 2015). They spend less time buying and preparing food and depend more on outsourcing and purchasing prepared meals. Eating at fast-food restaurant is less often seen as option because people considered that as expensive and unhealthy (Alm & Olsen, 2017).

Furthermore, financial restrictions are important barriers that healthy cooking and frequent cooking is not practicable for everyone (Costa et al, 2007). Several studies have examined the role of financial resources and the motivation to save money in home cooking behavior, which indicates the

importance of affordability (Wolfson, 2016). Time scarcity is more problematic for families with a low-income who do not have the optimal circumstances of resources such as flexible work schedules or money for adequate child care or prepared food, that would foster adaptation to high time demands (Wolfson & Bleich, 2015).

Kalenkoski and Hamrick (2013) suggest that time-poor consumers do not necessarily turning to unhealthy convenience products. Rather, they are likely to have alternative prepared foods, that is, other ‘convenient’ food options such as meal kits (Halkier, 2017; Hertz & Halkier, 2017). The meal kits could be a solution to balance competing time demands and reduce barriers to healthier eating habits at a low price (Ducrot et al, 2017). Buying a meal kit minimizes the physical and mental effort in meal planning and shopping.

The average Dutch person shops for groceries for three hours per week and cooks for five hours per week (Peil, 2016). Besides the time taken by shopping, determining the meal also takes a lot of time. A meal kit shortens the planning and shopping time because the meal kits have a recipe and pre-measured and prepicked ingredients. A meal kit contains almost all the ingredients you need to take each meal as well as a recipe and pre-measured items; therefore it is not necessary to decide which ingredients to use.

10

et al, 2017). Because the meals are planned with the meal kits, meal kitscould solve the time scarcity problem reported for meal preparation and thus could promote home cooking (Ducrot et al, 2017). Therefore, a meal kit has many benefits such as inspiring people what to cook, keep the price low, and reduces physical and mental effort when planning and shopping all ingredients.

However, a disadvantage of meal kits is that they contain the recipe and almost all ingredients, such that the meal is always the same. The lack of variation can lead to consumers not buying the product after a while. Extra variation suggestions can be considered as an added value of the product, which in turn can lead to an increase in the purchase intention.

The following sections will discuss research on variety, cooking skills and convenience orientation in more detail.

2.2 Variation suggestions in meal kits

While a consumer is shopping in a supermarket, the consumer’s purchase intentions can be

influenced by different marketing strategies. Multiple theories, such as brand loyalty, self-perception theory and cognitive dissonance suggest that consumers will buy the same products as the previous time. However, a broad selection of studies suggest that consumers seek variety in their decisions (Ellis, 2018). Moreover, consumer’ intrinsic desire for variety is recognized as an important

characteristic influencing their food choices (van Trijp & Steenkamp, 1992). Variety-seeking in foods is formally described as “the motivational factor that aims at providing variation in stimulation through varied food consumption”(van Trijp, 1995). A meal kit is highly convenient, but this quickly limits the variety in consumer meals and negatively effects consumers repurchase intention (Berne et al, 2001). If a meal kit shows multiple recipes and additional ingredients, this may result that people may be more likely to buy it again. Therefore, variation suggestions on a meal kit might be an effective strategy to increase purchase intentions. There are three main mechanisms through which variation suggestions on a meal kit have a positive influence on purchase intentions. The first is that people get satiated if they prepare and consume the same meal kit several times. The second mechanism is that consumers want to experience new food. The third mechanism is that when people have different options, they perceive higher satisfaction because they have autonomy and freedom of choice. These three mechanisms will be explained further below.

Firstly, feeling of satiation occurs as consumers consume products and experiences to the extent that they no longer enjoy them (Coombs & Avrunin, 1977). Satiation, particularly in the food domain, may be the result of physiological processes (Rolls et al., 1984), repetition the same food leads to

11

particular sensory signal. This implies that the more frequently a meal kit is purchased, the more bored the consumer becomes with that choice, which stimulating variety-seeking behavior.

Variety seeking in choice situations reduces boredom and satiation and increases stimulation to the preferred level (Raju, 1980, Steenkamp & Baumgartner, 1992). These assertions are based on

Berlyne’s Arousal theory and the Optimal Stimulations theory. Berlyne’s Arousal theory suggests that an individual is aroused by external stimulus properties, which can lead to variety seeking behavior (Berhyne, 1960). Consistent with this, the Optimal Stimulation Level theory assumes that individuals strive to achievea certain level of stimulation which one feels to be optimal. The optimal stimulation level is the amount of stimulation a person prefer in life. Thus, when stimulation is below the optimal level an individual is bored and desires to increase his or her stimulation. This stimulates variety-seeking behavior, so that the stimulation will be optimized to the preferred level (Baltas, 2017). Satiation can lead to searching for variety not only across products but also within a specific attribute, as buy a different flavour (Johnson et al., 1995). As a result, people do not want a completely different meal kit, but minor adjustments can lead to people becoming less satisfied. Transition to a different or new product increases stimulation during purchases (Menon & Kahn, 1995). This is in line with the results of the study of Ha & Jang (2013) that as customers in a restaurant become bored or satiated with that restaurant, they are more likely to seek a new restaurant. Therefore, it is expected that the stimulation from the same meal kit decreases with repeated use while competing meal kits become more attractive simply by being not the same meal kit. If a company offers more variation in a meal kit, consumers will see this as a different meal, preventing satiation. Most studies have discussed variety-seeking at the product-offering, with consumers brand-switching, but variation suggestion in meal kits can provide deeper insights into the pursuit of variety. This is not only important for understanding decision-making but also for providing sufficient variation suggestions within a meal kit.

This relates to consumers switching from their favored options because they are afraid of getting satiated by repetition (Read & Loewenstein, 1995). Consumers are inclined not only to seek variety because of actual and physiological needs but also to seek variety as a way of controlling satiation after initial consumption (Sevilla, 2019). Therefore, variety-seeking behavior can either be used as a reactive response to satiation (Menon & Kahn, 1995) or proactively used to avoid satiation with products (Read & Loewenstein, 1995).

12

(Meixner & Knoll, 2012). People look for something new if their hedonic motive to experience a new product is stronger than their motivation for benefits of the old product (Hirschman & Holbrook, 1982). Furthermore, extensive research has documented that consumers looking for something new require uniqueness. Uniqueness is one of the main requirements for people buying personalized items (Simonson & Nowlis , 2000). The requirement for uniqueness demonstrates the consumer’s desire to be distinctive and special (van Trijp & Steenkamp, 1992; Snyder & Fromkin, 2005). Trying new recipes relates to ambition to attain personal goals and interest in learning new skills (Costa et al, 2017). This also reflects the desire for new and novel stimuli. Uniqueness and novelty give people a chance to explore new foods with variations.

People buy a meal kit because they want to prepare and eat an extensive meal, therefore a people perceive meal kits as a hedonic product. Hedonic products enable consumers to feel pleasure and enjoyment from buying and using the product. They are linked to sensory products and are more linked to experimental use (Hirschman & Holbrook, 1982). Consumers search for more variety in hedonic product categories (Ratner et al., 1999; van Trijp et al., 1996), because the desire by hedonic products are triggered by sensory attributes. Consumers perceive sensory attributes in hedonic products to be more repetitive and want therefore to try something new (Baltas, 2017). Baltas (2017) suggests that if the product is perceived as hedonic, an appropriate strategy would be to expand the product line with products that vary in sensory attributes. Therefore it is interesting that a meal kit has different variation suggestions.

13

therefore less overall enjoyment over time. The sweet spot for a meal kit is unknown, but in this research is chosen for two additional variety suggestions.

It is suggested that by offering more variation suggestions in the meal kit, consumers will have a higher purchase intention because they have the freedom to choose. Furthermore, people become satiated if they make the same meal kit repeatedly and a meal kit with variation suggestions gives them new recipes that can fulfil people’s curiosity for new meals. Therefore, the following hypothesis is derived:

H1: Variation assumption in meal kits have a positive impact on purchase intention.

2.3 Cooking skills

Earlier research has addressed variety-seeking as means of reducing physical satiation (McAlister & Pessemier, 1982). Now, researchers indicate that the relationship between variety-seeking and reducing satiation is more nuanced, so behaviors can be driven and affected by exogenous factors (Sevilla et al., 2018). For decisions that are unimportant to the consumer, habit-based, repetitive choices are an efficient choice heuristic, and consequently, variety-seeking habits are less likely to occur. This is in line with van Trijp et al. (1992), who suggest that some degree of involvement with the product must be present for variety-seeking behavior to occur. Therefore, it is interesting to investigate if the effect of variation suggestions in meal kits is more pronounced for the people with high cooking skills. There are four drivers by which I posit that this effect is stronger for people with high cooking skills. Firstly, if people are more involved with the object, they switch more between products. Secondly, people with higher cooking skills perceive less risk when preparing a new meal. The third reason is that people with higher cooking skills eat more healthily, so variations with more vegetables or healthier options are popular among such people. The fourth reason is that people with higher cooking skills are more motivated to cook and are more motivated to make different meals.

Cooking skills are person-centered and have traditionally been described as a set of mechanical or physical skills used when preparing meals (Ternier, 2010). Cooking skills also include perception, planning, and fundamental skills of food nutrition (Ternier, 2010). Cooking skills include knowledge of what happens when food items are combined, the ability to cook, and knowledge of food (Ternier, 2010).

14

case with food choices (Brunner et al., 2010). Moreover, when customers are highly involved in buying a particular product, they show great interest in their decision-making process and search for a lot of information about the product (Lamb et al., 2007), this greatly affecting their behaviors (Homburg & Giering, 2001). Knowledge and experience of a product influence the amount of information processed by the consumer, thus affecting the consumer’s purchasing behavior (Bettman & Park, 1980). This is in line van Trijp’s research (1992) that demonstrated that variety-seeking is more likely to occur when consumers are more knowledgeable about the product. People also want to appear to have high cooking skills. If people make a variation suggestion from a meal kit, it implies that they are familiar with it. People may see variety-seeking as a sign of domain expertise and may use the variety strategically to show others expertise through their choices (Sela, 2019). More diverse recipes indicate familiarity with different options. Furthermore, food experts have been found to derive the most gratification from experimentation during the preparation of new meals (Buckley et al., 2007). These findings suggest that consumers who are more

knowledgeable about the product category are more likely to try new experiences. This is also the case with innovators, they have more product-related knowledge and are the first to try a new product (Gatignon & Robertson, 1991).

Secondly, people with high cooking skills are more confident and perceive less risk when buying new products compared to people with low cooking skills (Ellis, 2018). People with high cooking skills want more variety because they know how to use a wider variety of produce. Perceived risk impacts people’s confidence in decision-making or purchase intention. Situations of risk can be those where the probabilities of outcomes are not recognized and the outcome is known or unknown (Smith et al, 2006). Marketers claim that increased involvement in food selection and preparation is a viable strategy for encouraging consumption of a variety of foods (Barlow & Dietz, 2002). An example is with wine: people with more knowledge about wine are seek more variety. These consumers believe they know all that matters about this product category (Ellis, 2018).

The third reason is, people with higher cooking skills consume more fruits, vegetables, and whole grains (Sobal & Hanson, 2014). People with higher cooking skills have better dietary quality and make healthier food choices (McGowan et al., 2015). They buy a broader variety of items ready-prepared, and buy more fresh products as well (Levy & Auld, 2004). If people think that the variation

suggestions have more vegetables or are healthier options, these variation suggestions will be more popular among those with high cooking skills.

15

meals. People who are less involved with preparing food favor saving time and energy in food preparation, meal planning and purchasing (Botonaki et al., 2009). People who are unable to plan or create a meal to satisfy other people, may find it simpler to purchase a well-known product that will save time and energy, and be satisfying to everyone. Such people have a more routinized, recipe-following approach to cooking (Cohen, 2017). They are unsure about specific cooking techniques and lack confidence in cooking a wide variety of products (Caraher et al,. 1999), and were expected to rely on one simple recipe (Alm & Olsen, 2017). Moreover, people who described themselves as having high cooking skills were more able to handle time and stress, and added more variety to their meals prepared on time-stressed days. Even after a busy day, people with high cooking skills wanted variation in their meals (Alm & Olsen, 2017).

Therefore, it is expected that people with higher cooking skills have more motivation to try different recipes and feel less risk concerning new meals, because they have more knowledge and are more involved. Therefore, I assume that cooking skills play a role in the way variety in meal kits influence purchase intention. Hence:

H2: The positive effect of variation suggestions in meal kits on purchase intention is more pronounced for people with high cooking skills

.

2.4 Convenience orientation

Berry et al (2002) conclude that all goods that save consumers time and effort are more appreciated by convenience-oriented people. Convenience-oriented people have been operationalized by the use of convenience foods (ready to eat meals), paid services (domestic services, child care) and

timesaving durables (e.g. dishwasher, and microwave ovens;) (Berry et al, 2002). This may relate to variation suggestions in meal kits because this also takes more time and (physical and mental) effort. Therefore, it is assumed that convenience-oriented consumers are not variety seekers because they want to make simple meals. This is based on the three mechanisms: convenience-oriented people want not elaborate meals, they want to spend little time and energy in cooking, and they want no stressful and challenging situations thus choose for security.

16

consumer is inclined to save time and energy as regards meal preparation”(Candel 2001). He has only researched convenience orientation related to the meal preparation stage. However, there are different stages in the process of consuming food: product planning and shopping, storage and preparation of products, consumption of itself, and cleaning up and disposal (Gofton, 1995). I use in this study the idea of convenience orientation as saving time, physical energy, and mental effort in connection with planning and shopping activities for consumers.

The first reason why convenience-orientated consumers want fewer variation suggestions in meal kits is that convenience orientation is negatively correlated with cooking enjoyment and meal preparation(Candel, 2001). Convenience-oriented consumers are passive consumers and their priority is to live a carefree, pleasant, and comfortable life. They want a simple meal rather than an elaborate one. Convenience-oriented people try to accomplish tasks with the least expenditure of human energy in the shortest time (Berry et al. 2002; Morganosky, 1986). Therefore, cooking and planning a meal is easier if you eat something you have eaten before. If a meal kit has more variations, the purchaser has more to read and may need to learn different recipes. Therefore, convenience-oriented people do not need additional variations, which make the process of meal planning less convenient for them.

Secondly, convenience orientation is positively related to role overload. Candel defined role overload as: “the degree to which the amount of energy and time demanded of a person as a result of the roles this person performs is perceived to exceed his or her available time and energy

resources”(Candel, 2001). The more an individual experiences role overload, the more of a burden obligatory activities become, and the higher a person’s tendency to seek convenience in activities such as meal planning and shopping (Candel, 2001; Brunner et al., 2010). Consumers with a strong convenience orientation prefer activities and behaviors with intrinsic time- and effort-saving characteristics (Collier & Kimes, 2012). Psychological studies have emphasized that perceived time pressure is a stronger determinant of convenience orientation than the real number of working hours (Darian & Cohen, 1995). They want to spend as little time as possible on challenging issues and circumstances and want to avoid discomfort and pain (Gottschalk & Gunnesdal, 2018). Convenience-oriented people believe that the less effort needed the better and think that spending a long time on an issue would be a waste of time (Gottschalk, 2020). If eating a take-away meal is considered convenient, then an increase in convenience orientation will lead to an increase in preference for this food option (Candel, 2001). Convenience-oriented people are willing to pay more for the

17

The last reason is that convenience-oriented people want security. Brunso et al. (2004) found that, in convenience-oriented people, planning, food shopping and cooking is positively related to ‘security’, the opposite to stimulation. People who are primarily motivated by stimulation seek an exciting life, full of novelty and adventure. Convenience-oriented consumers prefer security and will choose the safest option. These consumers want to avoid more problematic, stressful, and challenging

situations. This is the case for meal kits with additional variation that have more unknown recipes. Meal kits are convenient in terms of time and effort saved on meal planning and shopping.

Convenience orientation relates to the amount of time and effort that is required to carry out a task. With additional variation suggestions, planning and shopping will take more time, effort, and risk. However, convenience-oriented people want to make their life comfortable. Therefore, I

hypothesise:

18

2.5 Conceptual framework

Based on the literature review and the hypotheses, a conceptual model have been developed. The conceptual model represents the relationship between the independent variable (meal kit variation suggestions), the moderators (cooking skills and convenience orientation) and the dependent variable (purchase intention). The hypotheses are as follows:

- H1: Variation suggestions in meal kits have a positive impact on purchase intention - H2: The positive effect of variation suggestions in meal kits on purchase intention is more pronounced for people with high cooking skills.

- H3: The positive effect of variation suggestions in meal kits on purchase intention is less pronounced for people with high convenience orientation.

This is depicted in Figure 3:

19

3. Methodology

Data from a convenience sample of consumers from the Netherlands was collected between the 12th

of April 2020 and the 26th _{of April 2020. To answer the research questions and check the three}

hypotheses, an experimental study was designed. The design of the experiment involves a between-subjects design, in which the participants are randomly assigned to one of two different conditions. The meal kits that supermarkets sell currently are shown to the control group (N = 118) (see Figure 4) and a meal kit with two additional variations is shown to the experimental group (N = 115) (see Figure 5). Data analysis was conducted using IBM SPSS Statistics 25.

Figure 4: Control condition

20

The following sections explain the participants of the questionnaire (Section 3.1), the procedure (Section 3.2), and the different measures (Section 3.3).

3.1 Participants

The participants of this experiment were individuals who live in the Netherlands. The target sample was the entire Dutch population above eighteen years old. All participants conducted a Qualtrics survey that took approximately 5 minutes to complete and was conducted at their time of preference. The data was collected by sending out the survey through social media (WhatsApp, Instagram, and Facebook). A snowball sampling procedure was used by asking participants to forward the survey to their acquaintances and friends, by sharing it on their profile or with relevant groups.

Before the actual survey was provided, a pre-test was conducted with ten people ensure construct validity as much as possible. Respondents who did not complete the survey and participants from the pre-test were excluded from the results. After screening the data of these participants, 233

responses were used for the analysis. After inspecting the data for outliers and odd figures, three other participants were removed from the sample because of the very short time (< 2 minutes) they took to fill in the questionnaire. For all answers, a response was compulsory, which resulted in no missing values. Thus, the final sample used for the analysis consisted of 230 subjects. The sample is described in more detail in the results section.

3.2 Procedure

A questionnaire was used to conduct this research. This section will explain the structure and content of the questionnaire. The full questionnaire is presented in Appendix A. All respondents received a Qualtrics link to participate and were asked to forward the questionnaire to other possible

respondents. The survey began with a brief introduction of the research, explaining that the study was related to food choices. The introduction was kept as general as possible ensure construct validity as much as possible. In the first part, there was a short explanation of meal kits and the respondents received two questions about the prevalence of meal kit usage. These questions considered the buying history and the likelihood to buy of the respondents and their household members regarding in-store meal kits.

21

purchase intention, the willingness to pay and to what extent they thought that this meal kit offers convenience and freedom (see Section 3.3 for details). Afterwards, a manipulation check was performed such that attention to the variation suggestions would be checked. This was only displayed to the experimental condition, to check whether or not they noticed the variation suggestions.

In the next part, all the participants were asked to reply to a series of questions about their cooking skills and convenience orientation. After these moderator questions, participants were asked about their health consciousness and to answer a couple of questions about their grocery shopping and meal preparation habits. The last part of the questionnaire focused on demographic characteristics, dietary restrictions andhunger level. Lastly, participants were thanked for their participation and were debriefed about the purpose of the study. The participants also could leave questions or comments behind, as well as to fill in their email to have the chance of winning a Bol.com gift card.

3.3 Measures

All items used in the questionnaire are based on existing scales. Each variable is explained in the following section.

Purchase intention: The participants received two questions to measure purchase intention as

applied by Sweeney et al. (1999). The first question was “I would consider buying this meal kit” and the second question was “there is a strong likelihood that I would buy this meal kit at the

supermarket”. The above questions were assessed on a seven-point Likert scale (1 = strongly disagree; 7 = strongly agree).

Cooking skills: The measurement of cooking skills is based on the cooking skills scale applied by

Brunner et al. (2010). Cooking skills were measured based on five items, using a binary scale where 1= “yes” (high cooking skills) and 2 = “no” (low cooking skills). These items included, for example, “I can prepare a soup from scratch” and “I can bake a cake from scratch”. To prevent participants from understanding which moderator would be measured, these statements were asked after the

experiment.

Convenience orientation: After the cooking skills, the moderator “convenience orientation” was

22

Health consciousness: To control for health consciousness among the respondents, questions from

23

4. Results

After creating the questionnaire and collecting responses, the output was analyzed to draw conclusions. In this section, I explain the different steps in the analysis, after which the results are described. Finally, the hypotheses are tested, and a follow-up analysis is performed.

4.1 Data analysis

The first step was to prepare the dataset. The independent variable was recoded (1 = participants were exposed to the meal kit with variation suggestions, -1 = participants were exposed to a meal kit without variation suggestions). After that, a reliability check for the scales was obtained to assess if the scales consistently reflect the construct that they should measure. This was measured using Cronbach’s alpha. For convenience orientation, the reliability analysis measured a reliable overall convenience orientation index (alpha) of 0.891. For the second moderator, cooking skills, the reliability analysis has a low Cronbach alpha, being 0.585. This can be explained by the fact that Cronbach’s alpha is strongly recommended for Likert-type scales, but there is no clear consensus about its usefulness when using binary data, which is used in our questions on cooking skills. For the control variable, health consciousness, the reliability analysis resulted in alpha = 0.908, showing that it is reliable. None of the scales would positively affect the reliability if items were deleted. The results of the three reliability analyses can be found in Appendix C.

Next, a mean value from participants’ answers to each item of the scales was found to determine the strength of the moderators (convenience orientation and cooking skills). After that, both moderators were mean-centered. This diminishes the multicollinearity and may render the B coefficients more easily interpretable. The participants’ convenience orientation of is moderate (M = 4.44, SD = 1.30 on a seven-point Likert scale). The distribution of the cooking scales scores can be found in Appendix D (Figure 10). Furthermore, most respondents (M = 1.26, SD = .25 on a binary scale) consider

themselves as having high cooking skills. The distribution of the cooking scale is slightly right-skewed, which means that the participants in this sample have high self-perceived cooking skills on average. The distribution of the convenience orientation scores are showed in Appendix D (Figure 11).

4.2 Descriptive analysis

24

sample is knowledgeable about the existence of meal kits and is more likely to have bought a meal kit than the overall Dutch population.

Furthermore, respondents purchase intention can be considered moderately high (M = 5.01, SD = 1.57). 158 respondents (68.7%) rated their purchase intention as likely to very likely. The distribution of the scores of the purchase intention is displayed in Figure 6. However, the willingness to pay for both meal kits together is slightly lower (M = 4.66, SD = 1.84) than the price of this meal kit in the Albert Heijn online store (€4.79). Appendix E shows an exact overview of all the characteristics of the sample.

Figure 6: Histogram purchase intention

4.3 Main and moderating effects

To test the hypotheses, I started with an independent samples t-test for the purchase intention to test the difference between the control and experimental group. The participants in the

experimental condition scored higher (M = 5.07; SD = 1.57; on a seven-point Likert scale) than the control group (M = 4.96; SD = 1.58). However, these results are not statistically significant (t (228) = -.545; p = .586). Based on these analyses, there is no evidence that the variation suggestions on a meal kit influences the purchase intention.

After the independent sample t-test, I tested the other hypotheses. The dependent variable and both moderators are continuous, and the independent variable is binary, so I performed a linear

regression. I performed the regression analysis PROCESS Macro for SPSS v3.4.1 from Hayes. Hayes offers several models and test the data more specifically. Model 2 of the PROCESS Macro analysis will be used to examine whether the effect of the variety in meal kits on the purchase intention is

25

The dummy variable (with variation suggestions = 1, without variation suggestions = -1) was used as the independent variable. The dependent variable was the calculated mean of the purchase

intention. The mean-centered cooking skills and convenience orientation scores were used as moderators. The overall model is significant, R = .232; R² = .054, p = .028. However, the low R² indicates that the variation of the purchase intention can be only be attributed to the independent variable and moderators to a small extent. The PROCESS Macro analysis verifies the t-test findings that the main effect (B = .033, t(224) = .318, p = .741) is not significant. Therefore, hypothesis 1 cannot be accepted because the purchase intention is not statistically significant higher if a meal kit have variation suggestions.

The second hypothesis predicts that high cooking skills would strengthen the effect of variation suggestions on the purchase intention. However, the interaction (showed in Figure 7) between cooking skills and the variation suggestion in meal kits (B = -.591, t(224) = -1.449, p = .149) is also not significant. Therefore, it can be concluded that the effect of variation suggestions in meal kits on purchase intention is not more pronounced for people with high cooking skills. However, the

direction of the effect is as expected and almost marginal significant. Therefore, the PROCESS Macro was performed again and used model 1 with only cooking skills as moderator. The variation

suggestions as the independent variable and purchase intention as the dependent variable. However, in this analysis was the interaction effect also not significant (p = .102).

26 1 2 3 4 5 6 7

Without variety With variety

Pu rch ase in ten tio n -1SD Mean +1SD

Figure 7: Interaction effect between purchase Figure 8: Interaction effect between purchase intention, intention, variation suggestions and cooking skills variation suggestions and convenience orientation

4.4 Follow up analysis

Further analyses were conducted to check for the influence of other independent variables and predictors on the purchase intention for the meal kits. Furthermore, I verified if the participants are willing to pay more for the meal kit with variation suggestions and analyzed the convenience and degree of freedom that a meal kit offers.

Control variables

I assessed the influence of other independent variables on the purchase intention of the meal kit with variation suggestions.The first PROCESS Macro was performed again with all the participants (IV = variety, DV = purchase intention, moderators = cooking skills and convenience orientation) and included various control variables. First, a multicollinearity check was performed. Multicollinearity arises when there are very high interrelations among predictors (Malhotra, 2010). The variance inflation factor (VIF) will be used for testing our data set for multicollinearity. A VIF score above four indicates a moderate multicollinearity, and a score above ten indicates a strong multicollinearity. All the control variables had a VIF factor between the one and two, which means that there was no correlation between this variables. The six control variables were: physical activity, educational level, age, gender, hunger level, dietary restrictions, and health consciousness. The overall model is

significant, R = .303, R² = .092, p = .046. The R-squared is higher than the model without control variables, so this model with control variables will fit the data better. However, the control variables may not affect the outcomes of the experiment. Age was only marginally significant and negative (B = -.021, P = .051). This model shows that there is some evidence that the younger people are, the more

27

likely they are to purchase a meal kit (all participants above 18 years old). The results of the entire regression can be found in Appendix G.

Predictors

Other factors may also significantly predict the purchase intention of a meal kit. By testing different independent variables, it is possible to learn more about which factors increase the purchase

intention for a meal kit. The mean score of the purchase intention for both meal kits was used as the dependent variable.Different independent variables were used: desire to improve cooking skills, enjoyment of grocery shopping, number of days per week the respondent grocery shops, number of days a week the respondent cooks dinner, and the modal number of people with whom the

respondent eats their main meal. The highest VIF score is 1.265, so all scores are much lower than the cut off value of 4, which is needed to prove multicollinearity. Therefore, I conclude that there is also no correlation between the predictors.

A multiple regression was used to predict the purchase intention from these independent variables. These variables significantly predict the purchase intention (F(5,224) = 2.776, p = .019, R² = .058). The results show that desire to improve cooking skills has a positive and significant effect (B = .212, p = ,008). Thus, people that want to improve their cooking skills have a high purchase intention for a meal kit. Moreover, the number of days a week the respondents cook dinner has a positive and marginally significant effect(B = .116, p = .058). The findings suggests that the more often someone cooks, the higher their purchase intention for a meal kit. None of the other results are statistically significant. The results of the multiple regression can be found in Appendix H.

Willingness to pay

28 Figure 9: Willingness to pay

The analysis was performed again and the participants that answered “no” on the question about the manipulation check were removed. The independent variable, dependent variable and moderators stayed the same. 37 participants were excluded from this test (N = 193). The overall model is still not significant, R = .208, R² = .043, p = .139. Among only those who payed attention to the label, the difference in willingness to pay is significant (B = .317, t(187) = 2.203, p = .029). Therefore, it can be concluded that people are willing to pay more for the meal kit with variation suggestions. The interaction effects of the moderators are not significant. However, the main effect of cooking skills on the willingness to pay is positive and marginally significant (B = 1.118, t(187) = 1.848, p = .066). Therefore, there is some evidence that people with higher cooking skills have a higher willingness to pay for a meal kit. The results can be found in Appendix J.

These analyses were repeated with the addition of the control variables: physical activity, educational level, age, gender, hunger level, dietary restrictions, and health consciousness. The overall model is not significant, R = .287, R² = .083, p = .192. The willingness to pay is still significant (B = .313, t(181) = 2.174, p = .031). Other results are not significant, except gender is negative and significant, (B = -.651, t(181) = -2.071, p = .040). Therefore, it can be concluded that females want to pay less for a meal kit than males. These results can be found in Appendix K.

Degree of freedom and convenience

Based on the literature review, it is expected that the meal kit with variation suggestions is perceived as offering more freedom and less convenience. An independent samples t-test was performed which revealed how much freedom the respondents felt that the meal kit gave them. The

independent samples t-test showed that the convenience is higher for the meal kits without variation suggestions (M = 5.65) compared to the meal kit with variation suggestions (M = 5.20) with a

significance of p = .002. In addition, the meal kit with variation suggestions offers more freedom (M = 4.36) than the meal kit without variation suggestions (M = 3.83) with a significance of p = .003. Therefore, it can be concluded that the meal kit with variation suggestions gives more freedom and less convenience. 4.67 4.85 0 1 2 3 4 5 6

Without variety With variety

29

Furthermore, the study investigated whether people have a higher purchase intention when they believe a meal kit gives them a high degree of freedom or when they believe it is convenient. A multiple regression was performed. The above-mentioned variables were used as independent variables and the mean of the purchase intention of both meal kits was used as the dependent variable. The results shows that these variables significantly predict the purchase intention (F(2, 227) = 28.010, p = .000, R² = .198). If people perceive the meal kit as very convenient, this positive affects the purchase intention (B = .487, P = .000). Furthermore, if people believe the meal kit offers a lot of freedom, their purchase intention is also higher (B = .281, p = .000).

30

5. Discussion

Where the previous chapters have described the literature behind this research as well as the methods and results of the research, this chapter will provides an in-depth analysis of the results. The following subsections provide an overview of the theoretical contributions, managerial implications, limitations, suggestions for future research, and end off with the conclusions.

5.1 Main findings

The study aimed to investigate whether variation suggestions in a meal kit have a positive influence on purchase intentions. Another goal of the study was to verify if this effect is influenced by cooking skills and convenience orientation. The results of the analysis showed that no hypothesis was supported. Table 1 presents an overview of the tested hypotheses.

Table 1: Overview of study hypotheses

Subject Hypotheses Supported

Hypothesis 1 Variation suggestions in meal kits have a positive impact on purchase intention

No

Hypothesis 2 The positive effect of variation suggestions in meal kits on purchase intention is more pronounced for people with high cooking skills.

No

Hypothesis 3 The positive effect of variation suggestions in meal kits on purchase intention is less pronounced for people with high convenience orientation.

No

31

Furthermore, the moderating effect of cooking skills was tested and was also non-significant. Despite this, the results were still in the right direction about the first and second hypotheses. It may thus be argued that future research with larger samples could lead to significant results and prove whether these hypotheses are correct. A possible explanation regarding this result could that participants’ self-perceived cooking skills were found to be relatively high, reducing the difference between the groups.

No significant proof for the third hypothesis was found. There was very little difference in the purchase intention among participants who were highly convenience-oriented and the people who were low convenience-oriented. However, the main effect was significant. This is in line with the literature: people who are highly convenience-oriented have a higher purchase intention for meal kits because meal kits are highly convenient. Furthermore, in this study there was an online shopping setting and online shopping does not require much effort. In previous studies, consumers have cited convenience and time savings as their main motivation for online grocery purchases (Morganosky & Cude, 2000). Therefore, online shopping is highly convenient for these consumers. This is worth investigating in a field study to determine whether there is an interaction effect as there is may be a greater difference in convenience in the two different meal kits.

A further analysis was performed which proved to be fruitful as some significant relationships were established. The further analyses, examined whether other control variables influenced the

participants’ purchase intention for the meal kits with variation suggestions. The control variable of age had a marginally significant negative effect, which means that the purchase intention for meal kits is higher among younger people. This is in line with the literature review that the meal kits are most bought by young families and young millennials (de Sena, 2020). Hence, it is optimal to continue to focus on this group.

32

addition of variation suggestions and increased price, the profit can increase considerably on an annual basis.

Furthermore, a multiple regression was performed to determine which other factors influence the purchase intention for a meal kit. I found that people who want to improve their cooking skills have a higher purchase intention for the meal kits. Furthermore, as people cook often had a marginally significant and positive purchase intention for a meal kit. Other interesting factors such as enjoyment of grocery shopping and the model number of people with whom participants eat their main meals did not show statistically significant results.

Finally, the research assessed if participants have a higher purchase intention and willingness to pay if they perceive the meal kit as offering a lot of convenience or a high degree of freedom. Both convenience and the degree of freedom have a significant and positive effect on purchase intention. With the degree of freedom, there is also a positive effect on the willingness to pay. From this, it can be concluded that it is good for meal kits to be seen as convenient and ,therefore, have the

premeasured and prepicked ingredients and the recipe. However, it is also very important to ensure that people perceive it as a product with a high degree of freedom. Meal kits offer convenience, but consumers also value a high degree of freedom. Consumers are likely to prefer activities that given them the ability to make decisions (Deci & Ryan, 1985). Including this freedom in the meal kits would increase the purchase intention and the willingness to pay.

5.2 Theoretical contributions and managerial implications

This paper explores a field in which little research has been conducted since most of the existing literature examines the variety between products but not the variation suggestions providing variety within one product. Furthermore, meal kits are a new but fast-growing product therefore, there is little available literature about them. In this way this study extends the current literature and proves four main theoretical contributions and managerial implications.

33

should be borne in mind that women have a lower willingness to pay. Therefore, marketers should conduct research first to establish if meal kits are not mainly bought by women.

Secondly, it is important for marketers that they target the right audience. The purchase intention of meal kits is higher for younger people. Therefore, for marketers it is good to focus on the millennials. While convenience is a much sought after value by millennials, they also are more likely than other to seek novelty in everyday products. They want to try something different, a new food, an authentic recipe that can make the necessity of eating into experiential novelty (Bonsignore, 2019). Moreover, people who want to improve their cooking skills and people who cook often have a higher purchase intention for a meal kit. New recipes are also related to an ambition to attain personal goals, and interest in learning skills (Costa et al., 2007). These people do cook often and therefore want to put different meals on the table and will also improve their cooking skills by learning to make different meals. For these consumers it can also be important to offer a wide assortment in the form of different types of meal kits.

Thirdly, it is important for marketers to know what the ideal meal kit looks like. The results show that a meal kit with variation suggestions has a higher degree of freedom and less convenience. This was expected but has now also been substantiated. In addition, purchase intent was found to increase when a product is perceived as convenient or providing a high degree of freedom. Moreover, there is a higher willingness to pay if the product indicates a higher degree of freedom. Marketers should research how to ensure that the product gives a higher degree of freedom whilst retaining convenience. Marketers can use advertising on television and online promotion to create positive feelings towards the convenience or the degree of freedom of the product, leading to consumers being more motivated to buy meal kits.

To conclude, this study contributes to the novel literature on variety-seeking in a product. It gives insights into the effects of a new form of variation (within one product) and not variety-seeking between products. There are increasing number of products in supermarkets that offer additional variation suggestions. Therefore, companies should continue to offer these recipes or variation suggestions because this ensures consumers are willing to pay more. Furthermore, this research can be used by marketers and public policymakers to acquire useful insights into consumers’ attitudes toward variety and a better understanding of efficient ways to promote home-cooking.

34

strategies with higher degrees of freedom and convenience in meal planning and shopping to make healthy food choices more relevant and appealing.

5.3 Limitations and future research

The methodology and results entail several limitations that are relevant for future research. A first limitation is that due to time constraints the number of participants was limited to 230. It could be that this small sample size is not representative of society. This makes it hard to draw conclusions for a larger population. Additionally, the majority of the sample is highly educated and aged between 20-30 years old. Moreover, 80% of the participants or their household member(s) had purchases an in-store meal kit in the past. This is much higher than the national average. This may be because meal kits are mainly bought by young people and the majority of the sample is young. Therefore, this sample is not representative of society in which older people and those of lower socio-economic status are included.

These findings need further verification, and it is recommended to explore these findings more extensively with a larger and more diverse portfolio of respondents. Recreating the study with a larger number of participants would have a greater statistical value and could lead to more

significant results. Furthermore, a more diverse portfolio would strengthen the internal validity and would better represent wider society. Some results could be significant if a larger and more diverse sample size is used in future studies. Furthermore, subsequent research is always necessary to either verify or disconfirm the results of a study.

Secondly, the manipulation check of the variation suggestions on the meal kit shows that 32.4% of the participants of the experimental group had not observed the variation suggestions. Future research could consider this understanding and ensure higher visibility of the variation suggestions. For example, the suggestions could be made more colorful and it could be emphasized more clearly that participants should also read that section. The title of the meal kit could also state “Thai green curry fresh package with additional variation suggestions.” Further research could also show the recipes which might make the variation suggestions more noticeable and better represent the entire renewed package.

Thirdly, the research setup was designed, to imitate a real-life online shopping situation. The

experiment used an image and text from the Albert Heijn online shop. This online setting could have led to different results compared to a real supermarket shopping. This study was executed within a relatively short period and focused mainly on internal validity. Future research is therefore

35

different results as participants would be completely free to do their shopping. In that case, other products to a Thai green curry should be used, since it is possible that people want more variation suggestions within one product than within another. Furthermore, it is also possible to investigate whether purchasing the same meal kit each time leads to satiation. The focus may then be more on the repurchase intention than the purchase intention. In addition, it can also be investigated whether it is wise to make minor changes in the meal kits every month and to use seasonal products.

Furthermore, this was the first study to research whether variation suggestions within one product lead to a higher purchase intention. It is recommended to conduct further research on this topic and replicate this study, to find consistent results. Further studies are necessary to verify the conditions that moderate the effect of variation suggestions on purchase intention. Moreover, it is important to find conditions in which this moderation effect would be supported and investigate other

moderators, such as cooking enjoyment or desire to improve cooking skills. Earlier research indicated that variety-seeking can be driven and affected by exogenous factors (Sevilla et al, 2018). Therefore, this study could be replicated with other moderators. In this research, the moderator of cooking skills has a low Cronbach’s alpha, so a different cooking skills scale should be utilized when using cooking skills as a moderator in a future study. It is possible and expected that the results will be different if any of these aspects are implemented in future studies.

36

5.4 Conclusion

The goal of this study was answer the following research questions:

1. “What is the effect of variation suggestions in meal kits on the purchase intention?” 2. “How is this effect influenced by cooking skills and convenience orientation?”

Variation suggestions in a meal kit do not lead to a clear increase or decrease in purchase intention. However, the findings provide evidence that the willingness to pay is higher for the meal kit with variation suggestions. Therefore, managers should offer variation suggestions on their meal kit. Variation suggestions in the meal kits can influence the potential success a meal kit might have. It takes little effort for managers to add variation suggestions to a meal kit and the profit would increase. Marketers can also investigate other small adjustments that would justify charging a higher price. The findings provide no evidence that the people with high cooking skills have a stronger preference for the meal kits with variation suggestions, not that convenience-oriented people do not care which meal kit they get. So, the variation suggestions on the meal kits are no problem for the convenience-oriented people, therefore the company’s do not have to make other meal kits for this group of people. It would be interesting to replicate this study with more participants and as a field study to get deeper insights about the consumer behavior in a real-life supermarket.

This study contributes to theory and practice by suggesting techniques that would allow researchers and managers to motivate consumers to cook for themselves more often. People have a higher purchase intention if they perceive a meal kit as convenient or as providing a high degree of