Analysis of Product Packaging Design

1 Introduction

Product design and specifically product shape and 'looks' have long been identified as factors that may contribute to product value and new product success. Design of products evokes both cognitive and affective responses in the mind of the observers and this can be used to tailor a more attractive product proposition.

While a lot of excellent research has been conducted on the positive effect that industrial design can have on the perceptions of customers about the product functionality, embedding issues like utility, safety and comfort, the importance of the perceived value by a customer on judgements about product elegance and social significance have not been extensively studied until recently. reference

In this thesis I am trying to test whether Yes please soup pot design not only communicates to the potential customer a series of qualitative attributes about its content, i.e. quality and healthiness, but also triggers positive emotional responses on the perceived beauty and difference with similar products and that can be leveraged by the company to command a price premium

2 NEW PRODUCT DEVELOPMENT AND PRODUCT DESIGN

New product development processes are the subset of standardized procedures that companies use to manage the new product project phases that lead to the launch of new products in the market.

The objective of these procedures is to implement a systematized approach to ensure the potential of new projects based on their financial and development feasibility while maximizing the value of new products as perceived by its target customers.

2.1 What is industrial design

Industrial design is the set of activities within new product development processes that deal with optimizing the functionality and appearance of a product to maximize its value for both consumer and manufacturer (1).

The ultimate objective of a product design is thus to align the set of attributes embedded within the product with the target customer preferences and to implement them in a way that they are actually perceived and valued by them. In order to achieve a successful design and implementation, companies engage in direct market research to elicit the target customer segment implicit and explicit needs ('the voice of the customer') and align those with product attributes, using techniques such as Quality Function Deployment (QFD) which systematically links the needs the product must satisfy with technical specifications while also prioritizing them based on the level of importance to customers.

At the same time, the identification of the customer segment preferences and the mapping of those within a perceptual map, comparing how well different products in the market fulfill the identified needs, allow for the design of specific product propositions that no other product does and thus achieving a unique positioning and successful product differentiation.

The dimension of product design has been recognized by several authors (Cooper, Trueman) as being critical to the ultimate success of the launch of new products.

The focus of design development is centered around the efficient implementation of the product features, ergonomics and quality form to maximize its utility to users, while at the same time embedding it with a pleasant appearance that is able to communicate positive attributes that contribute to the ultimate value proposition. (3) (4)

Trueman: "Design has the facility to improve product reliability and quality standards thereby raising the perceived value of goods and services in the eyes of the customer, allowing companies to increase profit margins"

As Trueman estates (6), A value proposition must successfully integrate a product within its own environment by combining and merging coherently the "different attributes, aesthetics, price and quality so that they are aligned similarly and reinforce each other".

Although mistaken by artists that only worry about the visual appearece of a product, or styling, successful industrial designers are able dig into such fields as engineering, Materials science, manufacturing, and marketing to embed new products with a set of attributes directly influencing new product success in the market.(2) (3) (4) (5) (6). Ultimately, the design of an object is "the specific configuration of elements, materials and components that give its particular attributes of function, shape etc. and determine how it is to be made and used". (13)

By embedding the design dimension into the processes, companies ensure that the final value proposition is increased as it contributes to the perceived value by the customer. A successful design increase the perceived quality of a product, ensures that is aligned with market and regulatory standards and thus increase the odds to satisfy customer expectations.

A consistent Design strategy in new product development processes also contributes to build a product and company image and helps to pull together the dimensions of company identity with branding and promotion (Trueman).

Also, by taking the design dimensions early on the NPD projects, companies can reduce the final time to market and product costs by simplifying the manufacturing processes and reduce the final costs of fabrication.

2.2 The contribution of Industrial design in the final product value proposition

There are many design attributes that can be embedded into new products, roughly separated within ergonomics and aesthetics, being the former more related to the experience of using the product, while the latter is focused into the experience of seeing the product. Aestethics, embedding all product parameters that determine the way the product look, are a an essential element of the purchasing process since customers base their preference on products by the subjective perceptions elicited by the product on the potential benefits it can provide.(7)

In that line, the Lens model first introduced by Brunswick (?), states that the potential customer makes a mental bundle of the information it receives about the product and from there triggers a set of perceptions that will ultimately lead to a set of preferences and choices.

The ways a specific design can lead to a positive perception and thus to a choice of preference vary and are entangled with other sources of information the customer receives and which align the propostion to the customer already decided preferences.

The perceptions that a product can evoke are immediately related to past information received and allow the person for example to relate it to a certain corporate and brand identity, a process that many companies have followed by implementing a sustained design strategy on their products. Brand identity allows to ultimately link the products observed to perceptions on company values and overall level of attributes of the products and has been used as means of effective product differentiation. (9) (10).

In different industries, companies tend to emphasize different attributes in their communication to be aligned with their specific company positioning and customers most importance preferences, like tastiness and safety in the food industry and reliability and environment friendliness in the car industry.

The physical form of a product has been researched to have an important impact in the way customers judge it and has ultimately a strong correlation effect with the final product success in the market ().

As Bloch (10) states "the physical form or design of a product is an unquestioned determinant of its marketplace success. A good design attracts consumers to a product, communicates to them, and adds value to the product by increasing the quality of the usage experiences associated with it."

2.3 Which financial benefits can it provide

Companies with an effective industrial design strategy achieve better perfoming products in the market in terms of several financial indicators as return on assets, return on sales and higher profitability, which can be linked to both the design differentiation factor as stated previously by Porter in the famous book Competitve strategy (1980) and to reduced costs due to more efficient use of materials and manufacturing processes. (14)

Also, the study of Roy (13) in 1993 on 221 small and medium sized UK manufacturers which received a government subsidy to promote the active use of industrial design in the development of new or improved products showed that 60% of all projects and 90% of the implemented ones were commercially successful and profitable with payback periods averaging under 15months, which show that the effective strategic approach to include design in new product development processes can be implemented in firms of different sizes.

Bloch in his research also collected previous studies that linked new product financial success factors with the inclusion of design as an inherent part of their NPD processes. He identified in a survey of senior marketing managers that, design was mentioned as "the most important determinant of new product performatice by 60% of respondents" by " only 17% considered Price most important" . Also and based on the work of cooper on the analysis of the performance of 203 new products identified that product design was the most important determinant of sales success . Most interestingly for the case of Yes Please foods product, which as will be explained later chose specifically a designer for the pot based on his previous award winning record, Some research has identified that the receipt of design awards is positively associated with profit margins above average and sales growth (Goodrich 1994; Roy 1994).

3 The purchasing process and the visual effect of a product design

3.1 Customer needs

The ultimate act of purchasing occurs as a result of a complex mental process where the information received is analyzed and weighted as per to measure to which extend the product satisfies the needs of the customer.

A general categorization of customer needs has been frequently compared with the Maslow hierarchy of needs which states that once most basic requirements have been satisfied by a product, the emphasis on a customer shifts to satisfy other more intangible needs related to symbolic and aesthetic attributes.

As a result of this the purchasing process is triggered by the fulfillment of the requirements for the intended use of the product but also by the satisfaction of more intangible needs like status, elegance or social significance.

In order to understand the ultimate behavioral response of a customer triggered by the visual appearance of a product it is critical to assess the cognitive and emotional processes that result from the act of observing the item under evaluation.

The cognitive processes take place when a customer uses his visual senses to observe the product and perceives certain information which mentally organizes to make some judgments about its attributes and which are influenced by previous visual references or similar product stereotypes, which suggest familiar usages of the product and ultimately help the observer to interpret the signals received.

It has been described a number of different approaches on how to categorize the judgments that a customer does based on the perception of a product observation. Crilly (?) has summarized all previous approaches and identifies a total of three main categories of cognitive responses to product appearance: Aesthetic, Semantic and Symbolic.

From those three, the semantic interpretation, the mental inferences that an observer does to judge whether a product is capable of performing the tasks for what is intended for, is the only processes where the tangible attributes of the product are assessed. During this process the practical qualities of a product like function, performance and efficiency are analyzed and mentally compared with other references to judge the utility a product will offer to the observer. In this category I include the information that is gathered by the customer when obtaining information from reading the label and which is directly processed to identify the physical attributes of the product. The emotional responses following this cognitive process are then aligned to assess the instrumental utility of the product which ultimately lead to satisfaction, when fulfilling the expected requirements, and dissatisfaction when the product is not fulfilling them.

The two other described cognitive processes are used to identify intangible attributes of the product that may or may not be perceived as valuable for the customer depending on a number of different factors, like current 'positioning' within the Maslow hierarchy of needs, consumer's cultural context and personal characteristics.

The symbolic association is the cognitive response that attaches to the product some socially determined symbolic meaning. During this process, as series of values are realized to be attached to the product and assumes that others must also associate them with it. As Crilly states 'This culturally agreed meaning will allow the customer project a desirable image to others, express social status or communicate its personal characteristics through it'. Examples of intangible values that can be associated to the product through symbolic meaning are exclusivity as the identification with certain economic status and environmentally consciousness.

Finally the aesthetic impression comprises all cognitive responses that are directed towards a perceived judgment of elegance and beautifulness. Even if still there is no unanimous consensus on what comprises beautiful objects, the perception of aesthetic attraction triggers positive emotional thoughts on the customer and contribute to attaching value to the product observed.

3.2 The Aesthetics Dimension

As researched by several authors (?) the definition of what makes an item beautiful or aesthetically pleasant is not conclusive.

It has been described though that cultural and social forces have an influence in the preferences for specific forms. Specifically, it has been described that a specific culture values and preferences may influence the acceptance of a particular style. Also seems to be proven that cultural norms may overwhelm an individual inner preferences and help shape its perceptions towards the acceptance of a specific design form. (blaich, Bloch)

Thus, and although cross cultural differences stay in the way of having an unified view of what can be considered as aesthetically pleasant, the current era of advanced information technology is working towards unifying the concepts that influence the perceptions of the soft values within a product design and thus working towards a more globalized and uniform set of criteria.

The cognitive processes described triggered upon the observation of a product lead to a series of emotional responses that will ultimately lead to the final decision on the purchasing process, being the most important the attraction or disgust towards the aesthetics, the satisfaction or dissatisfaction towards the fulfillment by the product of instrumental requirements for its use, the surprise or indifference based on the perceived product novelty and the admiration or indignation towards social significance.

4 DESIGN IN THE FOOD INDUSTRY

In the food industry, Tauber (8), collected through extensive market studies an exhaustive set of problems related to food products. By doing so, he was trying to identify potential opportunities for new products while also providing with a thorough analysis of the basic needs to be fulfilled by this kind of products. From that list it can be identified a series of attributes that the product under study is fulfilling and thus achieving a specific positioning. Some of them, like low calorie content, convenience of transport and preparation and adequacy of the serving size are efficiently communicated through the information contained in the label. Others like healthy, tasty, and high quality can only be perceived by the customer through the design of the pot and label and by some previous information it might have received.

Moreover, as Berkowitz researched, the shapes and images that have a more natural looking are associated with products that are fresher, taste better and have a better texture, and are ultimately more preferred by customers. At the same time, he found that aesthetic appealingness of the product, defined by bloch (10) as " the ability to evoke positive beliefs, positive emotions and sympathetic with consumer's aesthetic tastes", and by Crilly (11) as "the sensation that results from the perception of attractiveness (or unattractiveness) in products" was not of any specific interest to food products customers.

5 The Yes Please! Foods Company, products and market

The trigger for the writing of this thesis has been evaluation of the influence that the design of the Yes Please! Foods soup product packaging may be having in its market performance. Yes Please Foods is Germany based company that entered the chilled organic convenience food German market in 2007. Although belonging to the generally mature food industry, both the organic and convenience sub segments have been growing at an average rate of 12% in the period 2003-2008 in Germany (15). Yes please foods can be considered to be positioned in the Premium priced range of offers within the segment as its Price its on average 33% more expensive than that of the competitors. The company started its operations in the Berlin area and has been enjoying a high growth rate in volumes sold ever since (the actual numbers can not be disclosed in this paper as they are covered by a confidential disclosure agreement). The company is currently planning to expand its operations throughout Germany and is about to close its first round of external financing.

Based on the information provided by the company, it is know that the customer profile of this kind of products is that of mid to high educated people, working full time, with 'above average income' and health conscious. It is inferred from the information provided by the company that the product is intended to fulfill the need of high quality healthy food that can be prepared and ready to serve within a short period time frame. The motto of the company portrays this positioning of healthy convenient to prepare food for customers that do not want to spend time in the kitchen as "good food for busy people"

The design of the pot, as explained by the company owner and general manager, Gemma Michalski, was the central part of their strategy to build a company identity and resulted in one of the biggest expenditures incurred during the initial company launching. The design was commissioned to Willimas Myurray Hamm from London which used the artwork of Berlin based illustrator Martin Haake. As explained by Ms Michalski, they were specifically chosen to perform the design task because of their previous record of award winning product designs that resulted in highly successful products in the market. (16)

6 Use of conjoint analysis to elicit the importance of attributes in the eyes of the costumers

The first part of the field study I am going to conduct has been designed to measure a the potential correlation between price and design in the final willingness to pay expressed by respondents. Also I shall use the results to mesure the relative importance of each, for the whole of respondents and for potential segments I shall identify within the sample of respondents.

The technique used, an empirical marketing research using conjoint analysis, was first introduced as an effective marketing tool during 1970's and has been validated as an effective means to identify the most relevant features of product, its key design attributes, and the degree of importance that customers attach to them. (17) The usefulness of the tool is extended by the ability to sort the answers from respondents to some specific criteria which allow eliciting the kind of different features preferred by different segments of consumers within the same market.

At the heart of the technique, consumers are asked to rate as per their own preferences a number of different product prototypes that are embedding different features and levels of features. By doing so, the customer is making choices and trade-offs from those multi attribute alternatives based on the overall perceived utility or value of the product under evaluation. The statistical treatment of the data, using a multiple regression model, allows quantifying how much each of the single attributes is affecting the overall value of the entangled set of properties, as it is assumed that consumers have an implicit utility value for every single one of them.

The first part of the technique is normally a consumer attitude survey were the general attitudes of the consumer towards the product are collected. One of the main outcomes of this part of the research is to determine which features of the product are relevant for the consumers.

The aim of this first part is typically to find out why the products are purchased, which use they make of them and their attitudes toward them. Once the information is collected allows the design teams to elicit which features of the product seem to be more relevant for the customers and allows to potentially determine needs that still unresolved or problems existing with current products in the market. In this thesis I have overcome this phase since this part of the research is aimed at determining which physical features of the product are relevant, whether the focus of this thesis is aimed to the 'soft' values attached to physical appearance. Thus, by assuming that the current most important features of the product for the target customer segment are actually satisfied by all products in the study, I have been able to focus any significance preference in the actual aesthetic value of the product and its potential relationship with the price they would be willing to pay. In order to communicate to respondents that all products evaluated had exactly the same features and were only differing in design and price, the following statement was introducing the questionnaire:

'The products you are going to see are all soup products. They contain just natural organic ingredients without any conservatives and must be kept in the fridge. The portions are all 500ml'

Generally, The second part of the conjoint process uses the information gathered to determine the whole set of attributes that define all of the existing products in a market and introduces new ones to test their acceptance by consumers. Also, different levels for each attribute are defined to obtain a meaningful representation of the different ranges within each attribute that are or could be available in the market.

In order for the research to be significant, prototypes having different combinations of levels of all attributes have to be created to be ranked in preference by consumers. Typically, and due to the large number of possible permutations of attributes that can be created, a smaller sample is chosen to facilitate the consumer research study. It has been shown that eliminating combinations through an experimental design called orthogonal arrays or through judgment (those that are not possible physically i.e. by cost or conceptually i.e. by design), has no significant effect on the final outcome of the study (18)

Since the research to be conducted for this thesis is aimed at identifying any preferred designs for a soup product and potentially monetary value attached to specific product appearances, I have chosen to study two sets of attributes, design comprised by the three levels, 'picture of natural ingredient', 'artistic draw' and 'no draw' and the attribute price, also with three different levels: '€ 1,99', '€ 2,49 and '€ 2,99'. Due to the small number of total possible combinations (32), the empirical survey shall ask respondents to evaluate all possible combinations.

The rationale for choosing these type of designs has been based on previous literature on shape and images in food products (Berkowitz) and the need for the actual inclusion of the design of the Yes Please Food product to test its hypothesized perceived value on the design. The third design included 'no draw' has been arbitrarily selected hypothesizing it to be a representative sample of an unaesthetic design.

The three price level selected have been chosen on the basis of actual prices of products in the market for the 'picture of natural ingredient' (€2,49) and 'artistic draw' (€2,99) designs, while the third level price has been arbitrarily selected to represent a low price level within that attribute.

The third part of the procedure gathers a meaningful sample of the product consumers and asks them to rank the different prototypes (combinations of the different levels of the selected attributes) based on their preferences. The aim of this part is to gather the structure of consumer's preferences for different product features. In this part is important to define the question to be asked properly so it collects the opinion on consumers about the perceived value they attach to each specific product proposition. In my research I have chosen the sentence 'from 1-7 how likely would you be to buy this product at the stated price'.

One of the benefits of conjoint analysis is that it is able to achieve statistical significance on the results with a relative small sample of respondents. The aim of this research will be to achieve at least 33 respondents in order to be able to make some inferences about the direction of the proportional influence that design has in the final monetary value of the proposed prototype, being that either positive or negative.

One of the limitations of my study shall be that the random sample of respondents to the questionnaire shall only be validated as actual consumers of soup products by one of the questions in the demographic profile, ' do you like soup?' that is embedding three possible answers: ' no', 'sometimes is ok' and 'love it'. It shall be assumed that a positive answer to this question (all possible answers but 'no') allow to make inferences about their potentiality to be consumers of the product. That limitation is affecting jus this part of the study were the research is trying to elucidate whether design has a relationship with willingness to pay.

For the second part where it will be researched the consumer's perceptions about intangible attributes of the design, it will be assumed that cultural context and general profile of the respondents is similar to that of the consumers of fresh soups as explained before in the Yes Please Foods product and market chapter of this thesis.

A segmentation of the respondents by any kind of useful criteria like demographics, type of usage, attitude towards the products etc will also allow to identify the preferred attributes for each type of customer segment.

The final part is the statistical treatment of the results that tries to identify which attributes are preferred by consumers and which are considered to be of more relative importance to them, and thus to the final value of the product proposition.

7 Experiment

7.1 Objectives

The objectives of the experiments executed were designed to validate and refute previous research about the influence that design have in the purchase decision making process and to elicit specific findings about the design of the soup product of the sponsor company Yes Please! Foods.

In the first experiment, the objective is to validate whether a series of qualitative attributes about the product can be inferred from the customer just by looking at the design, specifically the respondents are required to rate each of the three designs as per their perception on how they consider them to be 'healthy, 'fresh, 'of high quality', 'different' and 'beautiful'.

One of the objectives of this part of the experiment is to validate the hypothesis that the design of the pot of the Yes Please! Foods soup product has is somehow original that can be clearly differentiated from that of the competitors and that is considered to be an aesthetically pleasant design.

Another objective of this part of the research is to validate previous research like that of Berkowitz (9), Bloch (10) and Trueman (6) which asserted that in mature markets, product form is one way to gain consumer notice and achieve a clear product differentiation.

By assessing the responses about the attributes quality, freshness and Healthiness, it is pursued to validate the research of Bloch (10) and Nusssbaum (?), which stated that exterior appearance of a product is an important channel to communicate information to consumers, that Product form allows to generate inferences regarding other product attributes. Also this experiment will help to validate the research from Berlkowtiz () which found that natural shapes displayed in the packaging of food products help consumers to make assumptions about the product as being more fresh healthy and of higher quality. Finally I intend to validate the research of Trueman (6) which found that products that are considered to have a good design (which I shall relate to the weight of the responses on the attribute 'beautiful) are considered to be of superior quality, by checking whether products to be considered 'beautiful' on my research are also considered to be of 'high quality'.

The inferences that I will try to make with the results of the second part of the research, a conjoint analysis of 9 different prototypes which result from the combination of three different design styles and three different sets of prices, will be dependent upon the results of the first part. If during the first experiment I am able to proof that some of the three designs are considered to be significantly more 'beautiful' than the others, I shall be able to validate with the results of the second experiment are aligned with the results found by Bloch. Kotler and Nussbaum when they found that given two identical products in terms of features and price, the one with the most beautiful design is preferred by respodents.

At the same time I shall try to validate whether a product considered to be more 'beautiful' can command a higher willingness to pay, which is stated in the survey designed as the likeliness to buy the shown product. If that is proven to be the case, I will try to identify the reasons why those customers would be willing to spend more in such product by relating it to their perceptions on the other attributes of the same product, either being the subset of attributes regarding product utility, (higher quality, more natural or more healthy) , the subset of aesthetic attributes (more beautiful, more different) or both.

Finally I shall use the results of the second experiment to validate the research of Berkowitz which found that products with natural shapes tend to be considered more natural, and healthy. I shall do this by specifically analyzing the results of one of the chosen designs which portrays a picture of a natural tomato on the label. Also I shall try to refute the findings of the same author which found in previous research that attractive designs where not of any specific interest to consumers in terms of aesthetic appeal.

7.2 Methodology

7.2.1 Sample

The survey will be send to all recipients of the address domain all.berlin@esmt.org, for which I expect a random sample of respondents within the employees of ESMT and all recipients of address mba2009@emt.org, for which I expect a random sample of ESMT MBA 2009 students I believe this sample shall be representative to elicit some significant differences on the aesthetic attributes of the prototypes studied. Since the ideal sample of respondents should target validated customers of the segment of chilled, all natural, organic soup products, inferences about any significance results on higher willingness to pay for a specific product prototype shall be taken in perspective.

7.2.2 Experiment

The Kansei part of the experiment will show three differently designed soup products, each at a time, and respondents will be asked to rank it from one to seven as per their perception on how much they considerer to possess certain attributes. The attributes to be evaluated are

• Healthy
• Fresh
• High Quality
• Different
• Beautiful

The prototypes have been chosen from actual soup products existing in the market which differ on their label design by specific aspects.

Product one, "mini pot noodle" from an unknown manufacturer, has been chosen to represent the hypothetical low design product. It portrays no pictures nor drawings and all visual information is limited to a set of texts of different sizes and colors and different color backgrounds.

Product two, SuppenGlück, is a fresh soup product from an unknown manufacturer that is currently commercialized in the Austrian retail market at a price point of €2, 49. This design has been chosen as a representative prototype of food products depicting images of natural ingredients on the label. The design of it shows the picture of a tomato together with the main ingredients of the soup recipe, tomato and basil.

Product Three is the actual product that Yes Please Soups is currently commercializing in the German market as a chilled, organic and with no conservants soup. It is currently sold through mainly Bio products specialized shops and has a current price point of € 4,99. Its label design depicts the contour of a man and resembles the draw of a child. I have hypothesized that this kind of design can be considered away from the standard designs of food packaging and thus may score high on the attribute 'different'.

Once the data of all respondents is collected ...

The conjoint part of the experiment will show the nine possible combinations of the two attributes selected, design type and price level, and respondents will be asked to rate them from one to seven as per their likelihood of purchasing such product.

Both the kansei part and conjoint part of the experiment have been uploaded to the web based survey program Qualtrics () that will allow respondents to answer the questions on-line.

The study is including a demographic and attitude questionnaire for the participants to respond. It includes a number of question that shall allow the gathering of data in clusters to analyse potential subsegments among the respondents.

The first three questions of this survey are general demographic questions, gender, income and marital status. While gender and marital status have not been identified as a selective parameter that define the type of customers of fresh soups, they have been included to evaluate possible differences among gender in the perceptions of the intangible attributes of the product.

The Income parameter has been included on the base of the information provided by the Yes Please! Foods company as being on key defining element of the customer profile of their products. Based on the information gathered from respondents it will be analyzed any possible trend of perceptions on product features based on this parameter.

In order to allow inferences about the respondents being potential customers of the products under study the question ' Do you like soup?' has been included. As explained in the limitations part of this thesis, although a positive answer to this question shall not allow to make direct inferences as per being potential consumers of the chilled organic soup category, it shall suffice for the purpose of this thesis

The last question of the demographic part of the survey, 'have you ever bought for yourself an Ipod, Iphone or a Mac computer' has been decided to evaluate a potential segmentation of respondents on the base of preferences for attractive designs. Thus, it has been arbitrarily inferred that the products mentioned from Apple Inc. company possess an intrinsic attractive design embedded.

7.3 Results

7.3.1 Results analysis using descriptive statistics

The questionnaire was launched online on Tuesday November 17th and collected on Thursday November 19th 2009 and was responded by a total of 49 people.

Revising the general data for all respondents it is found that there exist significant statistical differences between the overall willingness to buy each of the product prototypes presented with a specific design at an equal level of price in all but in one case: the draw design at the high price versus the picture design at a high price.

It can be stated that this findings are aligned with what had already been described by Bloch. Kotler and Nussbaum so given two exact same offers, consumers will end up choosing the one with what they consider to be a more attractive design. The case of not finding statistical significant differences when comparing at the high price level the designs of draw versus picture, could hypothetically be explained by having reached the top price point where price becomes high enough to make the overall offer unattractive and thus all offers regarded as similar. The utility to the consumer of each of the attribute's levels will be explained and measured during the analysis of the respondents using the conjoint methodology. For now, it may suffice to explain that each level of the two attributes has a fixed utility to the consumer, and by adding up the two utilities that represent for them one specific design and one specific price, the overall value of the offer can be measured. Since the survey was explicitly asking about their likeliness to buy such product, it might well be that as prices increases overall utility decreases for a given design down to a point where their likeliness to buy it drops below the medium point and this towards the 'unlikely' zone, making offers equally unattractive.

When comparing the responses from all participants for same designs at different price levels, it has been found there were no statistical significant difference for any of the designs from low price level to medium and from medium to high, in the latter excepting the picture design. There were significant differences though in all designs when comparing the low price level with the high one.

Again these results could be explained by realizing that if the utility added or subtracted by a different price is not making the offer significantly more or less attractive, the overall offers will be regarded as similar. It can be inferred from this results that generally speaking for the whole of the sample of respondents, the design attribute could be of much higher relative utility to consumer compared to that of the price. Since the prices used in the survey were actually existing in the market, this results would confirm the relative importance of design in the market success or failure of products. This assumption has been later confirmed when analyzing the results from the conjoint methodology.

When revising the average responses for each type of design, it is found that there is no design found likely to be bought (average of 5 or higher), with the picture design scoring on average 4,2, but still not significantly different that 4 (undecided).

Since the overall objective of this thesis is to find the difference in perceptions across different types of consumers, the next step has been to cluster the respondents among groups where they had a clear preference for one of the designs, measured as their willingness to buy it.

7.3.1.1 Picture Lovers

I have called picture lovers the segment from the total respondents that have scored significantly higher than the average to their likelihood of buying the product with a picture on the design, regardless of the price (+25% with p value of 0,0001). They show no significant preference for any of other two compared with the average respondent and moreover are significantly more willing to buy their chosen design over the draw one (+47%) and over the Letters one (+125%)

The differences on their perceptions of the picture's pot to be more healthy, fresh, of high quality, different or beautiful are not significant compared to that of the average respondents.

On the other hand they have significantly scored their choice of pot higher in all those intangible attributes compared with how they regard the other two, excepting that they don't perceive significant differences in the attribute 'different' from their pot to the draw design.

There is a strong evidence thus that this group, although having a mainstream perception of the qualities attached to the product with a picture, they basically 'downgrade' the others, maybe because of the reason of not finding them appropriate enough or aligned with their expectations.

The demographic data collected do not allow to further identifying the profile of consumer belonging to this group as there are no statistically significant differences with the average respondent.

When it comes to analyze the attitudes of this segment towards prices, I have found that that they significantly prefer the low price over the medium (+47%) and the medium over the high (53%).

When running a correlation analysis between the results for 'beautiful' to the results of 'healthy', 'fresh' and 'high quality', I have found statistical significance with all of them, specially with 'high quality' (p value 0,0091). During the assessment of what is triggering the willingness to buy the picture product, I have found a strong correlation between their perceptions of 'high quality' 'fresh' and 'healthy' and no correlation with ' beautiful' nor 'different' and their willingness to buy the product. Interestingly though and as mentioned before, their perception of 'beautiful' is correlated with 'high quality'.

Assessing the influences of their perceptions about the Draw design in their lack of willingness to buy it, I have found a very strong correlation between all intangible attributes and their purchasing behavior, especially in their perceptions of 'beautiful' and 'high quality'.

Finally when assessing what is triggering their lesser purchasing tendency towards the design with letters, I have found a strong correlation between their perceived 'beautiful',' High quality', 'fresh' and 'healthy' and their purchasing behavior towards this product. It is interesting to note that this segment had found this product to be significantly less 'beautiful' than the average respondent (-21%)

7.3.1.2 Draw Lovers

Following the same criteria to define this segment as with the previous one, the Draw lovers have expressed a significantly higher likelihood of buying the product with a draw in the design than the average respondents for that same pot (+37% with a p value of 2,23x10-5). In this case though we find that they significantly perceive the product to be more healthy (+19%), fresh (+23%), of high quality (+27%) , different (+20%) and beautiful (29%) than the average respondent.

In order to relate this findings with the work of Bloch (), which found that designs that evoke positive beliefs and emotions and are in line with consumer's tastes help them to make inferences about the product attributes, I have performed a linear regression with the rating of 'beautiful' against the ratings of 'healthy', 'fresh' and 'high quality'. The findings have been that 'beautiful' is just significantly correlated (p value 0,0077) to 'high quality', again in line with the findings of Trueman ().

Their taste for aesthetics is shown also when they express their perceptions about the product with design letters. They find particularly ugly (-39% in 'beautiful') and, if you allow me the expression, vulgar ('different' -37%) than the average respondent.

These perceptions on the letters design may be somehow affecting their willingness to buy it. It is 52 % and 49% smaller than their preference to buy the draw and the picture designs respectively.

The reactions to this segment to the pot with picture are positive and as a matter of fact, have a higher willingness to buy the picture product as compared with the average respondent and their willingness to buy it doesn't differ significantly from the one to buy the product with a draw.

When comparing their perceptions about their preferred product to buy (draw) to the one with the picture, there are no significant differences in all intangible attributes but one. They see both equally healthy, fresh, of high quality and beautiful but the draw product more 'different'. As a matter of fact most of respondents within the draw lovers segment are included within the picture lovers segment as well, so they share the same perceptions about the picture design.

Interestingly and bringing together these findings with the ones from the picture lovers segment analysis, the two groups agree into finding the design of draw significantly 'different' and this perception is correlated with their likeliness to buy it (which is within the likely range for the draw lovers and within the unlikely range for the picture lovers).

Translating these findings into buying behavior, it can be inferred that draw lovers find both the picture product and draw product mutually attractive, and that their positive perceptions on 'beautiful' and 'different' and influencing their likeliness to buy both.

Differing again from the picture lover group, this segment doesn't have a significance preference between the low and medium prices and between the medium and high prices. The only significant difference found has been between the low and high price levels (p=0,0043). Also this group has shown a significant higher likeliness to buy any product at any price point, while at the same time not liking soup more than the average respondent.

From the demographic data we find that the draw lovers have a significantly higher income that the average of respondents (p value 0,0426), which is curiously related to the customer profile of the Yes Please! Foods product in the market, as the sample that answered the survey was not previously filtered as potential purchasers of these kind of products.

7.3.1.3 Letters Lovers

I have called letters lovers the segment from the overall respondents that have shown a significantly higher preference to buy the Letters design product compared to the average respondent (+36% with p value of 3,5x105). This segment shows no significant differences with the average respondents as per a preference to buy any of the other two other designs.

When assessing how they perceive their product of choice compared to the average respondent, significant differences have just been found for 'healthy', 'fresh' and 'high quality', and thus not for 'beautiful' nor different'.

When assessing their behavior towards price, I have found they significantly prefer over the average respondent the low priced products (+26% with p value of 0,0023) and the mid priced products (+20% with p value 0,0303)

Overall the conclusions that can be inferred from evaluating this segment are that they consider the physical aesthetic attributes as a secondary element which is not triggering their purchasing behavior, and do not have high expectations on the design of the product to be aligned with its physical features. One hypothesis to explain their perception of fresh healthy and high quality, besides its relationship with their perception of the product design appearance (which they would consider to be normal), would be the fact that in the survey all products where introduced as to be fresh products with natural ingredients, so one could nickname this group also as the 'believers'.

Assessment of relationships between attributes and intended purchasing

In order to assess how their perceptions on the set of intangible attributes are affecting their scores on the likeliness to buy the products I have proceed through two steps. First I have calculated, for each of the segments (Picture lovers, Draw lovers and Letters lovers) a correlation table among all of the intangible attributes they perceived for each design, in order to analyze whether they were regarding some of those attributes as similar to each other. I have used a rule of thumb of a score of 0,85 or higher to consider two attributes highly correlated, and between 0,8 and 0,85 to be moderately correlated. The results for the correlation tables can be seen in the tables

Secondly I have proceeded to run a linear regression between the set of five attributes and the average scores on the likeliness to buy each design for each respondent from each segment, and for all of the three designs (one design at a time).

Assessing the results of the linear regressions, the model has shown correlations ranging from adjusted R2 of 0 to 0,715, depending on the segment and the design analyzed. Generally the lowest correlations have been seen for the letters lovers segment in all of the designs and for the picture lovers segment when assessing the design Picture. Also, The coefficient for the attribute 'high quality' is for all segments and all designs the highest when correlating the attributes to the likeliness to buy the product. It might be inferred that this findings indicate that the perception of a food product such as the ones presented about its apparent quality is key trigger in the likeliness to buy.

In order to find out whether the perception of some other attribute was influencing the perception for a specific design to be of 'high quality', I have run a linear regression for all designs and segments between the attributes 'healthy', 'fresh', 'beautiful' and 'different' and the attribute 'high quality'.

When analyzing the data obtained some general trends can be observed by identifying, for each segment and each specific design, the few other attributes that have scored 10 times or less than the 'high quality' attribute (the 'significant' range).

One of the main findings of this part of the analysis is that for all segments the perception of 'beautiful' becomes one of highest influencers in the perception of 'high quality' when assessing the products they showed preference into buying. This results are in line with with Trueman (), which found that attractive designs help to make inferences about the product quality.

Another main finding of the analysis is refuting previous research of Berkowitz (?), which found that natural shapes (the Picture design depicts a the picture of a tomato) elicit perceptions to consumers such as they find the product to be more fresher, with better taste and texture. When assessing the key elements related to the perception of 'high quality' of the picture design, the words' beautiful' and 'healthy have scored within the highest coefficients consistently across the different segments, and thus not the expected relationship with the word 'fresh'.

For the 'Draw' design, I have found that for all segments the perception of 'healthy' is consistently within the key coefficients most influencing the perception of 'high quality', and that it is also a key element when predicting the likeliness to buy. As seen in the previous analysis of respondents, just the Draw lovers scored the design to be above average in their perception of 'healthy' and this could explain their higher likeliness to buy this design. Also remarkable is that 'beautiful' is also a high coefficient in correlation to likeliness to buy for all segments but for the Draw lovers. So for this segment the perception of beautiful is influencing their perception of high quality but then doesn't seem to be a factor influencing their likeliness to buy the product. It is worth though mentioning that both Picture lovers and Letters lovers didn't find the draw design significantly more beautiful than then average respondent.

For the 'Picture' design, the coefficient for the attribute 'high quality' is the only one that is consistently present in all segments to the relationship to likeliness to buy. The coefficients for the regression of attributes to the 'high quality' one for this design show also a pattern as all segments have scored healthy and beautiful within the significant range, although again the Letters lover segment with a negative sign for 'beautiful'.

Finally for the Letters design and for all segments it is found that the attributes 'beautiful', ' fresh' and 'different' are consistently within the 'significant' coefficients when assessing likeliness to buy, and that the latter two are consistently across segments with negative signs. For this design the attribute 'fresh' has consistently scored the highest coefficient (healthy being second)in the correlation with the attribute 'high quality'. It can be inferred from this findings that the low perception of 'fresh' from the Draw and Picture lovers for this design is triggering their perception of quality and thus their lesser likeliness to buy it.

Picture lovers perceptions analysis

Assessing the results particularly for each different segment, it is found that for the Picture lovers the attribute 'different' is within the highest coefficients in all regressions against likeliness to buy for all of the three designs, although with a negative sign for the Letters design. For the general relationships of the scores of this segment to the attribute 'high quality' for all three designs, it is found that 'healthy' is present in all of them and 'beautiful' in their perceptions of the Draw and Picture designs.

Specifically observing their attitudes towards specific designs, it is found that the likeliness to buy their preferred product, the one with design 'picture, is linked mostly to the 'high quality' perception and that perception is linked mostly to perceiving the pot 'healthy' and beautiful'.

For the draw design, this segment is linking their relative lower likeliness to buy to 'high quality', healthy with a negative sign, 'different' and 'beautiful'. Their 'high quality' perception for this pot is similar to the patter observed for the Picture design, with 'healthy' and 'beautiful' scoring within the highest and including 'different'.

It can be inferred thus that the fact that this segment didn't show an above average likeliness to buy the draw product is primarily linked to their average perception of 'beautiful', different and 'healthy'.

For the Letters design, High quality' seems to be mostly related to their perceptions of being 'fresh', followed by different' and 'healthy'. Their significant lower likeliness to buy than their choice of preference is linked to their perceptions of 'high quality', 'beautiful' and 'healthy' with a positive sign and 'fresh' and 'different' with a negative sign. It is worth underlying that this segment scored significantly below the average respondent this design in their perceptions of 'beautiful' and 'different'.

Draw lovers perception analysis

The Draw lovers segment seems to care when stating their likeliness to buy to 'different' besides 'high quality, as it is present in their assessment of the three designs and with a negative signs in both the Picture and Letters designs. Their perceptions in 'high quality' have a consistent pattern across designs in their perceptions of 'Healthy' and 'Beautiful', although the latter with a negative sign when assessing the Letters design product.

The perceptions on the draw design with the highest impact on their score of likeliness to buy, besides 'high quality', are 'different' with a negative sign and 'healthy'. Curiously 'different' is also present in the perceptions linked the most to 'high quality', with a positive sign and together with ' healthy' and 'beautiful'. I have no hypothesis to explain this apparent dichotomy, on the one hand and for this segment 'different' helps to contribute to 'high quality' but then is negatively correlated in their likeliness to buy the product.

For the other design which this segment was found to be likelier to buy than the average respondent, the picture design, their likeliness to buy is related mostly to their perceptions on 'fresh', 'beautiful' and 'different', the latter again with a negative sign. 'High quality' seems to be mostly related to their perceptions of 'healthy' and 'beautiful' again as with the Draw design.

Finally for the Letters design, which they showed a significant lesser likeliness to buy than the other two designs, we find that 'fresh' and 'different' with negative signs are within the highest coefficients of the correlation with likeliness to buy, and that all attributes seem to play a, equal role in their perception of 'high quality' excepting for 'beautiful' which is present with a negative sign.

Letters lovers perceptions analysis

For the last identified segment, the Letters lovers, it has to be highlighted that the model used has shown particularly poor correlations when linking attribute perceptions and likeliness to buy (adjusted R2 of zero when assessing the letters and picture designs and 0,332 when assessing the draw design).

For the coefficients more 'significant' towards likeliness to buy it is found that 'fresh' is present within the range with a negative sign for all designs and 'healthy' is also present and with a negative sign for the 'picture' and 'draw' designs (thus not for their preferred design).The attributes with a highest influence in 'high quality' for this segment are found to be 'healthy' and 'fresh' for all designs. It can be observed that this segment scored above average in their perceptions of the Letters design to be 'fresh' and 'healthy' (+27% and +28% respectively). It is difficult thus to make generalizations about this segment, but seems apparently as if healthy and fresh would be attributes that influence negatively their likeliness to buy.

Assessing specifically their perceptions for each design, it is found that for the Draw design the coefficient for 'beautiful' is high in both the correlation with likeliness to buy and with 'high quality' in the same manner as with the perception of 'fresh', although the latter with a relatively high negative coefficient in the regression towards likeliness to buy. Also, For this same design I have found that the 'different' is one of the highest coefficients when assessing which elements are correlated to the perception of 'high quality' and with a negative sign, implying that the more different they perceive it the less high quality they also perceived in this design.

For their preferred design, the Letters one and besides 'high quality', 'different' and 'fresh' have high negative coefficients when linking the perceptions to willingness to buy and 'beautiful' is present with a positive sign. Their perceptions of 'high quality' for this design are mostly related to being' frsh', healthy' and 'beautiful'.

Finally in their assessment of the picture design when linking it to likeliness to buy 'different' is playing an important role together with 'high quality', while also 'healthy' and fresh' but with a negative sign.

Again like with the perception of the draw design for this segment, 'high quality' is linked in a relative similar manner with all other perceptions but playing a key role 'fresh' with the highest coefficient and 'beautiful' with a high negative sign.

7.3.1.4 Other segmentations

In order to validate the research of Bloch. Kotler and Nussbaum regarding that consumers, everything equal, prefer the product with the design they found more attractive, I have segmented the respondents by those that significantly find one type of design more 'beautiful' than the others. For the case of the respondents finding significantly more beautiful the picture design than any other, I have found that there is a significant greater preference to buy it given the same price compared to the designs of draw and letters, for all price levels for the letters design and for the low and medium prices but not greater significantly for the high price for the draw design.

For the group of respondents that found the draw design significantly more beautiful that any other (p value of 0,0018 against picture and 1,19x10-9 against the letters design), it was found that here was not a significantly greater willingness to buy it against the picture design at any given price point, but it was significant against the letters design in all price points.

It can be inferred by this findings that finding it 'beautiful' is not the same as finding it 'attractive' as they stated, and that there are other factors that are influencing the decision besides price. Altough it has been proven for all segments that 'beautiful' is consistently linked to the perception of 'high quality', other perceptions or influencers are playing a role in the behavioral process.

Finally it was not possible to find a significant sample of respondents that considered the product with design letters to be significantly more 'beautiful' than any other.

7.3.2 Estimating the Conjoint Model Using Dummy Variable Regression

The total of the 49 respondents rated each of the 9 product prototypes, the total of the combinations possible between three different design types ('Draw', 'Picture' and 'Letters') and three different set of prices (low price '€1,99', medium price '€2,49', high price '€2,99'), and the answers were collected.

There were seven possible ratings for each prototype where the respondents stated their likeliness to buy such product: Very Unlikely, Unlikely, Somewhat Unlikely, Undecided, Somewhat Likely, Likely and Very Likely. In order to codify the responses a number was assigned to each of them from 1 (very unlikely) to 7 (very likely).

In order to perform a multiple variable regression between the different product attributes and their likeliness to buy, the different prototypes were codified accordingly with dummy variables:

Dummy variable design: if design picture '1', if design draw '0', if design 'letters' '-1'

Dummy variable price: if price low '1', if price medium '0', if price high '-1'

Finally the average responses for all respondents for each prototype were calculated.

Note that there are 9 observations representing the 9 different prototypes and that the regression was statistically significant with a p value of 0,0001.

The utility or parts-worth of each attribute, design and price, can be calculated by multiplying the coefficient of the regression for each attribute with its representing dummy variable. In conjoint modeling, it is assumed as mentioned previously that the consumer utility can be calculated by adding up the value of the utility of each of the levels of the attributes of the chosen product. When adding up to each combination the intercept value, an estimated likeliness to buy for that product can be estimated.

Also, as each of the attributes has been proven to be statistically independent from each other, a relative importance of each attribute can be simply calculated by calculating the ratio of the coefficient of each attribute to the sum of both coefficients

Thus,

Importance of Design = 0,9014/(0,9014+0,5170)=64%

Importance of Price= 36%

So for the average respondent the design is significantly more important than price when deciding upon their likeliness to buy the presented soup products.

Also another type of dummy codification was performed in order to find the relative importance of each type of design and price compared to each other.

Dummy picture: If picture '1', otherwise '0'

Dummy draw: if draw '1', otherwise '0'

Dummy medium Price: If price €2,49 '1', otherwise '0'

Dummy Low price If Price: €1,99 '1', otherwise 0

Note that with this codification there is no need to codify for the Letters design and High price levels, as it will be considered when the rest of the levels for each attribute are both codified at 0 level (i.e. Letters Design shall be implicitly considered when dummy picture and dummy draw are both '0').

The analysis of the linear regression shows, by studying the coefficients, that for the average respondent the design picture is to most related to their willingness to buy the products, followed by the draw design with roughly half the utility. Again in line with economic theory, low prices are preferred that high prices (low priced preferred 2,2 times more that medium price)

Note that if adding up the coefficients for 'picture' and 'draw' and calculating the ratio against the sum of all coefficients (excepting the intercept coefficient) we obtain the same ratio of importance of the design attribute versus the price attribute as with the previous regression (64% versus 36%).

As with the previous analysis, I have also performed the same regression of product prototypes against average scores for each of the segments identified: Picture lovers, Draw lovers and Letters lovers, all with statistical significant results, in order to calculate the relative importance of the two attributes for each subgroup.

For the picture lovers we find that the importance of design is significantly higher compared to that of the average respondents, becoming 81% for design and 19% for price.

Again with the second dummy codification, it can be seen that for this segment the design of the picture if preferred over the draw design by nearly three times. The low price is preferred over the medium price by more than three times and there are relatively small differences between the medium and high price. Recalling the average preferences for the price levels for all designs mentioned in section 7.3.1.1 (low price over medium by 47% and medium over high by 53%) one of the big advantages of the conjoint model using multiple variable regression is highlighted: preference for one level is not the same as relative utility. Even if this segment had show a preference for lower prices, they are willing to trade them off easily for the preferred design.

Also it can be seen from the regression results that there is no significance in the correlation with the medium price, while in the analysis of their scoring of products within two consecutive price points they showed a significance preference for the lower priced products (significantly preferred the low price over the medium and the medium to the high. The explanation to these findings can be related to the small relative utility that price is offering to this segment compared to that of design.

For the Draw lovers it has been found a similar trend, becoming the importance of design 72% and price 28%. Also as show on the previous analysis of data through descriptive statistics, their preference for the draw design is no significantly higher than for the picture design, as shown on the respective utility curves. Also there seems to be a more proportional attitude towards prices, as utility decreases in the same proportion as prices increases.

Finally for the letters lovers we find a trend not observed before, as price seems to be significantly more important than the design, 71% versus 29%. The lower price is preferred twice as much as medium price

8 FINDINGS

        Complementarities and contradictions with previous knowledge

        Limitations and weaknesses of the research

        The sample chosen

        Proposal of potential ways to overcome identified limitations

9 References

(1) http://www.idsa.org/absolutenm/templates/?a=89

(2) P. Design and corporate strategy Blaich Book

(3) Industrial design and businessman HBR dreyfuss

(4) Product design and development Ulrich book

(5) Winning at new products Cooper

(6) competing through design Trueman

(7) Design and marketing of new products. Urban

(8) Discovering New Product Opportunities with Problem Inventory Analysis

(9) Product shape as a design innovation strategy Berkowitz

(10) Seeking the Ideal form, Bloch

(11)Seeing things: consumer response to the visual domain in product design. Crilly

(12) Motivation and personality Maslow

(13) The commercial impacts of investment in design Roy

(14) the impact of industrial design effectiveness on corporate financial performance Hertenstein,

(15) Euromonitor International : Country Sector Briefing June 2009

(16) http://www.packaging-gateway.com/projects/yes-please/

(17) Redesigning product lines with conjoint analysis; How Sunbeam does it. Page

(18) On the design of choice experiments involving multifactor alternatives (Green) from Page