FACTORS AFFECTING CUSTOMER BRAND PREFERENCE a CASE OF BEER INDUSTRY IN HANOI

Do beer characteristic, branding, beer types, situation appropriatenessstatements, packaging, social media and country of origin have relationships withconsumer brand preference in Hanoi

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION 3

1.1 The rationale of the research 3

1.2 Objectives of the research 4

1.3 Scope of the research 5

1.4 Research methodology 5

1.5 Structure of the research 6

CHAPTER 2: LITERATURE REVIEW ON CUSTOMER BRAND PREFERENCE IN BEER INDUSTRY 8

2.1 Brand 8

2.1 Customer brand preference in beer industry 9

2.2 Factors affecting Customer brand preference in beer industry 12

CHAPTER 3: RESEARCH METHODOLOGY 19

3.1 Research process 19

3.2 Research method 20

3.2.1 Qualitative research 20

3.2.2 Quantitative research 21

3.3 Research model 24

3.3.1 Variables, hypotheses and equation 24

3.3.2 Research model 26

3.4 Data analysis 27

3.4.1 Population and Samples 27

3.4.2 Data collection 27

3.4.3 Data analysis procedure 28

CHAPTER 4: RESEARCH FINDINGS 29

4.1 An introduction about beer industry in Vietnam 29

4.2 Demographics 32

4.3 Descriptive statistics 36

4.4 Scale test on factors affecting beer preference 39

4.3 Correlation matrix 41

4.4 Exploratory factors analysis 43

4.5 Confirmatory factors analysis 48

4.6 Structural equation model 51

4.7 Regression Results 53

CHAPTER 5: RECOMMENDATIONS AND CONCLUSIONS 56

5.1 Conclusions and research limitations 56

5.2 Recommendations 58

5.2.1 Recommendations based on beer characteristics 58

5.2.2 Recommendations based on branding 60

5.2.3 Recommendations based on situation appropriateness 62

REFERENCES 63

APPENDIX A v

LIST OF FIGURES LIST OF TABLES

Table 3.1: Acronym for variables22

CHAPTER 1: INTRODUCTION1.1 The rationale of the research

A critical component of marketing in any product is consumer behavior As aresult, this paper aims to answer what makes a consumer buy a product, especiallywhere other similar products and brands are available The study of consumerbehavior in purchasing beer would allow marketing managers to come up with aninsight on what factors governing a consumer's decision in order to put forwardstrategies and predictions that align with customer brand preference

Beer-making and its resultant beverage have been present for multitudinouscenturies alongside the development of human history However to Vietnamesepeople, beers as a topic only caught on since its first introduction in the 19thcentury, by the French colony Ever since, it has grown on the general populace,gradually replacing traditional home-brewed liquors as a national favorite, throughits lighter alcohol level, more refined taste and affordable price [ CITATION FPT18 \l

1033 ].

Even for today, the beer industry is still ceaselessly expanding in Vietnam inboth growth rate and production capacity, rivaling other big nations such asEngland, Japan and Poland [ CITATION FPT18 \l 1033 ] Consequently, there is anincreasing number of both local brand breweries and international brands emergingfrom within the market competing for market-share Therefore it was one primarygoal of this study to investigate beer brand preference pattern in Hanoi - acosmopolitan city as a focus of this study A set of identified 7 intrinsic andextrinsic [ CITATION Sunkamol16 \l 1033 ] factors will be examined for their potentialinfluence on customers' behavior

Myriad research has already been published internationally on the subject ofbrand preference in the beer market Brand preference in these studies werediscovered to be affected by both intrinsic factors [ CITATION Aqu15 \l 1033 ], [ CITATIONSunkamol16 \l 1033 ] and extrinsic factors [ CITATION Chr14 \l 1033 ] Nevertheless, this

topic of research was still few and far between in Vietnamese literature Hence, thispaper aims to fill in the gap; it will put brand preference into a conceptualframework in Hanoi.

1.2 Objectives of the research

This research aims to investigate factors affecting customer brand preference

in Hanoi beer industry

In detail, it resolves the 3 main objectives, as state below:

First of all, it sets to identify factors that can positively or negatively affectVietnamese customer's beer brand preference

Next, using an empirical framework, it aims to unravel the underlyingreasons and to explain the correlation between factors in the model, all has a basis

on the context of the overview of beer industry in Vietnam

Finally, the findings may be of use to marketers and policy makers to betterunderstand a customer's mindset and create corresponding strategies

In order to give light to the objectives above, this research will also aim toanswer some research questions:

1 Do beer characteristic, branding, beer types, situation appropriatenessstatements, packaging, social media and country of origin have relationships withconsumer brand preference in Hanoi beer industry?

2 Do beer characteristic, branding, beer types, situation appropriatenessstatements, packaging, social media and country of origin a with consumer brandpreference in Hanoi beer industry?

1.3 Scope of the research

The study is conducted in the area of Hanoi, through non-randomconvenience sampling method Data collected by questionnaires from 300respondents aged over 18 who had experienced drinking before will be analyzed toseek the correlation between independent variables: Beer characteristic, Branding,

Beer types, Situation appropriateness statements, Packaging, Social media, Country

of origin and the latent dependent variable: Brand preference

The study was conducted from March, 5th 2019 to May 30th 2019 With thefirst month (March) spent on reviewing the literatures and drafting research outline

as well as the questionnaires; from April 1st to May 5th dedicated to surveycollection and the remaining time spent on analysis and developing a conclusion

1.4 Research methodology

Most scientific research is done through either qualitative or quantitativemethod With qualitative research - a method primarily employed for exploratorypurposes, does not involve numbers or numerical data It often involves words andlanguages to express observations gathered from a variety of methods such asinterviews, focus group, secondary data or plain observations

Quantitative research on the other hand can be more easily identified anddefined The data produced are always numerical, and they are analyzed usingmathematical and statistical methods [ CITATION Mar09 \l 1033 ] The data forquantitative research are primary data through surveys, observations and othersecondary data These data will be further analyzed and visualized for correlationand statistical significance through quantitative programs

For the sake of this research, a mixed method will be utilized to both thecollection and analysis of data Qualitative method will be used in the surveydevelopment phase of the study when latent variables are encoded with representingquestions While quantitative method will be utilized in model developmentprocess, when IBM SPSS and IBM AMOS are used to analyze the resultingsurveys

1.5 Structure of the research

This research will be divided into 5 distinct chapters, each with a singularpurpose as detailed:

Chapter 1 - INTRODUCTION: We will discuss the rationale of the research,what this thesis aims to accomplish, its respective area of scope and also lightlytouch on relevant researches on the same field.

Chapter 2 - LITERATURE REVIEW ON CUSTOMER BRANDPREFERENCE IN BEER INDUSTRY: This chapter aims to cover all the literaturereview on all factors, such as customer brand preference, beer characteristic,branding, beer types, situation appropriateness statements, packaging, social mediaand country of origin with their respective previous findings and how these willcome together to build a feasible theory model

Chapter 3 - REASEARCH METHODOLOGY: This chapter is to buildresearch methodology and data analysis The study will discuss the data collectingprocedures, questionnaires, how data would soon be analyzed and finally come upwith a conclusive model to test out all my previous hypotheses

Chapter 4 - RESEARCH FINDINGS: This chapter aims to illustrate researchfindings from said questionnaires, and further analysis on both individual factorsand the model's regression validation, using Confirmatory Factor Analysis andStructural Equation Modeling These results are then compared to a set of indicesfor model fit and used to either approve or reject the hypotheses

Chapter 5 - RECOMMENDATIONS AND CONCLUSIONS: This finalchapter sums up all previous findings, and give recommendations for future furtherresearch, strategies for marketers and students who are interested in bettering thebeer industry in Vietnam

CHAPTER 2: LITERATURE REVIEW ON CUSTOMER

BRAND PREFERENCE IN BEER INDUSTRY

2.1 Brand

Brand is considered firm asset which can contribute identity and character,guide consumers' product choices and form the relations among consumers Otherthan that, brand has several benefits for firms, consumers and society [ CITATION Bah15 \l 1033 ]

In terms of consumers, brand is a quality indicator and creates awareness forproducts A highly sought after brand is great indicator that it is a brand of goodquality and reputation

In terms of firms, brand provides customer loyalty, consistent sales amountand a high profit margin As a result of consumer and firm benefits, brand plays akey role on social development, where both the incentives of firms and customersare satisfied in the act of exchanging goods, services and cash In intensecompetitive conditions, firms need to create strong brands in order to survive andgain a competitive advantage

However, the exact definition of a "brand" is still in debates Some wouldconsider it the personality of a business [ CITATION All16 \l 1033 ], where it must beable to illustrate several attributes: name, design, positioning, brand promise, targetmarket and awareness Others define that brand is the perception of customers aboutsaid companies, organizations or individuals [ CITATION Bri18 \l 1033 ] Since it is anintangible concept, it must be defined in the place where it shows Brand lives inpeople's minds, which includes all employees, investors, consumers and the media

A brand’s personality is the reason it is identifiable to its loyal customers and thebasis for the highly personal relationships they form with it

To sum up, brand is the sum total of how people perceive a particularorganization This is influenced by the brand compass - of where the direction of thebrand is going through the use of a clear vision statement, mission statement,purpose statement, values proposition and strategic objectives to attain these goals

In this regards, values proposition or core values are especially important in guidinghow the brand will engage and leave its mark with the world outside, and in turns,

be perceived favorably or unfavorably

2.1 Customer brand preference in beer industry

Amadi and Sunday [CITATION Chr14 \l 1033 ] definition of brand preferencebeing a measure of brand loyalty where a consumer would prefer a one brand inpresence of competing others, but will switch to substitute products should thatbrand is unavailable

The notion of preference has been considered by different disciplines, such

as economist, psychologists and sociologists Economists [ CITATION Sam78 \l 1033 ]state that preferences are exogenous stable, known with adequate precision and arerevealed through choice behavior, but many other refused to believe on thisexclusive exogenous characteristic In marketing, the concept of preference meansthe desirability or choice among alternatives This concept of Brand Preference isoften denoted differently by different academic authors But in summary, first brandpreference is created from consumers' differentiation and comparisons betweenvarious alternatives of brands considered by them Second, the resulting preferencedisplays several biases in certain brands in evaluation of alternatives And finally,the attendant predisposition of customers is reflected by affective, cognitive andbehavioral responses[ CITATION Reh13 \l 1033 ]

A successful brand would know how to take advantage of groups ofcustomers who have positive attitude towards their products or brand and inspirethem to prefer its brand over the competitors However, as a good attitude can onlylast so much, that is why obtaining brand loyalty is a highly desirable next step forfirms after brand preference [ CITATION Reb02 \l 1033 ] in that they retain existingcustomers instead of increase costs to find new ones, deter customers fromswitching brands, encourage related purchases and cause customers to be moretolerant in event of product failures Unfortunately, further research on brandloyalty will be needed as this concept is not covered by this research

As have established, customer loyalty towards a brand is reflected throughbrand preference It can be difficult to accurately measure brand preference, but thenumber of repurchases and referral programs can be a good starting point Plans todevelop brand preference should be regarded utmost importance Nevertheless,building brand loyalty was only the stepping stones to gain more advantages onbrand equity [ CITATION Hai08 \l 1033 ] In a market of similar producers with similarproducts, competitive advantages must be gained through the use of a higher brandequity - the manifestation of confidence customers have on particular brands.

In order to better understand brand preference more precisely, it is important

to get the distinction between brand preference and other relating constructs, such asbrand loyalty (as have mentioned), brand choice, brand attachment and brandawareness

Brand preference and Brand loyalty - in the dictionary brand preferencemeans “a measure of brand loyalty in which a consumer will choose a particularbrand presence of competing brands, but will accept substitutes if that brand is notavailable." Whereas, brand loyalty means “extent of the faithfulness of consumers

to a particular brand, expressed through their repeat purchases, irrespective of themarketing pressure generated by the competing brands [ CITATION Bus19 \l 1033 ] Astaken from dictionary definition, the main theme is that brand preference is related

to brand loyalty However, brand loyalty is depicted more consistent by long-termrepeat purchasing behavior, consistent with [ CITATION Reb02 \l 1033 ] research

About brand preference and brand choice, brand choice is concerned with theselection and consumption of the brand[ CITATION Bet98 \l 1033 ] Brand preferencecan be viewed as a motivator of brand choice Consumer choices are based on well-defined preferences through which consumers can determine the set of alternativesfrom which they will make their choices Marketing managers are more interested

in brand preference than brand choice to signal repeated purchases, since consumerpreferences tend to be constant across the different contexts, rather than choice-limited to a specific context

Finally, the distinction between brand preference and brand awareness.About brand awareness, it is defined as the primary knowledge of the brand incustomers' mind It reflects the biased position toward a certain brand developedafter exposure to the brand Therefore it can be said that brand awareness is thestepping where customers from which, can develop their own sets of brandpreference.

Previous research also discussed how cognitive consistency theory couldhave an effect on an object's preference According to Rosenberg attitude model[ CITATION Ros56 \l 1033 ] customer preference for brands is derived from intrinsicbrand benefits, as well as some degree of satisfaction with brand value This essence

is captured in the following formula:

N

Ao = ∑ (PIi)(VIi)

i=1Ao: is the consumer preference toward an object

PI: is the perceived instrumentally of the object; beliefs about the ability ofthe object to offer value i

VI: is the value importance; the degree of satisfaction with the value iprovided

N: the number of values offered

Another well-known one is from Fishbein [ CITATION Fis65 \l 1033 ] which isalso commonly represented in marketing studies This model originated frombehavior theory, as it made an effort in explaining the relationship between attitudeand behavior by investigating their impact As a result, Fishbein created an equationwhich measured brand preference on an algebraical scale by brand weightedattributes and belief associated with an object:

N

Ai = ∑ (Wi) (Bib)

i=1Ai: the preference toward a particular brand

Wi: the weight/importance of attribute i.

Bib: the belief toward attribute i for brand b

N: the number of attributes important in selecting brand b

From these study above, it can be concluded that customer brand preference

is a measure of brand loyalty where a consumer would prefer a one brand inpresence of competing others It is influenced by a combination of intrinsic (innatecharacteristics) and extrinsic factors (social, psychological, economical cues) Forthe next part these factors will be discussed in details

2.2 Factors affecting Customer brand preference in beer industry

Beer characteristics

Beer's inherent characteristics can be among the primary reasons for whybeer purchases happen in the first place Therefore beer attributes such as aroma,foam, perceived quality and carbonation do have an impact on the choice of beertypes [ CITATION Aqu15 \l 1033 ]

However characteristics alone cannot complete the full story As [ CITATION Car12 \l 1033 ] stated past customers experience with brand stimuli have a largerimpact on the choice and preference of beer; even surpassing innate characteristics

of beers

But taste is still a major factor on which customers base their repeatedpurchase decisions As the results from[CITATION Gne14 \l 1033 ], customers are morelikely to return to a beer that tastes good and even more willing to pay higher for it.Beer can be made of four main sources: malt, yeast, water, and hops These enabledifferent brewers to create a wide variety of beers These kinds of ingredientsfacilitate differentiations in terms of quality

This theory is further confirmed through a beer blind taste testing conducted

by [ CITATION Car12 \l 1033 ] His result illustrates how sensorial cues, such as taste,texture, color, appearance and flavor can have a lasting impact in the mind (negative

or positive) of customers The method of blind taste has been used widely inscientific research, as a mean to eliminate potential mental biases and past

preference, by focusing the participants solely on sensorial cues of the given testsubject.

In order to evaluate this factor more accurately, we need to understandprecisely what product quality means As [ CITATION Sta13 \l 1033 ] have discussed inQuality Management for Organizational Excellence, Product quality could bereferred either as the standard performance of the product or expected performance

of it The difference between expectations and the product performance would have

a significant impact on the perception of customers as well as on brand preference.Additionally, quality could be individually or collectively judged on the 5 stagesmodels of a customer decision making process: Need recognition, informationsearch, and evaluation of alternatives, purchase and post-purchasebehavior[ CITATION Cha14 \l 1033 ]

To sum up: aroma, foam, perceived quality, appearance of texture and color,these innate characteristics are found to be positively influencing brand preference

Branding

Branding through its 3 main purposes: product identification, repeat salesand new-product sale can have a significant influence on the choice ofcustom[ CITATION Cha14 \l 1033 ], because customers will often associate brandingwith quality Since the success of customers products largely depends in part on themarket's ability to distinguish one product from another, branding would be a usefultool in helping the firms gain an edge in the competition In short, a firm canmanipulate its own name, term, symbol, design or combination that can helpidentify its products and distinguish it from the rest of the market

This concept still holds true in the beer market Since [ CITATION Sie13 \l 1033 ]found that the branding of alcohol market was among the most influential factors oncustomer purchasing decision: there is usually a small concentration of alcoholbrands which dominate the total consumption In this competitive market, whenproduction quality ceases to be the differentiation factor, and product quality isclose to maximum, it is customer service, brand personality, in short, customerrecognition and identification that is key to success

The term brand equity also plays a vital role in the equation A brand whichhas high awareness, perceived quality and brand loyalty among customers havehigh brand equity[ CITATION Cha14 \l 1033 ] An organization's brand equity can be asindispensable as the goods or services it produces for that it makes people confident

in that the organization is trustworthy To develop a corporate brand is importantbecause a positive brand image will give consumers, and other interestedstakeholders, confidence about the full range of products and activities associatedwith a particular company[ CITATION Abd15 \l 1033 ]

Beer types

Beer is among the most consumed alcohol beverage among Vietnamesecitizenry According to finding by [ CITATION Ngọ18 \l 1033 ], Beer is preferred by bothmen and women, with a whopping 68.7% preference of the subject sample

How the beers are made can also be paramount in deciding customerpreference of beer types In fact, just the mentioning of the "secret ingredients" intheir packaging can negatively or positively influence a customer's perception of thebeer [ CITATION Leo06 \l 1033 ] This is due to arising expectations of the products thecustomer might give Generally beers are brewed from 4 main ways [ CITATION Sunkamol16 \l 1033 ]: wheat, malt, fruit, or craft beer

However, not all of these types are widely available in the Vietnamese beermarket In fact, fruit beer and malt is an uncommon ingredient since the tropicalclimate of Vietnam makes it difficult to brew with those However, since malt beer

is the ubiquitous staple drinks for Vietnamese men and women[ CITATION Ngu17 \l

1033 ], Vietnamese producers have resulted in numerous "filler" ingredients such asfrom common crops, corns, potatoes, etc besides malt, to cut down cost and makebeer more widely available as a national drinking beverage Beers consumptionusually concentrate on a small number of major producers: Sabeco (with 40%marketshare), Heineken (with 25% marketshare) and Habeco (18% marketshare)[ CITATION Kha17 \l 1033 ]

The term "draft beer" here can sometimes be used interchangeably with

"craft beer" But there is one major distinction to tell the two apart Draft beer - or

Bia Hơi undergoes both a short fermentation process and a pasteurization processwith rapid heating and cooling to kill off harmful bacteria; so they tend to have ashort shelf life Craft beer - or Bia Tươi while goes through the same process butusually brewed independently by restaurants or mini-breweries, is served as they arecreated without the pasteurization [CITATION Sap \l 1033 ]

Situation appropriateness

As beer is not indispensable products in every household's meals, itsconsumption often take place in special occasions: such as weddings, alone times,hang out with friends… Therefore its preference of one beverage over another isfound to be dependent on the contexts, or the appropriateness of the situation inwhich it happens Furthermore, beers are often found to be one of the manyelements to be consumed in these moments Which means that the food that oftenaccompany the meals can greatly influence beer choice among participants of theevents [ CITATION Car16 \l 1033 ]

People usually drink beers at the main 2 places: home consumption orrestaurant-context consumption At home, people tend to pay more attention to thebeers themselves - as these are the catalyst for people to share moments together.While at the restaurant, beer choices are, most of them time, affected by the foodwhich are present [ CITATION Car16 \l 1033 ]

In order to have clearer command of situation appropriateness, [ CITATIONCar161 \l 1033 ]'s research have discovered the most common situations among whichparticipants often found themselves indulging in beers: for casual social & sportingevents and for special occasions Casual social events often calls for beers “as athirst quencher”, for use “At a sport event”, “When I want to relax”, when I’m “Athome”, for “casual dining” and “To drink alone” while beers for special occasionscall for qualities that can help them “To impress someone”, “To taste with friendsinterested in beer” and “When I want something different”

Packaging

Packaging's design plays several roles to advocate for the products andbrand Therefore it needs to be creative and innovative to compete with the products

of other major producers [ CITATION Kli06 \l 1033 ] in Packaging Design: SuccessfulProduct Branding from Concept to Shelf has stated packaging design is used tocontain, protect, transport, dispense, store, identify and distinguish a product in themarketplace Therefore, packaging aids the ultimate marketing objectives incommunicating both of a product's functionality and personality

Packaging has also been shown to influence customers' perception andbuying behavior [ CITATION Cha14 \l 1033 ] through the use of designs, colors, shapesand materials Therefore, by enhancing the experience of customers with plentyinformation on product's function and personality, packaging is among theparamount factors which can affect the brand preference of customers

Another major advantage of good packaging design is that: it helps reducethe cost of advertising It sometimes even in so far as can have more impact on acustomer's decisions than that of a brand image As customers often place highervalues on convenient, environmentally friendly, organic, patriotism invoking,authentic & prestige products [ CITATION Cha14 \l 1033 ]

Social media

The book definition of the social media, in MKTG7 by Lamb, Hair andMcDaniel, is that "it is any tool or service that uses the internet to facilitateconversations" Because it is in every customers' needs to exchange information,collaborate and engage in conversations, it is the job of the marketers to find whichconversations profitable and ways to enter them

By closely monitoring for customer's insights, building a steady relationshipsbetween brands and their loyal followers, promoting information on products andservices, building reputation and engaging in dialogue with dissatisfied customers,the brand can enhance customers experience with their beers, especially to the tech-savvy millennial who are strong purchasers

Purchasing frequency has also been shown to be influenced by the socialmedia marketing Since the social media is a powerful tool to convey information,consumers today rely heavily on it to obtain products or brands related informationwhich influences their purchase intention [ CITATION Pre16 \l 1033 ].

Furthermore, social people tend to hold their influencers' opinions in highregards, especially those who they identify with or assume to be the experts of theparticular interest [ CITATION Gab01 \l 1033 ] As a result, social groups could possiblyhave a significant impact on brand preference of beer brands

Country of Origin

The country of origin is in fact one simple criteria for helping customerschoosing their beer brands, as they can draw information from unfamiliar brandsfrom their country of origin Country of origin is considered a factor here becausecustomers can derive from the past experiences and perception of the country’sproduction and strength of market to assume qualities in a particular product Thistheory is supported by[ CITATION Anj97 \l 1033 ]'s work illustrating a strong correlationbetween the country of origin factor and brand preference, noticeably amongunfamiliar brands than familiar ones Because beer is considered a low-investmentproduct, this extrinsic cue is more likely to influence customer's choice of favoriteproducts However, given more information on other values of the products,customers tend to rely more on intrinsic attributes (taste, etc) to evaluate a beerbrand

Secondly, consumers are inclined to process information about countries andbrands simultaneously in order to come up with a concluding evaluation of theproducts [ CITATION Yam15 \l 1033 ] But when their perceived country of origin andbrand are inconsistent, consumers would try to look for more information or at otheralternatives, with a focus look on operational performance instead of country origin

Since the market of nowadays are highly competitive, differentiation andvalue-added product strategies are of primal import With country of origin orcountry image having both a direct and an indirect effect on product preference, thekey for beer producers here is to create unique identity for their product from

country origin factor by communicating sufficient information on their source ofproduction, raw materials and components [ CITATION Koe03 \l 1033 ].

As a closing note on country of origin, customers can have differentinteractions between global and local beer brands Global beer brand tends to gainmore customer loyalty than local ones but they are not necessarily perceived to bebetter [ CITATION Cri15 \l 1033 ]

CHAPTER 3: RESEARCH METHODOLOGY

This chapter demonstrates applied methods studying factors which positivelyaffect beer brand preference of customers in Hanoi The chosen approach to thisresearch is deductive approach, which means "a theoretical position is developedprior to the collection of data", which will subsequently be tested using the said data(Mark, Phillip & Adrian, 2009)

3.1 Research process

The research process for the topic comprises of 5 basics steps which were done in order one after another

Figure 3.1: Research process

Source: Author compilation

The first step is identifying research topics, where the focus of futureresearch will be formed This is done by coming up with rationale, objectives,methodology as well as the scope of the research to facilitate the second step -Literature review In this steps a multitude of previous literature was scannedthrough for relevant information and conclusions The methodology of similarresearch is also taken account of The third step is Qualitative research, to build,modify the scale to match with research context and propose an official model toconduct research This will be combined with the next step Quantitative research, to

Identify research topics

Literature review

Qualitative research Quantitative research Recommendations and conclusions

describe statistics, reliability analysis using Cronbach's Alpha scale, EFA, CFA,Pearson's correlation matrix and multiple regression analysis in order to testhypotheses Finally, meaningful recommendations and conclusions will be drawn toprovide marketers in the beer industry with up to date insight.

3.2 Research method

As the focus of this research is to seek factors that positively influence brandpreference, quantitative approach will undoubtedly be the chosen method forcollecting and analyzing primary data Quantitative is dominantly used as asynonym for every data collection technique (like a questionnaire) and data analysisprocess (using graphs or statistics) which can generate numerical data[ CITATIONMar09 \l 1033 ].

3.2.1 Qualitative research

To begin with, questionnaires were developed on the foundation ofdimensions laid by previous research Observed variables were selected to measureBeer Characteristics, Beer Types, Branding, Situation Appropriateness, Packaging,Social Media, Country of Origin and Brand Preference The next step is to translatethe mentioned observed variables into the questionnaires In order to quantifyqualitative concepts, Linkert 5-point scale was used to best represent the opinions ofsurvey participants After the draft questionnaire was completed, a pilot test wassent out to 10 friends of the author to check for misleading and incorrect phrasing ofthe questions Comments from the previous pilot test will be based upon to developthe final questionnaire

On the content of the questionnaire itself, it is divided into 2 main sections.The first section will be on demographics information, where question on personalinformation of the participants will be asked, such as gender, age, marital status,level of education, monthly income brackets, consumption frequency of beer andfavorite beer selection This part serves 2 purposes: To give an outlineunderstanding of the research sample to confirm any potential biases and toeliminate responses which do not belong to the target sample (those who have neverdrank beer, for example, will be eliminated from analysis) The next section will

investigate the customers' opinions on why they prefer one brand of beer to another.There are total 40 multiple choice Questions using 5 point Linkert scale, which will

be used to explore more on the 8 predetermined affecting variables

3.2.2 Quantitative research

To start with, data distribution will be analyzed to describe the properties ofthe survey sample groups such as gender, age group, education level, marital status,income, frequency of beer consumption and favorite beer brands

After that, descriptive statistical analysis is used for observed variables of thescale according to the research model The Cronbach's Alpha will be next to comeinto play in order to analyze the reliability of variables The scale of the potentialfactors on consumer brand choice of beer industry in Hanoi consists of 8 potentialvariable groups, expressed through the criteria with Likert scale 5 points - 2 poles:(1) Strongly disagree and (5) Strongly agree This Scale test is designed for thepurpose of finding the question items and decide whether that need to be retained orremoved from the questionnaires [ CITATION Hoa08 \l 1033 ]; in other words, it is foreliminating inappropriate variables and scales that do not meet requirements

Several researchers in the academia agreed that when Cronbach’s Alpha isgreater than 0.8 and less than 1, the scale test is at its best From 0.7 to nearly 0.8 isserves as a usable scale There is also suggestion that a Cronbach's alpha greaterthan 0.6 is also acceptable in case that the concept of measurement is of novelty ornew to respondents in the research context (Hoang and Chu, 2008) Cronbach'salpha coefficients are used before eliminating the previous inconsistent variables,then item-total correlation which is less than 0.3 would be disqualified and thestandard for choosing scale when it has alpha of 0.6 or higher (Hoang and Chu,2008) For this study, the author proposed that the level of Cronbach’s alpha of 0.6

or more is acceptable

It is a proven standard procedure for factor analysis whose main objective is

to discover the basic relationships among calculated variable and a group ofvariables Researchers usually employ this while expanding a scale which is a set ofqueries used to evaluate certain research topics For EFA, the term ‘factors’ is used

for the continuous latent variables while the term ‘factor indicators’ used forindicating observed variables Factor indicators can vary to a great extent, becausethey can be censored, continuous, counts, ordinal (ordered categorical), binary, ormixture of all variables EFA considers that any particular indicator or measuredvariable can be linked with any common factor or unique factor

Exploratory Factor Analysis (EFA) is used to shorten the set of observedvariables with close relationships without reducing the amount of initial variables.Factor analysis is used to test the convergence of conceptual variables withConvergence validity while measuring discriminant value to help ensure differenceswithout correlation among variables, which is measured by Discriminant validity.Exploratory factor analysis should be based on specific and reliable standards

The research topic uses EFA method, extracting the factor PrincipalComponents with Varimax perpendicular rotation and stopping point whenextracting Eigenvalue factors of larger than 1 When conducting analysis, observedvariables that do not meet requirements will be disqualified The observed variableswill be selected based on the coefficient of KMO (Keyser - Meyer - Olkin ofSampling Adequacy), which is used to evaluate the appropriateness of the factoranalysis method with the collected data and must satisfy the condition of 0.5 ≤KMO ≤ 1, the Eigenvalue greater than 1 to prove an above 50% total varianceextracted, the significance level of Bartlett’s test (Bartlett Test of Sphericity) below0.5 to prove its statistical significance and a factor loading above 0.55 in accordancewith sample size (details in Chapter 4)

The Pearson correlation coefficient's main purpose is to test the degree ofstrictness of each linear relationship between two quantitative variables If r > 0, itshows a positive correlation, otherwise r < 0 indicates a negative correlation, r = 0indicates that two variables do not have a linear relationship The closer to 1 thevalue of | r | is, the stronger the relationship between the two variables, and thecloser to 0 the value of | r | is, the weaker the relationship between the two variables

On the other hand, if the independent variables are highly correlated with each

other, it is also a signal for identification of multi-collinearity in the linearregression model under consideration.

The regression model is used to determine the influence of some potentialfactors (independent variables) on consumer brand preference of beer consumers inHanoi (dependent variable) The method used in this research is the ordinary leastsquares method The coefficient of determination (adjusted R2) is statisticalmeasurement for the closeness of the data are fitted to the regression line, the F isfor testing the hypothesis of the equality of two population variances To summarizethe model's concern with brand preference, the following equation was drawn toexpress the relationship between dependent and independent variables:

BRAPRE = β1.PACKAG + β2.CHARAC + β3.BRANDI + β4.SITAWA +

β5.BEETYP + β6.SOCMED + β7.COUORI

Finally, all qualified variables will be passed to Confirmatory FactorsAnalysis to check for the fitness of research model and data collected for theobserved variables and Structural Equation Modeling to check for potentialrelationship between variables in accordance with the hypotheses The StructuralEquation Modeling or popularly known as SEM is a statistical analysis techniquesdeveloped for analyzing the inter-relationships among multiple variables in a model.The inter-relationships among variables could be expressed in a series of single andmultiple regression equations [ CITATION ZAw12 \l 1033 ]

CFA is often used to determine and confirm the factorial structure of analready developed measuring instrument in an application among a targetpopulation If the hypothesized CFA model fits data, we confirm the factorialstructure is valid for the population [ CITATION Jic12 \l 1033 ]

Compared with EFA, the advantages of CFA include: firstly, all factors inEFA are either uncorrelated (orthogonal) or correlated (oblique) In CFA,relationships among factors can be flexibly specified on a theoretical basis or based

on empirical findings Secondly, observed indicators/items in EFA are loaded ontoall the factors; while observed indicators/items in CFA are only loaded onto factorsthat they are hypothesized to measure However, an indicator may also be loaded

onto one or more factors in a CFA based on a theoretical concern As a result, aCFA model is not only theoretically more meaningful, but also is moreparsimonious because the factor loadings of indicators to the irrelevant factors areall fixed, a priori, at 0 in a CFA model, thus substantially reducing the number ofparameters to estimate Thirdly, measurement errors are not allowed to be correlated

in EFA, yet this is not the case in CFA However, appropriate specifications of errorcorrelations in CFA can be used to test method effects Fourthly, unlike thetraditional EFA, CFA can be simultaneously conducted in multiple groups so thatmeasurement in variance across groups can be tested Finally, covariates can bereadily included to predict the factors, thus expanding the CFA model to a SEMmodel

The Structural Equation Modeling or popularly known as SEM is a statisticalanalysis techniques developed for analyzing the inter-relationships among multiplevariables in a model The inter-relationships among variables could be expressed in

a series of single and multiple regression equations The Structural EquationModeling technique employs the combination of quantitative data and thecorrelational or causal assumptions into the model The relationships are modeledinto a theoretical framework represented by a schematic diagram The schematicdiagram presents the hypotheses of interest to be tested in the study SEM is capable

of estimating a series of inter-relationships among latent constructs simultaneously

in a model In fact, SEM is the most efficient method to handle the ConfirmatoryFactor Analysis (CFA) for measurement models, analyze the causal relationshipsamong latent constructs in a structural model, estimating their variance andcovariance, and test the hypotheses for mediators and moderators in a model[ CITATION ZAw12 \l 1033 ].

All variables in both CFA and SEM will be tested for 3 categories of fit:parsimonious, incremental and absolute Absolute fit indices do not use analternative model as a base for comparison They are simply derived from the fit ofthe obtained and implied covariance matrices and the minimization function Chi-square is the original fit index for structural models because it is derived directlyfrom the fit function Relative fit indices compare a chi-square for the model tested

to one from a so-called null model The null model is a model tested that specifiesthat all measured variables are uncorrelated (there are no latent variables) The nullmodel should always have a very large chi-square (poor fit) Parsimonious fitindices are relative fit indices that are adjustments to most of the ones above Theadjustments are to penalize models that are less parsimonious, so that simplertheoretical processes are favored over more complex ones The more complex themodel, the lower the fit index [ CITATION New05 \l 1033 ]

In summary, the key difference between path analysis and SEM is that theformer analyzes relationships among observed variables, while the latter focuses onrelationships among latent variables

3.3 Research model

3.3.1 Variables, hypotheses and equation

Taking into account the aforementioned 7 factors: beer characteristic,branding, beer types, situation appropriateness statements, packaging, social mediaand country of origin, all of which are latent variables by which would be furtherexplore through the use of questionnaires (detailed in section below) Each of theseseven independent variables will be represented by five questions to explore therelationship between factors, based on the questionnaire developed by the research

of [ CITATION Sunkamol16 \l 1033 ] Below is a list of the questions and variables:

Table 3.1: Acronym for variables

Beer characteristics I choose beer based on its aroma CR1

I choose beer based on its foam

I have a clear image of the type

of people who drink my favorite

beer brand

BRAN5

Beer types I can tell the difference between

malt, fruit or wheat beer

I prefer malt beer than any others BTY3

I prefer craft beer (Bia Tươi) than

Situation appropriateness I drink beer when I am at a

casual dining restaurant

SA1

I drink beer when I relax alone at

home

SA2

I drink beer when I am at parties SA3

I drink beer when I want to impress someone

SA4

I drink beer for a special occasion.

SA5

Packaging Buying beer with extravagant

packaging makes me feel good

packaging Beer's packaging design can be a source of satisfaction for me

I use social media to interact with

my favorite beer brand

SM2

I use social media to catch up with my favorite brand's promotional activities

International beer brands have

I often drink international beers more than domestic beers

CO5

Brand preference I am loyal to only a selected few

of my favorite beer brands

brand

BPRE3

I will definitely purchase my favorite beer brand in the future

BPRE4

I am completely confident in the quality of my favorite beer brand

BPRE5

Source: Author compilation

The questionnaire was created on the basis of foundations laid by pastresearches To be more precise, each of the seven latent variables will be explained

by five indicators in form of questions, encrypted both to best represent the mainvariable and later, to accommodate analysis To best quantify these observedvariables, a 5 points Linkert scale was developed This is, as per my research, thebest way to quantify quantitative research providing sufficient report on Cronbach’salpha coefficient for internal consistency and reliability testing in multiple variables[CITATION Jos03 \l 1033 ] More details on the questionnaire can be found on theappendix

From the conclusions drawn from a variety of researches done above, I sum

up the following hypotheses:

Hypothesis H1: Beer characteristics positively influence brand preference.Hypothesis H2: Branding positively influence brand preference

Hypothesis H3: Beer types positively influence brand preference

Hypothesis H4: Situation appropriateness positively influences brandpreference

Hypothesis H5: Packaging positively influences brand preference

Hypothesis H6: Social media positively influences brand preference

Hypothesis H7: Country of origin positively influences brand preference

3.3.2 Research model

Figure 3.2: Research model

Source: Author compilation

3.4 Data analysis

3.4.1 Population and Samples

The population of this research are able people who are currently living inHanoi and over the legal age of consuming alcohol over 18 years old and have atleast consumed alcohol in the past (either one-time consumption or frequentdrinkers) The sample was collected, using non-random sampling in terms ofconvenience sampling by gathering data from drinkers who are willing toparticipate with the surveyors by filling in the given questionnaires [CITATION Tro07 \l

1033 ]

The choice of sample size was influenced by past research on factorsaffecting beer choice in Nha Trang by [ CITATION Ngu17 \l 1033 ] In his research hechose 390 respondents to conduct his analysis However, according to Springer'sstudy [CITATION San \l 1033 ], any sample size above 200 or better yet, 300 can beconsidered ideal to conduct further analysis, in this case with my CFA and SEM.Therefore after removing all duds from my data list, which include those who havenever drank alcohol or those with unfinished answer, my chosen sample size for thisresearch is exactly 300

As mentioned above, this research is created with the purpose to put brandpreference in a conceptual framework with other affecting variables Seven latentvariables (beer characteristics, branding, beer types, situation appropriateness,packaging, social media and country of origin) are chosen to determine theircorrelation with a singular independent variable brand preference, as displayedabove

3.4.2 Data collection

Due to the lack of resources to commit a true random sampling (to eliminatepotential biases) or to perfectly survey the entire population of Hanoi, non-randomsampling was used to collect data on this research To be more detailed, thisresearch was conducted using convenience sampling by publishing online surveysand sending those to students, office workers and acquaintances of the surveyoraround Hanoi

As mentioned, the study was conducted from March, 5th 2019 to May 30th

2019 The first month of March was spent on completing the literature alongsidewith drafting the questionnaire in Appendix The first set of questionnaire wasinspired by that of another research on customer beer preference in Bangkok[ CITATION Sunkamol16 \l 1033 ] From April 1 st to May 5th, an online researchquestionnaire was published on Google Form, for convenience purposes, and sharedamong students of Foreign Trade University, National Economics University, ArenaMultimedia, Hanoi University of Science and Technology, Vietnam NationalUniversity, University of Science, University of Social Sciences and Humanities,Academy of Journalism and Communication, Diplomatic Academy of Vietnam,Chu Van An high school Alumni, Trung Vuong Secondary Alumni, a variety ofcompanies in Hanoi, friends of friends of the author and the neighborhood of Hai

Ba Trung district

3.4.3 Data analysis procedure

Primary data source from questionnaires will be analyzed to conclude thecausal relationship between determinants affecting brand preference in the beerindustry in Hanoi As mentioned above, these determinants are beer characteristic,branding, beer types, situation appropriateness statements, packaging, social mediaand country of origin

Data regarding these factors will be quantitatively analyzed via tools offered

by IBM SPSS, and further analysis by IBM AMOS, where scale reliability wastested using Cronbach Alpha, Exploratory Factor Analysis to test dimensionality,Confirmatory Factor Analysis to measure how well the measured variables canrepresent the number of constructs and Structural Equation Modeling (in AMOS) toexplore the connection of latent and observed variables

CHAPTER 4: RESEARCH FINDINGS4.1 An introduction about beer industry in Vietnam

While beers making has been present for thousands of years in Europeanregions, the history of beers in Vietnam only officially started in the 19th century -through the French colonial regime's introduction[ CITATION Vie19 \l 1033 ] With themain purpose at first to quench the thirst of French soldiers, beer gradually became

a favorite drink, especially to the middle-class Vietnamese due to its lighter andmore refined taste than the accustomed Vietnamese liquor, combined with anaffordable price for household consumption

According to report made by FPT Securities [CITATION FPT18 \l 1033 ]; the beerindustry is still debatably a growing industry - with an ever increasing capacity of 4billion liters annually nationwide This figure is a combined effort of big domesticbreweries (like Sabeco, Habeco…) and other international brands such as Heineken(VBL), Carlsberg… This trend illustrates how Vietnam has been so successful inattracting international investment onto this field, with such a short developmenthistory comparing to other nations As a matter of fact, in 2017 Vietnam ranked10th fastest growing, surpassing 15 other Asian countries since 2007 Vietnam evenhas the highest Compound Annual Growth Rate and year-over-year compared toothers in the beer industry This is furthermore, confirmed by a report on rapidincrease in national alcohol consumption [ CITATION Mar15 \l 1033 ] Taking account ofall these figures, Vietnam definitely has a lot of potential for growth in the alcoholbeverage sector

Table 4.1: Top 10 Asian countries with highest market growth

Liter)

%Worldcapacity

Yoygrowth

CAGRover10years

Capacity(Bil

Liter)

Yoygrowth

Source: Ministry of International Trade and Industry

Despite the expanding opportunities for big players in this particularindustry, statistics have shown Vietnamese breweries (such as Sabeco, Habeco orCarlsberg) all suffer a decline in popularity in market share At the same time,international brands (Heineken) rapidly expand theirs This demonstrates how localbeer brands are losing advantages over international ones As beer is such apromising industry, a well-design research is in dire need so that marketers caninitiate effective strategies to compete in this competitive Fast-moving consumergoods industry

Figure 4.1: Vietnamese market-share 2012 - 2016

Source: Euromonitor International

One possible explanation for this trend, according to report from FPTS[ CITATION FPT18 \l 1033 ], is changing consumer habits Out of the 3 main sectors:Premium beers, middle-class beers and low price beers, consumers are demandingbetter quality products and consumption experience, therefore, shifting theirpreferred beers upwards to premium and middle-class products Unsurprisingly, thissector is dominated by Heineken - due to their efficacious global brandingstrategies

One other explanation for this shift in habits can be explained throughpsychological factor Apart from consuming this kind of beverage to quench theirthirst, Vietnamese consumers have a high desire to make themselves "look good" infront of others in social meetings This finding coincides with that of [CITATIONPet03 \l 1033 ] which suggests furthering interpersonal relationships is among themajor factors influencing the decision to purchase and consumer beer products Myresearch will, additionally, explore how this factor can affect brand preferenceunder the category of "Situation Appropriateness Statement"

Alongside with that, Vietnamese consumers are reportedly to possessfavoritism attitude towards foreign brands, believing them to be better in terms ofquality and reputation Under the category of "Country of Origin", as suggested bySunkamol [ CITATION Sunkamol16 \l 1033 ], I will discover the correlation of brandpreference and this factor in this paper

This subject has not been widely covered yet in the literature of Vietnameseacademia Among those I found most notable was the paper in Nha Trang city byNguyen Tien Thong [CITATION Ngu17 \l 1033 ] which suggest how brand, price andpackaging format could have an effect on customer's choice of beers In addition,Sunkamol's in depth research in Bangkok has found out that beer characteristic,branding, beer types, situation appropriateness statements, packaging, social media,country of origin, and convenient and practical functions all could potentially have

an influence towards beer brand preference; while Gabriel A Okwandu [ CITATIONGab01 \l 1033 ] has confirmed these factors (price, taste, state of origin and

packaging) indeed affect those in Nigeria Because of cultural and geographicalreasons, this research will intensively focus on how 7 factors: Beer characteristics,Branding, Beer types, Situation appropriateness statements, Packaging, Socialmedia, Country of origin could together influence a latent variable - brandpreference.

Table 4.2: Demographics - Gender

Gende

r

Source: Author compilation

The sample consists of 192 male participants and 108 female participants.This, however, does not necessarily indicate a higher number of men who drinkbeers than women in the whole population Perhaps more exhaustive research withtrue random sampling could eliminate the bias of convenience sampling in myresearch, thus potentially giving a more conclusive resolution

Table 4.3: Demographics - Age

50+ 12 4%

Source: Author compilation

The sample was divided into 5 age groups Those who are under 23 take upthe majority of the sample (up to 49%), following after by those in their 24-29 (30%total participants) while those from 40 and above only accumulate a meager 7%.This is again, the result of sampling technique and might not necessarily reflect thewhole population

Table 4.4: Demographics - Marital status

Source: Author compilation

About marital status of the participants, the majority of those who take part

in the survey are unmarried (55%) This trend is further observed in the "age"category above, since the majority of participants are young people, they are lesslikely to have already started a family However I must take note of the existence of27% who chose "Other" option, as this could indicate an unexplored option

Table 4.5: Demographics - Level of education

Above graduate study

Source: Author compilation

Consistent with the conclusion drawn above, the utilization the nonrandomsampling technique might have left some biases in the representing sample In thiscase of Level of education, 82% of 300 people who enter the survey have a college/university degree while none has lower than high school diploma, which meansthose that has lower than high school diploma could be underrepresented by thesample.

Table 4.6: Demographics - Monthly income

Source: Author compilation

Taking account of monthly income, most of the participators enjoy above 8.6million VND bracket (at 57%) While 29% have lower than 3.5 million VND, only

a small percentage fall into the middle options 3.6 - 5.5 million and 5.6 - 8.5 millionwith 4% and 9% respectively

Table 4.7: Demographics - Consumption frequency

Consumptio

n frequency

Several times perweek

Several times permonth

Source: Author compilation

For the frequency of beer consumption, the most common answer is "Forspecial occasions only" with 64% This was confirmed by [ CITATION Ngọ18 \l 1033 ] in

a study conducted in Nation Economics University where "Special occasions" andcelebrations are among the chief reasons why Vietnamese people consume alcohol.The second largest group was those who drink "Several times per month" at 19%,followed by "Several times per week" at 10% Only a trifling 1% of people chosethey drink "1 time per month"

Table 4.8: Demographics - Favorite beer brands

Source: Author compilation

Finally, participants are asked to choose their preferred beers from acompiled list of available beers in Hanoi market [ CITATION FPT18 \l 1033 ] Since there

is no restriction in how many items a single person can choose, this question can beused to infer the most popular beer brands amongst the sample The most popular

choice is believed to be Heineken with a stupendous 87% pick rate, followedclosely by Sai Gon Beer at 80% Hanoi came third with being a favorite of morethan half of the participators (59%) And those at the bottom are Sư tử trắng, Halidaand Huda with respectively 4%, 6% and another 6%.

4.3 Descriptive statistics

Table 4.9: Questionnaire descriptive statistics

N Minimum Maximum Mean

Std

DeviationCR1 300 1 5 2.64 977 Innate characteristics (CR1)CR2 300 1 5 3.23 996 Innate characteristics (CR2)CR3 300 1 5 2.49 1.000 Innate characteristics (CR3)CR4 300 1 5 2.71 1.033 Previous experience (CR4)

2

Reputation (BRAN2)BRAN

3

Confidence (BRAN3)BRAN

4

Confidence (BRAN4)BRAN

5

Favorite brand (BRAN5)BTY1 300 1.0 5.0 3.500 8321 Ingredients (BTY1)BTY2 300 1.0 5.0 3.183 9519 Ingredients (BTY2)

SA1 300 1 5 2.17 803 Outdoor context (SA1)

SA4 300 1 5 2.23 856 External influence (SA4)