Segmentation of the car market in China

Segmentation of the car

market in China

2013

1

Table of Contents

Abstract ... 5

1. Introduction ... 6

1.1 Background ... 6

1.1.1 The growing Chinese car market ... 6

1.1.2 Government intervenes to control the car market ... 7

1.1.3 Changes in Chinese market segments ... 7

1.2 Problematization ... 7

1.3 Research questions ... 8

1.4 Problem formulation ... 9

1.5 Delimitations ... 9

1.6 Outline of the thesis... 9

2. Theory ... 10

2.1 Market segmentation ... 10

2.2 Bases for Market Segmentation ... 11

2.2.1 Geographic segmentation ... 11

2.2.2 Demographic segmentation ... 12

2.2.3 Psychographic segmentation ... 13

2.2.4 Behavioural segmentation ... 13

2.3 Macro Segmentation ... 14

2.4 Micro segmentation ... 15

2.5 Hybrid segmentation: the mix of micro and macro segmentations ... 15

2.6 Dynamic segmentation: the predictive implication of segmentation ... 16

2.7 State-of-the-art ... 16

2.8 Model, how the constructs relate... 19

3. Methodology ... 21

3.1 Introduction ... 21

3.2 Population and sample ... 21

3.2.1 Data collection method ... 22

3.2.2 Questionnaire Design ... 22

3.3 Operationalization ... 23

3.4 Data analysis method ... 28

3.5 Validity ... 28

2

4. Empirical Data ... 31

4.1 Geographic data: ... 31

4.2 Demographic data: ... 31

4.3 Behavioural data:... 32

4.3.1 Customer behaviour explained by Geographic data ... 33

4.3.2 Customer behaviour explained by Demographics: ... 34

4.4 Psychographic data:... 35

4.4.2 Respondent’s opinion on global preferences ... 36

4.4.3 Car owners’ current preferences ... 37

4.4.4 Car owners’ preferences for their next purchase ... 40

4.4.5 Non-car owners’ preferences for their first purchase ... 43

4.5 Macro data for the dynamic aspect ... 45

5. Analysis... 49

5.1 Control Variables ... 49

5.2 Hybrid Behavioural segmentation ... 49

5.3 Hybrid Psychographic segments ... 50

5.4 Dynamic Hybrid Psychographic segments ... 52

6. Conclusion and discussion ... 53

7. Managerial implications, reflections, further research ... 55

7.1 Managerial implications ... 55

7.2 Reflections ... 57

7.3 Limitations ... 58

7.4 Future Research ... 58

References ... 60

Appendix A ... 64

The Survey ... 64

Survey emails ... 68

Descriptive statistics ... 71

Appendix B

Linear regression analysis and linear curve estimates ... 77

3

B.1.1. Region of residence ... 77

B.1.2. Degree of urbanisation. ... 80

B.1.3. Behaviour by age ... 82

B.1.4. Behaviour by Gender ... 84

B.1.5. Behaviour according to annual income... 87

B.2. Psychographic data by Demographic and Geographic ... 89

B.2.1. Psychographs by age ... 89

B.2.2. Psychographs by gender. ... 103

B.2.3. Psychographics with Annual income (Yuan) ... 115

B.2.4. Psychographics by region of residence. ... 127

B.2.5. psychographics by degree of urbanisation ... 143

B.3. Customer psychographics by behaviours ... 156

B.3.1. Psychographics by car replacement rate ... 156

B.3.2. Psychographics by experience ... 169

B.3.3. Psychographics with frequency of use ... 187

B.3.4. Psychographics by non-car owners who plan to have a car in the future. ... 201

B.3.5. psychographics by style of use; long trips ... 216

B.3.6. Psychographics by style of use; short trips ... 229

4

ACKNOWLEDGMENT

This thesis was written between March 20th 2013 and May 25th 2013, at the Linnaeus

University, Campus Växjö. The authors have received support from several people during

this period. In this acknowledgment page they want to thank all these wonderful people, for

their assistance during this thesis.

First of all we would like to thank their examiner Doctor Sarah Philipson who has been

giving the authors support, feedback and inspiration. The authors would also like to thank

their tutor Professor Anders Pehrsson who has supported them by giving much needed

feedbacks and inspiration; he has also given inputs in developing the model used throughout

the thesis. The authors also want to send out their gratitude to the respondents of their survey.

Their contribution to the thesis has been of great value and the authors are very much

thankful for their precious time.

Växjö, May 25th, 2013

5

Abstract

The Chinese car market has, through the last decade evolved into the major market in the

world. Its car market from has become the world’s largest market from 2009 until today.

With the emerging market that is China, the demand for cars is supposed to grow even more

in the next decade.

The thesis starts by studying the theories of consumer market segmentation with a hybrid and

dynamic aspect. A quantitative investigation was conducted with the help of a survey. The

survey was sent out to car consumers and potential car consumers who are residing in China.

From this study the authors were able to anticipate possible preferential profiles.

6

1. Introduction

1.1 Background

The history of the automobile is complicated and rich. It dates back to the 15th century, when

Leonardo da Vinci was designing and modelling transport vehicles (library of congress,

2013-3-22). According to the library of congress (2013-3-22) Nicolas Joseph Cugnot built the

first self-propelled vehicle for the French army in 1769. 1832-1839, Robert Anderson in

Scotland built an electric carriage. George Baldwin Selden integrated an internal combustion

engine in the United States in 1876/95.

The first companies to produce automobiles for the public market emerged in 1896,

manufacturing electric gas-powered vehicles (Bentley historical library, 2013-3-22). The love

of cars and their ever-growing necessity in our lives has increased the number of automobiles

to an estimate of over 1 billion vehicles in 2010, from 500 million in 1986 (Autos Cars,

2013-3-22).

1.1.1 The growing Chinese car market

During the last two decades social and economic events have brought immense opportunities

in soaring emerging markets. Many multinational corporations have made substantial

investments in these countries, as a part of their global expansion strategies, creating more

and more jobs thus increasing the purchasing power of customers. Such an emerging market

is China (Cui, & Liu, 2000).

Half a century ago bicycles were considered a mark of family fortune in China (The guardian,

2013-3-22). Since then, China has become the world’s largest automobile producer and the

largest market with annual sales of some 14 million vehicles in 2009, continuing to expand in

2010 according to APCO (2013-3-22).

7

1.1.2 Government intervenes to control the car market

The Chinese government has implemented a number of tax adjustments and subsidies for

hybrid electric vehicles, electric vehicles, and traditional vehicles with small engine

displacement, attempting to impact the national consumption (APCO, 2013-3-22).

1.1.3 Changes in Chinese market segments

Richard Cant of Dezan Shira & Associates at the HSBC International Exchange to China said

“For the first time ever there are more people living in China’s cities than in the countryside

and this trend will only continue,” (HSBC, 2013-3-22). There is an increasing middle class in

China (HSBC, 2013-3-22). The Chinese consumers are now spending their disposable

income rather than saving it, as they did before. They are also educated, skilled, price savvy

and internet savvy (HSBC, 2013-3-22). Foreign brands, principally for visible products, are

very popular and trendy among China’s young consumers, as they have status appeal (HSBC,

2013-3-22).

McKinsey (2013-3-22) points out that more of China’s middle class is shifting to premium

cars, notably foreign cars like VW, BMW, Volvo, Aston Martin, et cetera. McKinsey

(2013-3-22) also states in his report that on average Chinese families change car every six to eight

years on average. Premium car owners say that they would like to change their car two to

three years ahead of the average (McKinsey 2013-3-22).

1.2 Problematization

The Chinese customers have shown a strong interest in cars, making it a challenging

environment for marketers, and the industry now sees the market as the most important in the

world. The demand evolved radically and is prone to continue changing. As the Chinese car

market is emerging, the marketing processes are evolving with it as the market changes really

quickly. Segmentation, a key marketing process to reach customers, has to be very efficient

for car manufacturers that want to perform well in the Chinese car market.

8

gain market shares on the promising market and be ahead of the competition in the heart of

Chinese people. This is where the importance of segmenting efficiently the market appears.

Segmentation can be made on basic macro dimensions: economic, geographic, culture & and

technology (Grundy, 2006). Specific segments can also be identified by carrying out

segmentation based on micro dimensions, such as life styles, behaviors etcetera (West, Ford

and Ibrahim, 2006). Those segmentation styles are chosen by car makers and have a static

dimension, and feature no flexibility or enough details to gain a clear understanding of the

demand.

The whole environment affecting the segments is bound to evolve (Grundy, 2006). This will

leave the company marketing their cars in an obsolete way, positioning themselves on

segments that they could not target as efficiently, as if they were adapting to them. The

surroundings of car manufacturers are influencing the way segmentation is carried out and in

its results. In this emerging market, with its quickly changing environment, “hybrid

segmentation”; mixing different styles of segmentations, is said to be necessary to have

detailed customer segments (Hassan, Craft and Kortam, 2003).

Micro and macro dimensions are connected and influencing each other continuously, causing

a challenge for companies as their segments evolves (Cannon and Yaprak, 2011). Marketers

and academics like Hasan, Craft and Kortam (2003); Amine and Smith (2009); Cannon and

Yaprak (2011); and Quinn, Hines, and Bennison (2007) have discussed the mix of macro and

macro to anticipate possible segmentation, these studies are rather recent. But there has been

little studies done on segmenting the Chinese car consumers. The study of consumer

preferences of the Chinese car consumers is there for a gap to be recognised. Finding the

different segments and build predictions of their evolution in the near future is a really

practical source of information to actually serve the challenges imposed to firms marketing

cars in China.

1.3 Research questions

9

1.4 Problem formulation

The purpose of this paper is to study customer segmentation in the Chinese new car market,

and the changes and movements of segmentation that could be predicted for the next

purchase.

1.5 Delimitations

The study is delimited in time and location. The research is made only on the Chinese market

of new cars. The focus is also time-based since the authors will study trends for the next

purchase.

1.6 Outline of the thesis

10

2. Theory

The four ‘Ps’ of strategic marketing is useful to analyse a marketplace according to Kotler

(1989).

He terms them as ‘probing’. The second step is ‘partitioning’ the market; this is the process

of segmenting clusters of the customers of a market. The third step is ‘prioritising’ the market

segments that a company has possibly greater advantage in pursuing. The fourth step is

‘positioning’; “pinpointing the competitive options in each segment that you're going to

target.” (Kotler, 1989). Partitioning is the term referring to customer segmentation, which is

the focus of the thesis.

2.1 Market segmentation

Market segmentation is an essential element of marketing for goods and services. The

recognition of heterogenic demand needs to be considered before a newly produced product

is being considered to be sold in a market (Wedel and Kamakura, 2000).

Markets as a whole are not entirely homogeneous (Beane and Ennis, 1987). Market segments

are homogeneous subsets of consumers in a particular market (West et al., 2006). Market

segmentation is a vital process to succeed, identifying suitable segments of consumers with

shared preferences for targeting purposes (West et al., 2006; Kotler and Keller, 2009).

Beane and Ennis (1987) gave the following major reasons for market segmentation:

(1) To look for new product opportunities or areas that may be accessible to current

product repositioning; and

(2) To create improved advertising messages by gaining a better understanding of

one’s customers.

In segmenting a market to sell a product it is not necessary to identify all segments, only the

groups that appear to be yearning for the product (Beane and Ennis, 1987).

There are many ways to assess market segmentation. To determine its effectiveness and

profitability some criteria have often been put forward: identifiability, measurability,

sustainability, accessibility, responsiveness and actionable (Frank, Massy, and Wind, 1972;

11

Identifiability and measurability are the extent to which marketers are able to distinguish and

approximately measure the clusters of customers according to its size, location and market

from within a market (Wedel and Kamakura, 2000; Beane and Ennis, 1987).

Sustainability/ stability; the segment must be large enough to justify attention (Beane and

Ennis, 1987). Kotler, in Beane and Ennis (1987) says: "a segment should be the largest

possible homogeneous group of buyers that it pays to go after with a specially designed

marketing program". Wedel and Kamakura (2000) states market segments that are steady for

a long enough period can provide a foundation for the development of successful market

segmentation

Accessibility; Segments must be accessible to the managers to some degree. It can be

accessible for example through promotional efforts or through secondary data (Wedel and

Kamakura, 2000).

Actionable; “Market segments are actionable if their identification provides guidance for

decision on the effective specification of the marketing instruments”, (Wedel and Kamakura,

2000).

2.2 Bases for Market Segmentation

There are many ways to segment a market, amongst those, Beane and Ennis (1987)

acknowledge Kotler to divide market segmentation variables into four main parts;

Geographic;

Demographic;

Psychographic;

Behavioural;

2.2.1 Geographic segmentation

12

“Consumers in the Southeast use more vegetable shortening than in any other part of the US.

Since Northeastern and Midwestern regions have more locally unique beer drinking

segments, they have more small beer breweries than any other region. The consumption of

menthol cigarettes is greater in the Southeast than in any other region of the country”. These

are few examples given by Beane and Ennis (1987).

2.2.2 Demographic segmentation

Beane and Ennis (1987) refer to demographic segmentation to be the most prevailing variable

of market segmentation. They think that’s probably because consumers are placed on certain

scales of measurement that can easily be understood; information is easily gathered,

interpreted, and convertible from one study to another. Satisfying the segmentation criteria of

identifiability, substantiality, accessibility, and actionability (Wedel and Kamakura, 2000),

three of the most common demographic variables employed in domestic and international

segmentation include: age, gender, income (Beane and Ennis, 1987; Cleveland et al, 2011).

Examples are Life Stage vitamins (four types depending on age and sex) and disposable

diapers according to the age of infant (Beane and Ennis, 1987). Younger techno-savvy

individuals are less committed to definite patterns and are more open to new perspectives and

products especially when it comes to high-technology products (Cleveland et al., 2011).

It is well recognised that the demand for categories of goods and services alters as they age

and pass through the various life cycle stages (Cleveland et al., 2011).

Income also has a strongly effect on individuals demand or choice of a product (Cleveland, et

al., 2011). Consumers with higher- income tend to buy better and expensive products, which

are also status- enhancing, examples are household appliances, consumer electronics, and

luxury products (Cleveland, et al., 2011).

13

2.2.3 Psychographic segmentation

According to Beane and Ennis (1987) psychographic or life-style segmentation is tougher to

describe as it is no longer looking at clearly definable, quantitative measures. Psychographic

segmentation, unlike geographic and demographic segmentation, reflects consumers’ choices

depending on personality, values and lifestyles, though there has been no clear agreement

about the variables that should be included (Lancaster and Williams, 2002; Kim and Lee,

2011; Wells, 1975).

Wells (1975), while writing a review article about psychographics, stated: “Twenty-four

articles on psychographics contain no less than 32 definitions, all somewhat different”. While

reviewing a Product-Specific Psychographic Profile Wells (1975) argued that when a

psychographic study is devoted to a product category, it is not essential to rely on item

diversity to get useful relationships. The investigator can focus upon a limited set of relevant,

product-related dimensions that would make more sense regarding the psychographic

dispositions of the population investigated towards the product in question.

Product-specific psychographic profile is a useful tool used to get insights of the market

dispositions towards a product. Such a study is investigating profiles through series of

questions about product attributes (Wells, 1975; Chin-Feng, 2002).

According to Chin-Feng (2002), products attributes are reliable to study personalities and life

styles. He states that the psychographics regarding a product are driven by the brand equity of

the current actors. Due to the personality associated to brands, clusters can be found

investigating the preferences of the customers. The analysis of personality traits and life

styles through their linkage to product attributes is preparing ground to the response an

individual can have when confronted with a purchase decision on the market (Chin-Feng,

2002).

2.2.4 Behavioural segmentation

14

segmentation”. Behavioural segmentation pays less attention to economic or age related

issues unlike psychographic segmentation, and concentrates more on consumer related

behaviour, concentrating on psychological attributes and product orientation (Lancaster and

Williams, 2002).

Behavioural segmentation, segments consumers based on knowledge of the product, attitude,

or response to the product. This includes such areas as purchase occasion, benefits sought,

user status, degree of usage, degree of loyalty, readiness stage, and marketing factor

sensitivity (Beane and Ennis, 1987).

Holbrook and Hirschman (1982) recognise that people buy products not only for its use, but

also for what they signify or mean. Consumer behaviour also relates to the multi-sensory,

fantasy and emotive aspects of one's experience with products. These facets of consumer

behaviour are defined as Hedonic consumption; Multi-sensory means several sensory

modalities including tastes, sounds, scents, tactile impressions and visual images. Emotions

represent motivational phenomena with characteristic neurophysiological, expressive and

experiential factors; they include feelings such as joy, jealousy, fear, rage and rapture.

Holbrook and Hirschman (1982) further proposed:

Differences in consumer emotional and imaginary response to products are closely tied to a

variety of subcultural differences, for example ethnicity or nationality.

Differences in consumer emotional and imaginary response to products appear closely tied to

a variety of subcultural differences.

Consumers of many hedonic ally-experienced products are characterized by very distinctive

social class profiles.

2.3 Macro Segmentation

15

2.4 Micro segmentation

Gaston-Breton and Martin (2011) stress the need of micro segmentation at a consumer level.

They present three classes of micro-segmentation that have been used in most European

segmentation studies, from product-specific characteristics, such as attitudes toward

attributes, to domain-specific features such as lifestyles, and general characteristics such as

central values. It is important to note that a country's culture has also been recognized as a

significant environmental trait and essential to understand systematic variances in behaviour

(Steenkamp, 2001). For example, markets around the world are getting more and more

cosmopolitan, but the consumption of local foods and fashions are across most countries

dominated by products compliant with the specific country’s ethnic identity (Cleveland et al,

2011).

Holbrook and Hirschman (1982) pointed out in an article some years back, about the different

aspect of consumption. He states that all products, no matter how ordinary they are, may

carry a symbolic significance. This significance in some cases is rich and noticeable to the

users. Holbrook and Hirschman (1982) also mention that many products projects essential

nonverbal indications that must be seen, heard, tasted, felt, or smelled to be appreciated

properly.

2.5 Hybrid segmentation: the mix of micro and macro segmentations

The concept of hybrid segmentation is relying on the principle that segments can be made

more profoundly by combining different segmentation methods. Hybrid segmentation in

different areas of the world, standardising some marketing elements, and adapting others,

could be a good solution (Hassan, Craft, and Kortam, 2003).

Segmenting a market should start with the macro dimensions and refined by micro factors

(Amine and Smith, 2009) to adapt to local behaviours and thus getting better performance.

16

previous sections is resulting in mass customization. If macro data is associated to micro data,

the segmentation allows the distinction of clear precise segments that can be targeted to

achieve greater customer satisfaction through coherent value delivery (Jiang, 2000).

2.6 Dynamic segmentation: the predictive implication of segmentation

Very few authors have emphasised the importance of dynamic segmentation. Among these

few is Weinberg (1972), who stressed its importance dynamic segmentation. He stressed, that

certain episodes can be predicted by the use of already available records. Weinberg (1972)

referred to Dr. Spock’s example of segmenting the baby product market: on the average,

three months after birth a baby begins to eat solid food, five months later the birth weaning

should begin, and a year later, a baby normally grows six teeth and can chew. Those different

periods are called time sequences. Episodes like these or the application of time sequence can

be beneficial to marketers and researchers in anticipating the future, while segmenting

international market and its trends. A similar approach can be applied to develop time

sequences for other areas such as automobiles (Weinberg, 1972).

According to Holbrook and Hirschman (1982), post purchase consumer behaviours are driven

by a learning effect, transforming the initial behaviour. This phenomenon results in a

feedback loop and depends on the interaction between the purchase experience and the

consumers’ inputs. The consumers’ inputs are defined by the macro and micro factors

influencing them. The customers’ behaviour is thus ever changing, due to the different

contexts the customers could be in. (Holbrook and Hirschman, 1982)

Quinn, Hines and Bension (2007) explained and made sense of the particular observed

phenomenon of marketing segmentation, ongoing. The aspect of ongoing, explained by

Weick (1995), is unending and continuous sense-making. In a milieu of dynamic

segmentation, in a world, which is becoming more and more cosmopolitan, cluster behaviour

is always ever hanging and ongoing.

2.7 State-of-the-art

17

accepted and it reviews Kotler (1980) theories. Wedel and Kamakara (2000) dominating

theory on market segmentation is also well accepted and validated.

Cleveland et al. (2011) Lancaster and Williams (2002), and Kim and Lee (2011) are the other

sources that have been used to explain bases and variables for market segmentation. The for

variables geographic segmentation, demographic segmentation, psychographic segmentation

and behavioural segmentation are the most common segmentation variables by Philip Kotler.

The sources are recent and do not have a low acceptance or validation and so are proposals to

fill a gap. These sources have been used to discuss the different segmentation variables.

Gaston-Breton and Martin (2011) and Douglas and Craig (2010) are again very recent articles

and has a low acceptance and validation. Douglas and Craig (2010) is a study that provides a

framework to understand contextual factors as a means to refine entry strategy and develop

effective segmentation strategies. Gaston-Breton and Martin (2011) proposes a two stage

model of international market segmentation. These sources have only been used in our thesis

to understand Macro segmentation.

Cleveland et al, 2011) and Steenkamp (2001) have been used for micro segmentation.

Steenkamp (2001) has moderate acceptance and is well validated. Steenkamp (2001) reviews

and discusses the role of national culture in international marketing research, the frameworks

presented by Hofstede and the Schwartz. Holbrook and Hirschman (1982) is another very

dominating article that has used to explain micro segmentation and to explain behavioural

segments. It has further been used to elaborate more on how knowledge that is gathered

through time effect segmentation. Holbrook and Hirschman (1982)’s article have been highly

accepted and is well validated.

Wells (1975) article review has also been very important in understanding psychographics.

Wells (1975) is well accepted and well validated.

18

(2011) have also been used as source for the integration of macro and micro segmentation.

This source has very low amount of acceptance but is thoroughly validated.

19

2.8 Model, how the constructs relate

Figure 1: The Analysis Model for dynamic segmentation

This model shows the micro and macro segmentation styles. The model features the influence

of the different segmentation styles: micro factors are influenced by macro ones.

Geographic and demographic criteria influence the behaviour of customers and their

preferences. For instance, according to where they live and factors such as their purchasing

power will have different behaviours towards a product offer. Moreover those criteria also

influence their way of thinking towards a product and why they are buying it, their tastes and

preferences. Here, the demographic and geographic data help predicting a behaviour and

psychographics of a person.

20

The model of analysis therefore includes a diverse set of criteria aimed at a complete

understanding of segmentation, and offers the possibility to predict eventual changes in

segmentation needs with the help of purchase-intention-oriented questions. This last aspect is

particularly important in an uncertain environment, such as the Chinese car market.

21

3. Methodology

3.1 Introduction

This chapter presents the different stages of the research and how it has been used in the

thesis. The data collected derives from quantitative research, an online survey, which was

sent out to China. The authors presented a survey based cross tabulation to study

segmentation of the Chinese car market with a time aspect; they also included secondary data

from different time period to capture a dynamic view of the Chinese car market.

3.2 Population and sample

A population is all the people belonging to a group or a geographical area. A segment of the

population is chosen for research, when sampling is done. Bryman and Bell (2005) states, a

sampling frame is the total list of units who are a share of the population from which a

sample is derived. A convenience sample means that the people in it happens to be available

to the researcher, making it easier, cheaper and less time consuming than a probability sample

which demands more preparations (Bryman & Bell, 2005).

The population consists of students and people from all walks from China. Survey questions

were sent to various companies. The authors found list of China-based companies and

manufacturers; there were no specific choice criteria only they had to be from China. From

the list a lot of emails were invalid so they could not be reached. A major number of peoples

contacted emailed back that the thought the survey link was a spam, and they refused to

answer the survey. The authors have also contacted friends who are residing in China to

answer the survey. The authors have also contacted various universities asking if they could

help the authors circulating the survey within their universities. Unfortunately no

confirmation that the email had been dispatched further to the students has been received.

The authors have also tried approached Chinese individuals with the age group of 65 and

above, by trying to contact various nursing homes for elderly in China via email, but

unfortunately the authors were not able to contact them.

22

Due to time restrain the authors had to close the survey with 89 respondents with a response

rate of 17.4%. As the survey was answered by individuals with a varied age and income

group and also from a scattered geographical location, the number of answers is judged

important enough to carry out an analysis.

3.2.1 Data collection method

Surveys can be conducted on paper or electronically. The respondents can answer

individually. The electronic survey is the most common research method today. (Bryman &

Bell, 2007).

The primary quantitative data collection was made via the online survey provider called

Keysurvey. This service provider is convenient to use and to collect data from, since it is easy

to export SPSS files. The SPSS file was ready to use directly in SPSS, the statistical program

used. The authors have used an email–marketing service called MailChimp to send out mass

emails, with this tool the distribution of the surveys was made more efficient than it would

have been, if it had been send out to each email, manually.

3.2.2 Questionnaire Design

Surveys are cheap and quick to administer, and the interviewer does not affect the answers

with their presence, no variation of the questions occur and they can be made when the

respondent has the time for it Bryman and Bell (2007). Bryman and Bell (2007) also pointed

out few disadvantages with surveys; Misunderstandings may occur, regarding what the

researchers are asking about, since the participants cannot ask the researcher directly.

Participants cannot ask follow-up questions, and it is difficult to ask many questions since

researchers risk losing the respondents’ will of participating.

23

Chinese translation of the survey was done with the help of a Chinese friend, later the

language and simplicity of the language was reassured by two other Chinese individuals who

were profound in the language. The survey consisted of 16 questions with the simplest

language possible. The survey had mostly closed question. To study the psychographic and

behavioural segmentation the authors used likert scales.

3.3 Operationalization

Questions

Theoretical

Reference

Concept

Motivation

1. Where do you

live?

- East Coast

- Central China

- North East China

- Western China

(Beane and Ennis 1987);

(Kotler and Armstrong

2011); (Gaston-Breton

and Martín, 2011);

(Douglas and Craig,

1982)

Geographic

segmentation

and Macro

Segmentation

(control

variable)

Helps segmenting the

sample geographically.

Controls the respondent

to be only from China.

2. What size is the

city you live in?

-Municipality of

China

- Sub provincial

level city

- Sub prefecture

level city

- Town, village

(Beane and Ennis 1987)

& (Kotler and Armstrong

2011); (Gaston-Breton

and Martín, 2011);

(Douglas and Craig,

1982)

Geographic

segmentation

and Macro

Segmentation

Refine the first question

on geographical

segmentation

3. How old are you?

(age groups)

16-19

Beane and Ennis (1987) ;

(Cleveland, M. et al,

2011); (Wedel and

Demographic

segmentation

and Macro

Helps segmenting the

sample by age

24

20-34

35-49

50-64

65 and above

Kamakura, 1998);

(Gaston-Breton and

Martín, 2011); (Douglas

and Craig, 1982)

Segmentation

(control

variable)

control variable to base

the analysis on and make

sense of the other results.

4. Gender:

- Male

- Female

Beane and Ennis (1987) ;

(Cleveland, M. et al,

2011); (Wedel and

Kamakura, 1998);

(Gaston-Breton and

Martín, 2011); (Douglas

and Craig, 1982)

Demographic

segmentation

and Macro

Segmentation

Tool to segment the

sample by gender

5. Annual Income

- 30 000 to 46 000

- 46 000 to 87 000

- 87 000 to 120 000

- 120 000 to 163

000

- above

Beane and Ennis (1987) ;

(Cleveland, M. et al,

2011); (Wedel and

Kamakura, 1998)

Demographic

segmentation

and Macro

Segmentation

This question is giving

an insight on the

revenues available for

the sample

.

6. What factor do

you think is the most

important for most

of the people:

- Safety

- Brand image

- Design

- Fuel efficiency

- Quality

- Technology

- Affordability

- Comfort

- Chinese brand

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Psychographic

segmentation

and Micro

Segmentation

25

7. How many cars

have you bought?

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Psychographic

segmentation

and Micro

Segmentation

For the experience?

8. Do you have a car

now?

Micro and

behavioural

segmentation)

This question is more of

a preparatory question to

introduce the other

psychographic and

behavioural questions

8.1.1 How often do

you use your car?

- every day

-only on weekends

- less than once a

week

(Beane and Ennis 1987);

(Lancaster and Williams,

2002); (Kim and Lee,

2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Behavioural

and Micro

segmentation

To determine the

frequency of use of the

car

8.1.2 You mainly

use your car for:

- long trips

- short trips

(Beane and Ennis 1987);

(Lancaster and Williams,

2002); (Kim and Lee,

2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Behavioural

and Micro

segmentation

To determine the sort,

type or nature of use of

the car

8.1.3 How often do

you change your

(Beane and Ennis 1987);

(Lancaster and Williams,

Behavioural

and Micro

26

car?

- every year

- every 2 years

- in between 2 and 5

years

- less than once

every 5 years

2002); (Kim and Lee,

2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

segmentation

car consumers in China.

8.1.4 If you have

ever owned a car, to

what degree were

the following items

important when

purchasing it?

- Safety

- Brand image

- Design

- Fuel efficiency

- Quality

- Technology

- Affordability

- Comfort

- Chinese brand

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Psychographic

segmentation

and Micro

Segmentation

Understanding the

motivation for the last

car purchase

8.1.4.1 Would you

buy your next car

for the same reason?

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Behavioural

and Micro

segmentation

This questions the

loyalty of the respondent

and the possibility that a

purchasing factor lasts in

time. It gives an

27

8.1.4.1.1 If you

would buy a

different car, what

attribute would you

consider the most

important for the

next purchase?

- Safety

- Brand image

- Design

- Fuel efficiency

- Quality

- Technology

- Affordability

- Comfort

- Chinese brand

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Psychographic

segmentation

and Micro

Segmentation

The question investigates

on the possible change in

the purchasing reasons

for car owners. It gives

an indication for

dynamic segmentation.

8.2 If you do not

own a car, do you

plan to buy one in

the future?

Behavioural

segmentation

and Micro

Segmentation

This question prepares to

the next one and helps

assessing the potential of

the demand.

8.2.1 If you plan to

buy a car, to what

degree would the

following items be

the most important

to you?

- Safety

- Brand image

- Design

- Fuel efficiency

- Quality

- Technology

(Beane and Ennis 1987);

(Holbrook and

Hirschman 1982);

(Wells, 1975); (Kim and

Lee, 2011); (Steenkamp,

2001); (Gaston-Breton

and Martin, 2011);

(Laroche, 2011)

Psychographic

segmentation

and Micro

Segmentation

With this question, the

current non owner of a

vehicle is taken in the

equation, and it

28

- Affordability

- Comfort

- Chinese brand

3.4 Data analysis method

The survey’s data was analysed with the use of the statistical computer PASW Statistics 18

from SPSS Inc. According to Bryman and Bell (2005), the process of deciding how to

analyse the data should be done while gathering it and not after the collection. The choice of

the kind of analyse and the variable choice are crucial. The authors chose the linear

regression analysis, with curve estimates in order to predict sets of variables for behaviours

and psychographics relying on the macro factors included in Demographics and Geographics.

The statistical significance is fixed at 10%, and all results found over this level will be

rejected.

3.5 Validity

Validity could be seen as the most important research criteria, validity is concerned with the

integrity of the conclusions which is generated from performing a research (Bryman and Bell

2007). Bryman and Bell (2007) categorises different types of validity which are typically

used for quantitative research:



Measurement validity – is also referred to as construct validity. It essentially deals

with the fact, whether or not a measure that is formulated of a concept really does

reflect the concept that it is supposed to be representing. It is important to find

suitable measurements of terms which are stated in the research. Otherwise the

construct validity will be poor and the results questionable.

The authors used simple regression analysis to measure significant relations between

variables. Managerial decisions are often based on these (Anderson et al, 2002). The

questions that were constructed are linked to the theories, as shown in the

operationalization. The questions have been also translated in simple Chinese as the

focus is only on the Chinese car market.

29

can be seen that it is one specific variable that affects another and makes it to vary,

and that it is not a third variable causing the effect.

This study is based on a single survey, and the relations of the variables are based on

that survey. With time the variations in the independent variable cannot be predicted.

So the internal validity is low.



External validity – relates to transferability. It stresses whether the results of a study

can be generalized beyond the context at hand for the researcher.

The study is focussed only on China car market which gives it external validity.



Ecological validity - stresses whether the results of a study can be applied in people’s

everyday life and their natural social context. The ecological validity might become

poor when using surveys, because they disturb the natural situation that the

participants are in.

As this study was based on an email survey, and the presence of the surveyor was

absent, the natural social context was not disturbed so the authors believe the study

consists of ecological validity.

3.6 Reliability and Replicability

Reliability is defined as whether a research would provide the same results if the research

would be done again, with other researchers but with the same circumstances as the original

researchers; that there were no random or temporary conditions in the original research. If the

measurement of the study is not reliable enough this may influence the results differently and

the results may be not reliable. Replicability can be described as, if the investigation of the

study is reproduced by another researcher to study the reliability of the original study.

Detailed explanation of the procedure of the original study must be provided. (Bryman &

Bell, 2007).

31

4. Empirical Data

In this chapter the authors have used both descriptive statistics and simple linear regression to

estimate relationship between the variables.

Managerial decisions are often based on this relation. In regression terminology the variables

being predicted are called dependent variable. The variable or variables being used to predict

the dependent variable is called independent variable. Only the relations with significance has

been used in the study; the significant value has been set to sig<0.10. (Anderson et al, 2002)

All the regression tables have been presented in appendix B.

4.1 Geographic data:

When asked their residence region in China, the respondents mainly indicated to be living in

Eastern China. The other regions had far less represented but nevertheless equally mentioned

at around 10% each, whereas the people living in Eastern China amount 72.4% of the

responses. The region of residence is also a control variable in the study. (See Appendix

A.1.1)

The urbanisation rate is really high since it can be observed that 63.2% live in a Municipality

of China. Sub provincial level city, sub prefecture of China and village inhabitants

respectively amount for 11.49%, 17.24% and 8.05%. (See appendix A.1.2)

4.2 Demographic data:

In this part of the questionnaire, people are asked basic information about themselves: their

gender, age and annual incomes.

The results show that the survey reached both male and female respondents in a balanced

way. Women are slightly less represented with 45.98% of the responses whereas men

represent 54.02% of the sample. (See appendix A.2.1)

32

that is above 65 years old shows no entry. Age is also the second control variable in the

study.

Concerning the question about incomes, the majority of respondents are earning very low

incomes for more than half of them: 34.48% have no incomes at all and 20.69% are in the

first category (10 000 to 30 000 Yuan a year). The other income categories are dispatched at

around 7% of the responses, except for the highest earners who responded for11.49% as well

as the 3

highest earners (87 000 to 120 000 Yuan a year). (See appendix A.2.3.)

4.3 Behavioural data:

For the questions regarding consumer behaviour, the sample had to respond according to their

consumption habits, and answered questions in order to know if they currently own a car or

not, their level of experience with the purchase of cars, the frequency and style of use, their

replacement rate and their loyalty. For non-car owners, the respondents had to indicate if they

were planning to buy a car.

In the sample, most of the answers show that people do not have a car yet (64.37%), but that

they are very likely to buy one in the future: 93.59% of the respondents are thinking of

buying one. (See appendixes A.3.1. and A.3.2.)

The persons who answered the survey have not purchased many cars in their lives yet, and

can be considered as non-experienced as car purchasers with all responses showing that they

have bought less than four. Half of the sample has never purchased any car and 32.18% only

one. 10.34% have purchased two cars and 3.45% of respondents have bought three cars. (See

appendix A.3.3.)

For the car owners, the consumption style is clearly in favour of short daily trips. (See

appendixes A.3.4. and A.3.5.)

33

67.92% of respondents mention that they would use their car a long time and buy a new one

after more than six years, and 22.58% in between 2 and 6 years. People replacing their car

within a short period of less than 2 years represent only 6.46% of the sample. (See appendix

A.3.7)

4.3.1 Customer behaviour explained by Geographic data

This part is dedicated to the following relation of the model:

Figure 2: Customer behaviour explained by Geographic data

Customer behaviour explained by region of residence (See appendix part B.1.1):

Experience is not significant (0.181), neither is frequency of use (0.173), long trips (0.558),

car replacement rate (0.709), and loyalty to attribute (0.

586).

The link between region of residence and short trips has an acceptable significance (0.093).

Customer behaviour explained by the degree of urbanization (See appendix part B.1.2):

34

Experience is not significant (0.784) and neither frequency of use (0.293), long trips (0.97),

short trips (0.756), replacement rate (0.836) nor loyalty to attribute (0.776) has acceptable

significances.

4.3.2 Customer behaviour explained by Demographics:

Figure 3: Customer behaviour explained by Demographics

Consumer behaviour explained by age (see appendix part B.1.3.):

The statistical significances for the frequency of use (0.55), long trips (0.92), short trips

(0.615), replacement rate (0.937), and loyalty (0.442) are showing no clear relations.

Nevertheless, the experience of a driver appears significant (0.008).

Customer behaviour explained by the gender of the respondent (see appendix part

B.1.4.):

Experience is not significant (0.721), and nor are the frequency of use (0.945), long trips

(0.218), short trips (0.685) and replacement rate (0.101).

35

Customer behaviour explained by incomes level (see appendix part B.1.5.):

Again, many behavioural variables cannot be explained by the income level due to

insufficient significance level: experience (0.195), long trips (0.29), short trips (0.14),

replacement rate (0.364), loyalty (0.391), are all over our maximum of 10%.

On another hand, the frequency of use is significant (0.039).

4.4 Psychographic data:

Figure 4: Customer psychographics explained by both macro data and customer behaviour

In this part, the psychographic side of the customer is explained by both the macro

dimensions that are demographics and geographic variables, and by the other micro

dimension that is customer behaviour.

36

4.4.2 Respondent’s opinion on global preferences

The first question is about what the respondent thinks about the global preferences in China,

and the results show that safety (4.7/5 average) and quality (4.6/5 average) are rated as really

important by most of them. Then come fuel efficiency (4.1/5 average), comfort (4/5 average)

and technology (4/5 average) which are still very important but less than the two first

attributes. The design (3.8/5 average), affordability (3.7/5 average) and brand image (3.4/5

average) are rated averagely and finally, the fact that the car is a Chinese brand (2.1/5

average) is rated as non-important. (See appendix A4)

Respondents’ opinion on others’ preferences by age of the respondents (see appendixes

B.2.1.1 to B.2.1.9):

Seven out of nine psychographic variables cannot be explained by the age of the respondent:

safety (0.373), brand image neither (0.568), fuel efficiency neither (0.742), quality neither

(0.288), technology neither (0.833), affordability neither (0.628), Chinese brand (0.707).



Design and comfort’s importance perception for the global market are the only

significant relations (respectively 0.006 and 0.014).

Respondents’ opinion on others’ preferences by gender of the respondent (see appendix

part B.2.2.1 to B.2.2.9):

The results fail to give an acceptable explanation of what the respondents think of others’

preferences according to their gender due to statistical significances higher than the limit

fixed by the authors: safety (0.297), brand image (0.135), design (0.346), fuel efficiency

(0.564), quality (0.839), technology (0.731), affordability (0.208), comfort (0.313), Chinese

brand (0.898).

Respondents’ opinion on others’ preferences by annual incomes (see appendixes B.2.3.1

to B.2.3.9):

The statistical significance for each variables being too high, the results could not be

considered valid and the respondents’ opinion on others’ preferences cannot be correctly

explained relying on the annual incomes of the respondents:

37

Respondents’ opinion on others’ preferences by region of residence (see appendixes

B.2.4.1 to B.2.4.9):

The same happens with people’s opinion on others’ preferences according to where they live

in China:

Safety shows no acceptable significance (0.352), nor brand image (0.502), design (0.194),

fuel efficiency (0.297), quality (0.268), technology (0.453), affordability (0.952), comfort

(0.551), and Chinese brand (0.555).

Respondents’ opinion on others’ preferences by degree of Urbanisation (see appendixes

B.2.5.1 to B.2.5.9):

All the psychographic variables cannot be related to the degree of urbanization due to

non-acceptable statistical significances:

Safety (0.237), brand image (0.786), design (0.114), fuel efficiency (0.433), quality (0.131),

technology (0.149), affordability (0.918), comfort (0.103).



Only one psychographic variable, Chinese brand (significance = 0.060), can be

explained by the degree of urbanisation of the respondent.

4.4.3 Car owners’ current preferences

4.4.3.1 Car owners’ current preferences explained by demographics

For car owners, safety (4.7/5 average) and quality (4.6/5 average) are also the most important

reasons of purchase. Fuel efficiency (4.1/5 average) and comfort (3.9/5 average) appear to

have been important attribute for their last purchase as well. Technology (3.8/5 average),

design (3.7/5 average) and affordability (3.7/5 average) represent an average attribute for

them, which are rated as slightly more important than brand image. Chinese brand (2.4/5

average), once again, cannot be considered as a valid purchase reasons when looking at the

importance granted by the respondents. (See appendix A4)

Car owners’ preferences for their current car according to the age (see appendixes

B.2.1.10 to B.2.1.18):

The other variables are not exploitable:

38



Design is an exploitable variable with a significance of 0.082.

Car owners’ preferences for their current car according to the gender (see appendixes

B.2.2.10 to B.2.2.18):

Safety is not significant (0.457) and cannot be treated, the same happens with the other

psychographic variables: brand image (0.862), design (0.772), fuel efficiency (0.165), quality

(0.712), technology (0.375), comfort (0.171), Chinese brand (0.610).



The gender though, explains the perception of affordability for car owners

(significance for affordability is 0.084).

Car owners’ preferences for their current car according to their annual incomes (see

appendixes B.2.3.10 to B.2.3.18):

None of the psychographic variables can be explained with annual incomes for car owners:

Safety is not significant (0.253), and neither are brand image (0.233), design (0.972), fuel

efficiency (0.784), quality (0.253), technology (0.652), affordability (0.495), comfort (0.393),

and Chinese brand (0.966).

4.4.3.2 Car owners’ current preferences explained by geographic variables

Car owners’ preferences for their current car according to region of residence (see

appendixes B.2.4.10 to B.2.4.18):

The relation regarding car owners’ perception of safety and their region of residence is not

significant (0.250) and neither it is with their perception of brand image (0.347), the design

(0.148), the fuel efficiency (0.519), the technology (0.254), the affordability (0.667), the

comfort (0.281), or the fact a car is from a Chinese brand (0.394).



The relation regarding car owners’ perception of quality and the region of residence is

significant (0.020).

Car owners’ preferences for their current car according to degree of urbanisation(see

appendixes B.2.5.10 to B.2.5.18):

39



Chinese brand is significant (0.090) and is the only significant relation with the degree

of urbanisation of car owners

4.4.3.3 Car owners’ current preferences explained by behaviours

Car owners’ preferences for their current car according to the frequency of use (see

appendixes B.3.3.10 to B.3.3.18):

Most of the variables cannot be used since they do not reach the 10% limit for statistical

significance that the authors imposed:

Safety is not significant (0.413), brand image (.568), design (0.897), fuel efficiency (0.393),

technology (.32), comfort (0.641), Chinese brand (0.641).



But a relation in between affordability importance and the frequency of use by a car

owner can be observed (significance is 0.096), as well as a relation with the

importance of quality (significance is 0.021).

Car owners’ preferences for their current car according to the style of use: Long trips

(see appendixes B.3.5.10 to B.3.5.18):

Safety (0.869), brand image (0.27), design (0.208), fuel efficiency (0.795), quality (0.316),

comfort (0.152), and Chinese brand (0.648) cannot be explained by the behaviour variable

“Style of use: long trips” since their significance does not meet the standard of 10%.



Nevertheless, technology is significant (0.01) and can be related to long trip drivers.

Affordability is significant too (0.012).

Car owners’ preferences for their current car according to the style of use: Short trips

(see appendixes B.3.6.10 to B.3.6.18):

Only one variable is having a statistical significance meeting the 10% standard, all the others

are higher:

Safety (0.521), brand image (0.481), design (0.5), fuel efficiency (0.823), technology (0.347),

affordability (0.125), comfort (0.547), Chinese brand (0.415).



This significant relation shows a link in between quality perception for short trips

drivers (significance=0.022).

40

The majority of variables having higher statistical significance are therefore not highlighting

any acceptable relation:

Safety (0.323), design (0.439), fuel efficiency (0.898), quality (0.807), technology (0.236),

affordability (0.258), comfort (0.372), Chinese brand (0.396).



Brand image is the only significant variable (0.039) enabling the observation of a

relation between its importance and the replacement rate of car owners.

4.4.4 Car owners’ preferences for their next purchase

When asked what importance they would grant to the different attributes for the next

purchase, car owners answered that, again, quality (4.5/5 average) and safety (4.7/5 average)

would be the most determining criteria. Then come comfort (4.1/5 average), fuel efficiency

(4/5 average) and technology (4.1/5 average). Affordability (3.8/5 average), design (3.8/5

average) and brand image (3.6/5 average) still appear important but less than the first

attributes. Chinese brand is not judged as important (2.6/5 average). (See appendix A4)

4.4.4.1 Car owners’ preferences for their next purchase explained by

Demographics

Car owners’ preferences for their next car according to their age (see appendixes

B.2.1.19 to B.2.1.27):

Most of the variables cannot be related to the age of the respondent because of the

significance level:

Safety is not significant (0.40), brand image (0.147), design (0.148), quality (0.191),

technology (0.835), affordability (0.508), comfort (0.483), Chinese brand (0.766).



Age is a valid variable to explain the car owners’ preferences for the next purchase

(significance = 0.086).

Car owners’ preferences for their next car according to their gender (see appendixes

B.2.2.19 to B.2.2.27):

None of the psychographic variables could be effectively linked to gender for the next

purchase by a car owner, statistical significances not being satisfactory:

41

Car owners’ preferences for their next car according to their annual incomes (see

appendixes B.2.3.19 to B.2.3.27):

Eight variables could not be studied more since the significance is over the 10% rate:

Brand image (0.446), design (0.715), fuel efficiency (0.598), quality (0.36), technology

(0.868), affordability (0.808), comfort (0.83), Chinese brand (0.705).



Only safety is meeting the 10% significance rate chosen by the authors (0.082). A

relation can thus be observed.

4.4.4.2 Car owners’ preferences for their next purchase explained by

Geographic variables

Car owners’ preferences for their next car according to their region of residence (see

appendixes B.2.4.19 to B.2.4.27):

There are still seven variables from which no relation can be drawn due to insufficient

statistical significance:

Safety (0.182), design (0.194), fuel efficiency (0.414), quality (0.502), technology (0.132),

affordability (0.929), comfort (0.464).



The results from the survey indicate that there is a relation in between the region of

residence of a Chinese person and what how he would rate Brand Image and Chinese

Brand attributes for his/her next purchase. Brand image significance is 0.06 and

Chinese brand significance is 0.01.

Car owners’ preferences for their next car according to their urbanization degree (see

appendixes B.2.5.19 to B.2.5.27):

Most of the variables do not meet the significance level of 10% and cannot be studied

meaningfully:

Safety (0.726), brand image (0.91), design (0.952), fuel efficiency (0.627), technology

(0.424), affordability (0.717), comfort (0.646), Chinese brand (0.632).



The urbanization degree of a respondent influences his/her preference level for quality

(significance = 0.052) when thinking about their next car.