The influence of social media marketing on
the behavior of consumers purchasing
cosmetic product
— A comparative study of China and Sweden
Author：Lu Zhouyan
Yu Yingpei
Master’s Thesis 15 credits
Department of Business Studies
Uppsala University
Spring Semester of 2020
Abstract
The purpose of this study was to investigate the influence of social media marketing on the behavior of consumers purchasing cosmetic products and discover the behavioral differences in Chinese and Swedish consumers. The authors propose five factors—social media marketing (SMM), trust, risk, word of mouth (WoM), and perceived ease of use (EoU)—and hypothesize their relationships with purchasing behavior (PB). Based on a review of previous research on consumer behavior towards social media, a theoretical framework was created, which became the foundation for a survey. A total of 251 questionnaires were completed online, with thesame questions being asked of consumers of cosmetics in both Sweden and China . While the generalizability of the survey has some limitations regarding sampling, this thesis provides some key findings after analyzing data with SPSS. It was found that consumers’ PB in Sweden and China is affected by SMM. Trust, WoM, and EoU can regulate PB in both countries, while risk can only regulate the behavior of Swedish consumers.
These insights can help cosmetic marketers to develop better growth strategies to enter the Chinese and Swedish markets and provide a better understanding of how different factors influence purchasing behavior in the social media environment. Though SMM is increasingly popular, the empirical research on cross-cultural consumers’ purchasing behavior in the digital environment is still in the preliminary stage. The results also provide some instruction meaning for further study.
Table of Contents
1. Introduction... 1
1.1 Social media marketing and promotion of cosmetic products...2
1.2 Behavioral differences in consumers from different countries...2
1.3 Research question... 3
1.4 Structure of the thesis...3
2. Literature review... 4
2.1 Social media marketing...4
2.2 Purchasing behavior...6
2.3 Different cultural traits in Sweden and China...7
2.4 SMM and PB in different countries... 8
2.5 Moderating factors... 9
2.5.1 Trust...9
2.5.2 Risk...10
2.5.3 Word of mouth...10
2.5.4 Perceived ease of use...11
2.6 Theory framework...12
3. Research design & data collection method... 13
3.1 Deductive research...13
3.1.1 Comparative design... 13
3.1.2 Quantitative method...13
3.1.3 Variables and measurements... 14
3.3.3 Reliability analysis...21
4. Quantitative empirical findings and analysis... 23
4.1 Correlation analysis...23
4.1.1 Bivariate correlation... 23
4.1.2 Partial correlation...23
4.2 Standard multiple regression...24
4.2.1 The regression analysis of hypothesis 1: SMM and PB... 24
4.2.2 The regression analysis of hypothesis 2: SMM, trust, and PB... 25
4.2.3 The regression analysis of hypothesis 3: SMM, risk, and PB... 26
4.2.4 The regression analysis of hypothesis 4: SMM, WoM, and PB...27
4.2.5 The regression analysis of hypothesis 5: SMM, EoU, and PB...28
4.3 Chapter Summary... 29
5. Discussion... 30
5.1 Basic data discussion and comparison...30
5.2 Discussion of research results...30
5.2.1 Trust...31
5.2.2 Risk...32
5.2.3 Word of mouth...33
5.2.4 Perceived ease of use...34
6. Conclusion...36
6.1 General conclusion...36
6.2 Implications...37
6.3 Limitations and recommendations for future study... 37
Reference...39
Appendix... 49
1. Introduction
There were 3.5 billion social media users worldwide in 2019, and that number is still growing (Emarsys, 2019). Additionally, people are spending more time on social media, and one study shows that an average of three hours is spent per day per person on social media (GlobalWebIndex, 2019). Because of this trend, social media has become extremely relevant for online marketing, as statistics also show that 54% of social media users not only research but also receive product information nowadays (GlobalWebIndex, 2018).
Social media marketing refers to the use of social media platforms like Facebook, Instagram, YouTube, WeChat and Weibo to persuade consumers that an organization, item, or solution is worthwhile (Lazer & Kelley, 1973). Social media marketing has now been developed into a prevalent instrument when it comes to effective marketing. Many young people, such as Kendall Jenner, who publishes cooperate advertisements for Tod’s, are digital role models. In China, WeChat is one of the most popular social media platforms, and it actively pushes advertisements, like Dior bags and so on, to feed females’ needs. The former uses the model’s influence to create a trend, while the latter implements customized advertisements according to the characteristics of customers. The relevance of social media for online marketing is increasing by the day based on the current statistics.
1.1 Social media marketing and promotion of cosmetic products
Brands have increasingly taken to social media to promote products to consumers. In 2018, there were 3.7 billion brand-sponsored posts on Instagram and this figure is projected to surpass 6 billion posts in 2020 (Statista, 2019). Among them, the cosmetics industry is one of the most engaging, and they have put influencer content center stage (Alexandria, 2018). Makeup brand owner, Kylie Jenner, who has 178 million followers on Instagram, launched her cosmetics brand solely on social media. Also, L’Oreal is one of the best examples of a well-known cosmetics brand that has had a few well-documented campaigns to boost brand awareness, and social media has been central to the spread of its positive message (Valentine, 2019).
According to Statista (2019), nearly 90% of respondents claim they would be interested in buying certain cosmetic products after seeing these the promotions for these products online. In addition, 73% of marketers in the cosmetic industry also believe the efforts and resources they put into marketing products through social media will have a great influence on the business (Buffer, 2019). The influence of social media marketing on consumer purchasing behavior is set to grow in the future. 1.2 Behavioral differences in consumers from different countries
While social media is growing in popularity in many countries and it is connected, consumers may have different perceptions of social media marketing and may behave differently. Goodrich and de Mooij (2013) compared the use of social media and traditional media across 50 countries and found that culture has a significant influence on purchasing behavior.
but this concern does not seem to be shared by Chinese consumers, who are very likely to use social media at every step of their buying journey.
Consumers from different countries may act differently due to cultural differences (Keisidou et al., 2011). Consumers are imprinted by their cultures with a set of values that influence their behavior (Kotler & Keller, 2006). Existing studies comparing consumers’ purchasing behavior generally focus on traditional media (Ibrahim, Aljarah, & Ababneh, 2020; Milwood, Marchiori, & Zach, 2013). There is less understanding of consumers from different countries and how they behave regarding the influence of social media marketing. Since the development of social media and online environments is relatively fast and new, more research is needed to investigate consumer purchasing behavior and SMM in different countries.
1.3 Research question
In this study, we examine the influence of social media marketing on customers’ purchasing behavior for cosmetic products. We also investigate consumers from different countries and try to understand their behavioral differences. We focus on Chinese and Swedish consumers, and we conducted a questionnaire to collect data. Our research question asks:
How do Chinese and Swedish consumers behave differently (or similarly) toward social media marketing when purchasing cosmetic products?
1.4 Structure of the thesis
2. Literature review
2.1 Social media marketing
Social media is defined as an online application built on the ideological and technical foundations of Web 2.0 that enables communication between users (Kaplan & Haenlein, 2010). Social media includes online applications, platforms and media that aim to facilitate interactions, collaborations, and the sharing of content (Richter & Koch, 2007). It is divided into a variety of forms, including weblogs, social blogs, microblogs, podcasts, pictures, text, videos, etc. As different forms continue to appear and become popular, the number of users is booming; not only existing social networkers but also businesses are joining and using these platforms as communication tools. Kim and Ko (2010) indicate that social media can have a dramatic impact on a brand’s reputation. Also, DEI Worldwide (2008) states that a company will miss an opportunity to reach customers if they do not engage in social media as part of their online marketing strategy. Customer-based characteristics show that companies should use social media as a marketing communication medium (Hootsuite, 2018; Yadav & Rahman, 2017).
Social media marketing (SMM) refers to the promotion of products, brands, or organizations by interacting on social media with current or prospective consumers, and it is intended to persuade them to change their behavior (Saravanakumar & SuganthaLakshmi, 2012). SMM is primarily internet-based but has similarities with non-internet-based marketing methods like community marketing. The social aspect of SMM can be compared to community marketing. It aims to promote “information, ideas, and methods to enhance social and economic ends” (Lazer and Kell, 1973). In order words, SMM is the use of social media to create value for stakeholders and achieve organizational goals (Reto, Philipp, & Hinsch, 2017).
their fans, the brand engages its target customers more broadly. Once consumers are aware and engaged, they are in a position to communicate their opinions to other consumers. Satisfied and loyal consumers communicate their positive attitudes toward the social application created by the company and use the social media application to influence their friends or fans, which is referred to as WoM.
The scholarly understanding of SMM, developed in Kim and Ko’s (2012) work, proposes five activities for SMM: entertainment, interaction, trendiness, customization, and WoM. This original model explains how SMM can enhance client assets in the luxury fashion brand space. According to Muntinga, Moorman, and Smit (2011), entertainment is considered a strong motivator for users to engage in social media and produce user-generated content (UGC). Within the social media context, Godey et al. (2016) describe interaction as the discussion and conversation between users and other participants. According to Muntinga, Moorman and Smit (2011), trendiness indicates the ability of social media to spread trendy information. Schmenner (1986) defines the level of customization as the extent to which tailored services or information can fulfill the individual’s preferences. Muntinga, Moorman and Smit (2011) assert that WoM is linked with customer-to-customer communication. In evaluating SMM’s impact on the luxury industry, purchase intention is positively influenced by entertainment, interaction, and WoM (Kim & Ko, 2010).
2.2 Purchasing behavior
There have been many previous intensive studies on consumer purchasing behavior (PB) (Blackwell et al., 2006; Kuester 2012), and the field covers a lot of areas (Hasslinger et al., 2008; Solomon et al., 2010; Solomon, 2010). According to Kuester (2012), PB is “the study of individuals, groups, or organizations and the processes they use to select, secure, use, and dispose of products, services, experiences, or ideas to satisfy needs and the impacts that these processes have on the consumer and society.” In addition, PB is regarded as the purchasing of a particular product or service that has been affected by customers’ perceptions, attitudes, and satisfaction. PB is considered an essential factor for enterprises to find, retain, and pursue consumers in the competitive environment (Morrison, 1979; Spreng, MacKenzie, & Olshavsky, 1996; Taylor & Baker, 1994). In the definition of Internet consumption, Goldsmith and Bridges (2000) defined PB as “gathering information passively via exposure to advertising; shopping, which includes both browsing and deliberate information search, and the selection and buying of specific goods, services, and information.”
Ajzen and Fishbein (1977) indicate in their attitude model that PB is predominantly determined by intention. According to Hasslinger et al. (2007), consumer PB and intention are inextricable. Also, many studies show that PB is closely related to the attitude and preference towards brands or products (Kim, & Johnson, 2010; Kim & Ko, 2010; Kim & Lee, 2009). Komiak and Benbasat (2006) further developed the belief-attitude-intention framework and proposed a trust model of electronic commerce adoption. The belief-attitude-intention framework, which relates to belief, attitude, and behavioral intention, is widely used in the study of PB. This framework suggests that the attitude toward a particular object depends on the direct effects of beliefs about the object, while attitude has a direct positive impact on behavioral intention toward the object. In most circumstances, prospective consumers usually get information regarding purchasing and the product they are interested in before they take action.
Hence, we define PB as a combination of consumers’ interest in products and the possibility of purchasing. As discussed before, belief, attitude, and behavioral intention are important components of PB, which is reflected in our questionnaire measurement.
2.3 Different cultural traits in Sweden and China
Cultural characteristics have the broadest and deepest influence on the consumer’s PB (Kotler & Keller, 2006; Zhao, 2012). Culture is regarded as the most basic cause of consumer wants and needs (Hasslinger et al., 2007). As people grow up, they are imprinted with a set of values that affects how they behave (Kotler & Keller, 2006). Culture consists of smaller subcultures, which contain more specific traits that affect PB (Keisidou et al., 2011). Hofstede conducted extensive research on the differences between cultures in 76 countries (Hofstede, 2001). A four-dimensional model was developed of cross-cultural work-related values; this consists of individualism-collectivism, power distance, masculinity-femininity, and uncertainty avoidance (Hofstede, 1983).
scores low in individualism, equal to 20, in comparison with Sweden, with a score of 71 (Hofstede, 2001). Power distance refers to the way that a society deals with inequality. China’s score of 80 demonstrates the inequality of power and wealth in the country. Cultural heritage, along with the history of political control, affects the power distance status of China. Sweden’s score of 31 affirms its low power distance, with its decentralized power and equal opportunities. Communication in social media is direct and informal (Hofstede, 2001). Masculinity-femininity refers to the extent of role divisions between genders. China, with a score of 66, is influenced by high masculinity and is success-oriented, stressing role division and financial achievement. Sweden scores 5 and has a more feminine culture (Hofstede, 2001; 2010). The uncertainty avoidance index deals with the extent of the uncertainty and ambiguity that a society can tolerate (Hofstede, 1980). Sweden and China, with ranks of 29 and 30 respectively, have low scores for uncertainty avoidance. In Chinese culture, people are more sensitive about risk but are flexible based on real cases. In Swedish culture, people put more effort into handling risk only when it is necessary (Jandt, 2006). 2.4 SMM and PB in different countries
Elena, and Florian (2013) shed light on social media adoption in two major markets: US marketing organizations go “all in” to adopt social media tools, which appear to be fully supported by leadership, while Swedish marketing organizations tread lightly. Based on the broad application of social media in different markets and the correlation between social media and consumer behavior obtained by numerous scholars after verification, hypothesis 1 is established:
H1: SMM has a positive effect on consumer PB in both China and Sweden.
2.5 Moderating factors
Cultural, social, personal, and psychological factors also significantly influence a consumer’s buying behavior (Kotler & Keller, 2016). Based on the previous understanding of consumer behavior, this study uses four moderate variables: trust, WoM, risk, and perceived EoU. Based on the relationship between these four variables and consumers’ PB, four hypotheses were developed and are outlined below. 2.5.1 Trust
The definition of and research into trust involves many fields. In economics, it is believed that a rational person will generate trust only when he perceives that the potential gain is greater than the potential loss, and the trust gradually increases with the deepening of the relationship (Cole, 1998). In the field of marketing, trust refers to the expectation that consumers believe the company can realize its value proposition and promise, which is embodied in their attitude towards the company’s products and services (Sirdeshmukh & Sabol, 2002).
According to the characteristics of Chinese and Swedish consumers, we predict that collectivism will make it easier for Chinese consumers to have a sense of trust in SMM, and this sense of trust will have a greater influence on their willingness to purchase. Thus:
H2: The factor of trust has a greater influence on Chinese consumers’ PB.
2.5.2 Risk
The concept of risk was first introduced into the field of marketing by Bauer in 1960. Perceived risk is the potential loss caused by consumers’ decisions based on uncertain and unclear situations and must be borne by consumers (Bauer, 1960). With the emergence of online shopping and digital marketing, the shopping impulse means people face risks in the online market. Forsythe & Shi (2003) found that concerns about product performance, financial loss, privacy issues, and convenience collectively constitute the risks consumers face online (Forsythe, 2003).
Perceived risk has also been shown to have a moderating effect on consumer buying intentions. Consumers with higher perceived risks will only choose the most reputable and experienced companies, and they will not change easily (San Martín & Camarero, 2009). Individuals with lower perceived risks are more likely to choose a purchase channel based on satisfaction with previous interactions, rather than security and privacy terms (San Martín & Camarero, 2009).
We assume that individualism will guide Swedish consumers to consider more risk factors when shopping, so we make the following hypothesis:
H3: Consumers in the Swedish market pay more attention to risk factors when they purchase cosmetics.
2.5.3 Word of mouth
set between those who have already purchased the product and those who want to purchase the product (Helm & Schlei, 1998). In the Internet environment, WoM is further transformed into “Internet WoM.” Areas of the Internet devoted to communication and sharing, such as social media, are also considered as essential areas for word-of-mouth formation (Sun et al., 2006).
WoM is an important factor affecting consumer decision-making. Consumers are more reliant on WoM for purchasing decisions because WoM is often more reliable in consumers’ eyes than information sourced from companies (Chu & Kim, 2011). The development of the Internet and social media has accelerated the spread of WoM and amplified its role (Yan, 2012).
Combining observations and literature, we believe that the success of SMM in China can be largely attributed to the word-of-mouth effect, and the utility of WoM is amplified in a collectivist society, so we assume:
H4: Consumers in the Chinese market are more concerned about WoM. Chinese consumers are more inclined to purchase products with better WoM and easily give up products with lower WoM.
2.5.4 Perceived ease of use
Perceived usefulness and perceived ease of use (EoU) originate from the technology acceptance model (TAM). TAM shows that behavioral intentions determine consumer purchase decisions, and the crucial components of behavioral intentions and attitudes are perceived usefulness and perceived convenience (Davis, 1985). Later research combines the characteristics of the online market based on TAM. It gives a definition: Perceived EoU refers to the convenience of consumers in the purchasing process without leaving home, including the comparison of price, quality, and other related services provided by the website (Jing, Wang, & Zhou, 2005; Wang, 2014).
H5: Perceived EoU will have a more substantial impact on consumers in the Swedish market. Swedish consumers rely more on this aspect to determine whether to make a purchase.
2.6 Theory framework
In this study, the previous literature review shows that the five aspects of SMM and the four moderate factors generated by consumers can all influence PB. Therefore, this paper’s logical sequence starts with proving the positive relationship between SMM and PB. We further explore the individual effect of each moderate factor and its effect on consumer purchase decisions. Then, we compare the specific conditions of the Chinese and Swedish markets and summarize the performance of various factors in different markets to put forward the research conclusions.
3. Research design & data collection method
3.1 Deductive research
Taking into consideration the review of previous literature to create several hypotheses, which are tested in this study, it can be said this research is following a deductive approach (Bell & Bryman, 2007). Deduction means that we will work with an observable social reality, which leads to the production of credible data (Saunders et al., 2009). Another critical aspect of the deductive approach is that the research should be undertaken in a value-free way without the researchers being affected by the research (Saunders et al., 2009). This study began by reviewing the vast amount of already existing literature on consumer behavior in the digital context. The focus is on the effect that SMM and some moderate factors can have on consumers’ behavior towards cosmetics. Then the hypotheses were developed, which are an important part of deductive research.
3.1.1 Comparative design
In order to answer the research question, comparative studies are used to collect enough samples for statistical analysis to examine similar characteristics in China and Sweden. Comparative studies, sometimes called cross-cultural studies, in sociology, psychology, economics, and political science use field data from many societies to examine the scope of human behavior and test hypotheses about human behavior and culture (Annamoradnejad et al., 2019). These studies can be seen as belonging to different types: They are exploratory or test hypotheses, include or do not include contextual variables, or compare score levels obtained in different cultures (Van de Vijver, 2009). The current study is a comparative study; through previous research on the topic it was discovered that further investigations of cultural differences could be made, especially considering China and Sweden. These countries represent the emerging market and the developed market, respectively.
3.1.2 Quantitative method
should collect a large sample to ensure the generalizability of the results (Saunders, Lewis, & Thornhill, 2009; Bell & Bryman , 2007). According to Boutellier et al. (2013), quantitative research is widely used in economics, psychology, and demography. The objective of quantitative research is to conduct systematic empirical tests of observable phenomena, using models and statistical data (Given, 2008). According to Wright (2005), quantitative research allows researchers to easily and quickly summarize research findings by replicating them in different populations. This paper takes the PB of consumers in China and Sweden as the research object. The identification of the influence of the main factors requires many responses to understand the PB of consumers in these two countries. Therefore, quantitative research is an appropriate research strategy for this paper.
3.1.3 Variables and measurements
The content of the questionnaire is divided into three parts. The first part aims to understand the interviewee’s personal and demographic information. Some of the questions are control variables and were used as the basis for screening. The second part is used to prove that SMM can enhance consumers’ willingness to purchase, in order to confirm the authenticity of this phenomenon. The third part concerns the psychology and attitude of the respondents when interacting with SMM, which is the central part of this survey. We measured all factors on a five-point agreement scale (1 = strongly disagree, 5 = strongly agree), which is based on the Likert 5-level scale. The latter two parts are used for subsequent analysis and comparison. The detail for all construct items is shown in Table 1.
Table 1: Variables and measurements
SMM (Ashley & Tuten, 2015; Bernoff & Li, 2008; Bianchi & Andrews, 2015; Hoffman & Fodor,
2010; Schultz & Peltier, 2013; Kim & Ko, 2012)
SMM1: I feel fun to collect information on cosmetics through social media.
SMM2: I always killed time to collect information on cosmetics through social media. SMM3: It is a leading fashion to use social media following cosmetics.
SMM4: The cosmetics content found on social media is up to date.
SMM6: I am interested in the lively feed of cosmetic information on social media.
Trust (Jarvenpaa, Tractinsky, & Saarinen, 1999; McKnight, Choudhury, & Kacmar, 2002; Chen &
Dibb, 2010)
TR1: I believe the information received from social media can lead to good cosmetics. TR2: I think the cosmetics information received from social media is honest.
TR3: I think the cosmetics information received from social media is well intentioned.
Risk (Forsythe, 2003; San Martín & Camarero, 2009)
RI1: I think the quality/price of cosmetics promoted through social media is difficult to judge. RI2: I am worried about fraud from social media promotion of cosmetics.
RI3: I am worried about my personal privacy when browsing information on social media.
Word of mouth (Sun et al., 2005; Chu & Kim, 2011)
WoM1: I like to share the information on cosmetics I received with friends on social media. WoM2: I will be attracted by the information about cosmetics shared by friends on social media. WoM3: I will be attracted by cosmetics information shared by public figures on social media.
Perceived ease of use (Jing & Zhou, 2005; Wang, 2014)
EoU1: Information on cosmetics on social media is valuable for me.
EoU2: Receiving information on cosmetics from social media saves my time to shop.
EoU3: I found it is easier to shop after reviewing information on cosmetics through social media.
Purchasing behavior (Morrison, 1979; Spreng, MacKenzie, & Olshavsky, 1996; Taylor & Baker,
1994; Fishbein & Ajzen, 1975; Cheung, Chan, & Limayem, 2005)
PB1: I feel more interest in shopping for cosmetics after receiving information from social media. PB2: I prefer to shop for cosmetics after receiving information from social media.
3.2 Data collection 3.2.1 Primary data
The data was collected through online surveys of Swedish and Chinese social media users. The firsthand data collected through the questionnaire is the most important data of this research, and the subsequent research analysis is based on the collection of these data. Furthermore, the firsthand data collected has enabled us to understand consumers’ psychology and behavior more intuitively, so as to enhance the reliability and effectiveness of the research (Kate, 2007).
3.2.2 Choice of group
As mentioned earlier, the reason this study selected the cosmetics industry as the research object is that the SMM model has achieved widespread success in China for cosmetics, and more and more companies recognize this model as one of the mainstream marketing models for the future. In Sweden, this model is also in a stage of rapid development. For these reasons, we believe that cross-border comparison and research analysis of the cosmetics industry can intuitively depict the characteristics of SMM in different markets and its attraction to customers, which is of great significance to commercial companies and future research.
3.2.3 Choice of method
A questionnaire survey is an powerful method of gathering information. The distribution of questionnaires on the Internet can achieve the amount of data needed for research and help researchers reach more consumer groups and increase the breadth of research (Hoonakker & Carayon, 2009). Moreover, the data obtained from an online survey is explicit. This method makes it easier for researchers to collect feedback information and further improve the efficiency of data analysis and collation (Walliman, 2017).
Before distributing the questionnaire, we invited some consumers to conduct a question test. After the respondents filled out this initial questionnaire, we also conducted a separate interview to better understand their feelings about the questions they had answered. Based on this feedback, the logic and structure of the questionnaire were modified. This step ensured that the questions in the final questionnaire were easily understood by the interviewees, which increased the reliability of the survey as much as possible. The design of the questionnaire also considered factors such as culture, language, and way of thinking, and it was divided into Chinese and English versions. It is easier for Chinese consumers to understand and provide more realistic answers in a Chinese version. After observation, it was found that Swedish consumers have a good understanding of English and find it easy to use English to express problems clearly and avoid misunderstanding. Therefore, we distributed the English version of the questionnaire in Sweden.
3.3 Data analysis
variable and compared the adjusted R-squared of each equation to explain the regulating degree.
3.3.1 Descriptive analysis
Gender, age, and expenditure are conside red as relevant demographic characteristics in previous studies, so we asked about these in this study (Dehghani, 2018; Wu, 2016; Choi & Kim, 2016; Hsiao & Chen, 2018). Also, because our study is based on SMM and PB, we expanded to five control variables, considering how often the respondent uses social media and purchases cosmetics. We collected the distribution of each variable in this group of respondents to draw a sample profile.
The basic situations of respondents from both China and Sweden are shown in Table 2. Respondents were mainly between 18 and 35 years old. The proportion of men and women surveyed in both countries is the same. The monthly expenditure of respondents in China and Sweden is evenly distributed. The proportion of respondents spending under $350 a month is 42.1% in China and 35.5% in Sweden. The other most heavily weighted indicator is the monthly expenditure of more than $1,400, which is 24.5% in China and 29.8% in Sweden. Questions 4 and 5 revealed the daily behavior characteristics of respondents in social media and cosmetics purchasing. We found that more than 50% of Swedish and Chinese people use social media more than ten times a day; however, Swedes bought cosmetics more often than Chinese people, 71.2% of Swedish respondents buy cosmetics less than three months, compared with 46.2% in China. This result is related to the cultural differences between China and Sweden.
Table 2: Demographics
Gender Male 37.4 30.8 Female 62.6 69.2 Monthly expenditure $150 or less 12.2 24 $150 to $350 29.9 11.5 $350 to $700 19.7 13.5 $700 to $1400 13.6 21.2 Above $1400 24.5 29.8 Frequency of using
social media per day
3 times or less 8.3 12.5 4–6 times 20.7 23.1 7–9 times 12.4 13.5 Above 10 times 56.6 51 Frequency of purchasing cosmetics 1 week or less 5.5 10.6 1 week to 1 month 16.6 21.2 1–3 months 24.1 40.4 3–6 months 20.7 14.4 Above 6 months 33.1 13.5 3.3.2 Validity analysis
Validity determines the authenticity of the research results and whether we can accurately test our pre-constructed hypothesis. Increasing validity can enhance the value of research conclusions and research claims (Heale, 2015). In this study, we use the support of relevant theories to improve internal validity. Most of the questions in this study are from the consumer’s perspective and fit the customer’s real life. In this way, respondents’ misunderstandings are largely avoided, which improve research effectiveness. In addition, this study collected data from China and Sweden separately and covers the major consumer groups, which maximizes the effectiveness and accuracy of the study.
Swedish data to check the validity of the sample. In the EFA test, the Kaiser-Meyer-Olkin (KMO) test for sampling adequacy and Bartlett’s test of sphericity (BST) were applied to test whether the data were appropriate for factor analysis. Tabachnick and Fidell (2007) suggest that if the KMO is greater than 0.6 and the BTS is significant at α < .05 then the factor ability of the correlation matrix is assumed. When factor loading is significant and the coefficient value is higher than 0.4, we can say that the validity is good (Coakes & Steed, 2003; Hair et al., 2010).
We explored the factor analysis of the independent variable (SMM) and the dependent variable (PB) communalities and the rotated factor matrix. We found the value of customization 1 had the lowest communality value (0.428) and the entertainment 2 was cross-loading; thus we removed entertainment 2 and customization 1. Then we conducted the factor analysis again. All of the extraction values of communality were then over 0.5 and no cross-loading was shown anymore, as shown in Table 3.
Table 3: Factor analysis of SMM and PB
Factors Items Factor loading Communality Social media marketing entertainment 1 0.719 0.677
trendiness 1 0.746 0.616 trendiness 2 0.75 0.592 customization 2 0.598 0.662 Purchasing behavior PB1 0.693 0.735 PB2 0.833 0.774 PB3 0.787 0.653
solution with principal component analysis. As seen in Table 4, each factor performed well, with the high communality, and there was no longer any cross-loading.
Table 4: Factor analysis of MV and DV
Factors Items Factor loading Communality
Trust trust 2 0.723 0.812
trust 3 0.897 0.909
Risk risk 2 0.977 0.99
Word of mouth WoM2 0.823 0.891
WoM3 0.767 0.897
Ease of use EoU1 0.734 0.917 Purchasing behavior PB1 0.588 0.795
PB2 0.763 0.829
PB3 0.872 0.865
We made the analysis valid in terms of both content and structure. We used the support of relevant theories to improve content validity. Moreover, factor analysis was used to test the validity of the constructs. This indicates that the entire case is valid.
3.3.3 Reliability analysis
In addition, Cronbach’s alpha was used to obtain the reliability index of the data. For analysis, the reliability index range is between zero (α=0) to one (α=1). A high alpha value means higher reliability (Pallant, 2000). Also, a generally accepted rule is that α of 0.6–0.7 indicates an acceptable level of reliability, and 0.8 or greater is a very good level (Hulin, Netemeyer, & Cudeck, 2001). In this case, the Cronbach’s alphas of SMM, trust, risk, WoM, EoU, and PB are 0.880, 0.818, 0.695, 0.830, 0.868, and 0.867, respectively. After deleting the low communality items, the Cronbach’s alphas of SMM, trust, risk, WoM, EoU, and PB are 0.859, 0.797, 0.684, 0.827, 0.868, and 0.867, respectively, which are all higher than 0.6. Therefore, the reliability of this study is acceptable.
Table 5: Reliability analysis
Factor Cronbach’s alpha Cronbach’s alpha (item deleted) Social media marketing 0.880 0.859
Trust 0.818 0.797
Risk 0.695 0.684
4. Quantitative empirical findings and analysis 4.1 Correlation analysis
4.1.1 Bivariate correlation
By combining our previous EFA analysis, we got the new variables SMM, trust, risk, EoU, WoM, and PB by calculating the average of the remaining items. According to Pallant (2016), the correlated relationship between each independent variable and dependent variable needs be above .3, and correlation is significant at 0.01. The Pearson correlation between the independent and dependent variables is 0.661, and the coefficients between the moderate variables and the independent variables are 0.621, 0.393, 0.717, 0.699, respectively. We could then carry out the next step of regression analysis on this basis.
Table 6: Correlation matrix for all variables
Pearson correlation 1 2 3 4 5 6 SMM 1 trust 0.595** 1 risk 0.243** 0.181** 1 WoM 0.692** 0.587** 0.264** 1 EoU 0.601** 0.641** 0.287** 0.622** 1 PB 0.661** 0.621** 0.393** 0.717** 0.699** 1
** Correlation is significant at the .01 level (2-tailed).
4.1.2 Partial correlation
4.2 Standard multiple regression
Before launching the regression analysis, we examined the sample size requirements. Tabachnick and Fidell (2013) give a formula for calculating sample size requirements as follows: N > 50 + 8m (where m = number of independent variables). Then we conducted a regression analysis between SMM and PB. We added the moderate variables, in turn, to check for changes in the equation and to see if the moderate variables affect. We used the adjusted R-squared, sig., and corresponding unstandardized beta to compare the equation and each moderate variable’s influence. As predictors are added to a model, each predictor will explain some of the variances in the dependent variable simply due to chance. The adjusted R-squared attempts to produce a more honest value to estimate R-squared for the sample.
4.2.1 The regression analysis of hypothesis 1: SMM and PB
As seen in Table 7, the R-squared is 0.411 and 0.450, respectively, which indicates the predictor SMM can explain over 40% of the change in sample purchase behavior whether in China or Sweden. This result is quite respectable. In the results of ANOVA, F is 101.124 and 83.598 for both groups, and the P-value is less than 0.05, which is statistically significant. Accordingly, we describe the equation as follows:
Purchasing behavior(China)= 0.586+ 0.731*SMM+error
Purchasing behavior(Sweden)= 0.514+ 0.7*SMM+error
Table 7: Regression of SMM and PB
China Sweden Model summary R-squared 0.411 0.45
ANOVA F 101.124 83.598
Sig. 0.00 0.00
Coefficients (Constant) 0.586 0.514 Unstandardized beta 0.731 0.7
Sig. 0.00 0.00
As can be seen from the coefficient rows, the coefficient of SMM is 0.731 and 0.7, respectively, indicating that SMM has a positive impact on consumer PB. Therefore, hypothesis 1 in this study is supported.
4.2.2 The regression analysis of hypothesis 2: SMM, trust, and PB
We added the moderate variable trust to the regression analysis of SMM and PB. As shown in Table 8, there is a significant regression relationship between these three factors in both the Chinese and Swedish data, indicated by the R-squared (0.47 and 0.57) and P-value (0.00 and 0.00).
Then, we compared the adjust R-squared with the equation of hypothesis 1. For China, the adjusted R-squared is 0.463, which is larger than the original adjusted R-squared value (0.407), indicating that trust improves the model. The P-value of trust is lower than 0.05, and the corresponding unstandardized beta is 0.323. Similarly for Sweden, the adjusted R-squared is 0.573, which is larger than the original value (0.450), suggesting that trust improves the model. Moreover, the P-value of trust is 0.00, and the unstandardized beta is 0.515. Accordingly, we describe the equation as follows:
Purchasing behavior(China)= 0.18 + 0.568*SMM + 0.323*trust +error
Purchasing behavior(Sweden)= 0.082+ 0.355*SMM+ 0.515* trust +error
Table 8: Regression of SMM, trust, and PB
China Sweden Model summary R-squared 0.47 0.579
Adjusted R-squared 0.463 0.571 Sig. 0.00 0.00 ANOVA F 63.902 69.414 Sig. 0.00 0.00 Coefficients (Constant) 0.18 0.082 Unstandardized beta (SMM) 0.568 0.355 Unstandardized beta (trust) 0.323 0.515
Sig. (SMM) 0.00 0.00
The unstandardized beta of trust for China (0.323) is lower than it is for Sweden (0.515), which indicates that trust has a more substantial impact on PB in Sweden. Thus, we can reject hypothesis 2.
4.2.3 The regression analysis of hypothesis 3: SMM, risk, and PB
We added the moderate variable risk to the regression analysis. It can be seen in Table 9 that R-squared values are 0.421 and 0.473, respectively, indicating that the predictors could account for more than 40% of the changes in consumer PB in both China and Sweden. The adjust R-squared of the Chinese data is 0.421, slightly higher than the original adjusted R-squared value (0.407), indicating that the risk makes a limited improvement on the model. The P-value of risk is higher than 0.05, indicating that this factor is insignificant. In contrast, the adjusted R-squared of the Swedish data is 0.473, slightly higher than the original value (0.450), indicating that risk improves the model to some extent. The P-value is 0.039, which is less than 0.05, and the corresponding non-standardized beta value is 0.141. Accordingly, we describe the equation as follows:
Purchasing behavior(China)= 0.346 + 0.686*SMM +error
Purchasing behavior(Sweden)= 0.098+ 0.685*SMM+ 0.141* risk +error
Table 9: Regression of SMM, risk, and PB
China Sweden Model summary R-squared 0.421 0.473
Adjusted R-squared 0.413 0.463 Sig. 0.00 0.00 ANOVA F 52.263 45.364 Sig. 0.00 0.00 Coefficients (Constant) 0.346 0.098 Unstandardized beta (SMM) 0.686 0.685 Unstandardized beta (risk) 0.106 0.141
Sig. (SMM) 0.00 0.00
Compared with the insignificant P-value of risk in the Chinese data, although the unstandardized beta in Sweden (0.141) is very low, it still indicates the influence of risk on Swedish consumers’ PB. Therefore, we partly accept hypothesis 3.
4.2.4 The regression analysis of hypothesis 4: SMM, WoM, and PB
We added the moderate variable WoM to the regression analysis. As shown in Table 10, the R-squared values are 0.514 and 0.618, respectively, indicating that the predictors could account for more than 50% of the changes in consumer PB in both China and Sweden. The adjusted R-squared of the Chinese data is 0.507, higher than the original adjusted R-squared value (0.407), indicating that WoM distinctly improves the model. The P-value of WoM is lower than 0.05, and the corresponding unstandardized beta is 0.403. Also, the adjusted R-squared of Swedish data is 0.611, which is higher than the original value (0.450), indicating that WoM improves the model more than would be expected. The P-value is 0.00, and the corresponding unstandardized beta value is 0.605. Accordingly, we describe the equation as follows:
Purchasing behavior(China)= 0.389 + 0.412*SMM + 0.403*WoM +error
Purchasing behavior(Sweden)= 0.168+ 0.245*SMM+ 0.605* WoM+error
Table 10: Regression of SMM, WoM, and PB
China Sweden Model summary R-squared 0.514 0.618
As shown in the equations above, the unstandardized beta of WoM for China (0.403) is lower than it is for Sweden (0.605), which means that WoM has a more substantial impact on PB in Sweden. So we reject hypothesis 4.
4.2.5 The regression analysis of hypothesis 5: SMM, EoU, and PB
We added the moderate variable perceived EoU to the regression analysis. As shown in Table 11, the R-squared values are 0.507 and 0.643, respectively, indicating that the predictors could account for more than 50% of the changes in consumer PB in both China and Sweden. The adjusted R-squared of the Chinese data is 0.5, higher than the original adjusted R-squared value (0.407), indicating that EoU noticeably improves the model. The P-value of perceived EoU is 0.00, and the unstandardized beta is 0.342. The adjusted R-squared of Swedish data is 0.636, which is profoundly higher than the original value (0.450), indicating that EoU improves the model more than expected. The P-value is 0.00, and the corresponding unstandardized beta is 0.459. Accordingly, we describe the equation as follows:
Purchasing behavior(China)= 0.392 + 0.459*SMM + 0.342*EoU +error
Purchasing behavior(Sweden)= 0.242+ 0.373*SMM+ 0.459* EoU+error
Table 11: Regression of SMM, EoU, and PB
China Sweden Model summary R-squared 0.507 0.643
Adjusted R-squared 0.5 0.636 Sig. 0.00 0.00 ANOVA F 74.129 90.96 Sig. 0.00 0.00 Coefficients (Constant) 0.392 0.242 Unstandardized beta (SMM) 0.459 0.373 Unstandardized beta (EoU) 0.342 0.459
Sig. (SMM) 0.00 0.00
Sig. (EoU) 0.00 0.00
As shown in the equations above, the unstandardized beta of EoU for China (0.342) is lower than it is for Sweden (0.459), which means that EoU has a stronger impact on PB in Sweden. So we accept hypothesis 5.
4.3 Chapter Summary
From the above analysis, as shown in the figure below, it possible for us to conclude that hypothesis 1, hypothesis 3, and hypothesis 5 of our five hypotheses can be supported, while hypothesis 2 and hypothesis 4 are rejected. In the next chapter, we discuss the outcome from our survey, make practical recommendations, and provide examples of how researchers can contribute with further research in this area.
5. Discussion
5.1 Basic data discussion and comparison
By looking at the data overall, we can see the general attitude of consumers towards SMM through the values of various factors. In the average statistics, the risk values of China and Sweden are the highest in their respective data (3.66 and 3.41), which indicates that consumers in both markets are clearly aware of the higher risk problems in SMM. In terms of positive influencing factors, the lowest value in China is trust (3.08), but this is still higher than the midpoint set by the survey (3.00). This shows that the psychological attitude of Chinese consumers towards SMM in terms of cosmetics is generally positive, which is consistent with the previous research results of Yan (2012) and others.
The Swedish data shows different situations: both WoM (2.79) and perceived EoU (2.94) are below the midpoint. What is interesting is that through regression analysis, we found that WoM and perceived EoU have the greatest impact on consumers’ purchase intentions. We interpret this phenomenon as showing that Swedish consumers are more likely to be influenced by the opinions of other consumers, and this influence directly leads to the negativity of consumer behavior (2.76). Because the implementation of SMM is closely related to the information accumulation process of consumers, there are many different information providers involved (Sun et al, 2005). Their subjective opinions constitute negative word-of-mouth values and have an impact on consumer behavior. In addition, previous research has found that WoM has an important effect on Swedish consumers’ decisions to purchase (Grönroos, 2008), which is consistent with our findings.
5.2 Discussion of research results
The results of regression analysis show that H1 established, that is, in both the Chinese and Swedish markets, SMM can have a positive effect on consumer purchasing intention. As described above, SMM can positively influence consumers’ willingness to purchase and change brand preferences and can also improve customer loyalty (Godey, 2016). Therefore, SMM highlights the importance of social media in helping brands get a lot of attention. The online environment built by social media offers the possibility of enhancing the relationship between producers and consumers (Vinerean, 2013). On the basis of hypothesis 1, we are able to further analyze what factors enable SMM to attract consumers and show the most value and which factors are worthy to be considered by firms when entering the cosmetics market in these two distinct countries.
The results of multiple regression show that there are obvious differences between the influential psychological factors displayed in the Swedish and Chinese markets: trust, risk, word of mouth, and perceived EoU have a positive effect in the Swedish market, while in the Chinese market, it is trust, word of mouth, and perceived EoU that have the strongest effect on the dependent variable. In horizontal comparison, all the above factors have a stronger impact on consumer behavior in Sweden than in China, which proves that hypothesis 3 and hypothesis 5 are true. In the following sections, we discuss each factor separately.
5.2.1 Trust
Past research has shown that the cultural values of the target group can greatly influence perceptions of trust (Schumann et al., 2010). In Schumann’s (2010) research, in the category of trust (including the four dimensions of integrity, ability, predictability, and benevolence), it is found that customers in a high power distance culture attach more importance to experiencing the trust brought by quality services; compared with low power distance customers, they pay more attention to accumulating trust through integrity. According to the 2020 Global Power Distance Index, China (80) and Sweden (31) have low and high power distance relationships, respectively. When discussing SMM, the expansion of information acquisition channels and the virtual network environment make it difficult for Chinese consumers to experience high-quality services and integrity. This makes it difficult for SMM to make Chinese consumers build trust that can influence their consumption decisions. In our study, the average Swedish consumer trust index is significantly higher than that of Chinese consumers, which is consistent with the results of Viklund (2003), which also indirectly confirms the above explanation.
Yan (2012) found in her research that trust significantly affects the consumer behavior of Chinese consumers in the Internet environment, but our research results do not support this claim. The reason for this difference in results may be related to the product specificity of cosmetics or the setting of control variables. We believe that these factors may cause deviations in the trust utility gained by consumers. Therefore, it is worth noting that our research results only apply to mainstream consumer groups of cosmetics in the SMM environment, and this conclusion does not apply to other products or marketing techniques.
5.2.2 Risk
Camarero, 2009), we added risk factors to the survey to test whether it would affect the dependent variable.
The regression results indicate that risk in Sweden has a positive, small β value (0.141) and significance of less than 0.05. In China, there is a significance value of 0.112, which does not have statistical significance. This shows that risk has the ability to influence the dependent variable of consumer behavior in Sweden but has no similar effect in China. From the perspective of basic data, Chinese consumers show a higher degree of risk perception and believe that SMM is more likely to involve risk, but this does not affect Chinese consumers’ consumption through SMM.
5.2.3 Word of mouth
As mentioned earlier, WoM has a significant positive effect on the dependent variable in both Sweden and China. WoM has the greatest impact on Swedish consumers’ behavior, which indicates that WoM is the most critical factor for Swedish consumers to decide whether to choose cosmetics using SMM. According to the survey, consumers’ evaluation of WoM may come from communication between friends on social platforms or celebrity promotion on public channels; Swedish consumers generally show low interest in sharing cosmetics on social media and are not easily influenced by celebrity promotion.
an avid desire to buy. This directly caused the weakening of the impact of WoM on consumers. Our conclusion supports the discovery of Chu et al. (2011).
Past social research also tells us that one of the biggest differences between the two markets is that China is a collectivist culture and Sweden is an individualistic culture. Collectivism emphasizes the clustering of personal interests, and under certain conditions, the amplification of personal interests can be achieved by gathering the power of individuals; in contrast, individualism attaches more importance to individual freedom and emphasizes that individuals enter society to achieve their personal interests (Triandis, 2018). This can explain why hypothesis 4 is rejected by the survey results. The results of this study show that in an individualistic culture, consumers are more likely to decide whether to buy cosmetics based on the reputation of others. This is consistent with the research of Kacen (2002). His research shows that the needs of the self and the individual are highly valued by Western culture. Consumers affected by individualism are more likely to actively meet their personalized needs. The phenomenon in our research is that Swedish consumers are more willing to collect word-of-mouth information and judge it separately. At the same time, collectivist consumers pay more attention to emotional control and moderation, which ultimately weakens the impact of psychological and emotional factors (including WoM) on consumption (Kacen, 2002).
Another study that supports this explanation comes from Lee (2005). He found that individualistic consumers are more likely to consider information on the Internet to be targeted at the majority of the public. Such information cannot meet the unique needs of individuals. Therefore, individualistic consumers conduct further “emotional processing” after receiving the information to satisfy their desire for uniqueness. Therefore, the underlying reason involved in hypothesis 4 is likely to be directly derived from the difference in social culture; in other words, the difference between individualism and collectivism causes this phenomenon.
5.2.4 Perceived ease of use
6. Conclusion
6.1 General conclusion
The purpose of this study was to investigate the differences in factors such as trust, risk, word of mouth, and perceived EoU displayed by Swedish and Chinese consumers in the context of social media marketing of cosmetics and how these factors affect the purchasing intentions of consumers in the two countries. We conducted a quantitative study of cosmetics purchase behavior and social media use in mainstream groups of consumers. After collecting data through online questionnaires and conducting further analysis, we can answer the research question: How do Chinese and Swedish consumers behave differently (or similarly) toward social media marketing when purchasing cosmetic products?
First of all, through multiple regressions, we found that both Chinese and Swedish consumers have a positive attitude towards the SMM of cosmetics. SMM can increase the purchase intention of consumers in both countries to a certain extent. Further, we found that four different factors all play a role in the Swedish consumer group, and three factors (other than risk) have some effects on the purchasing intentions of Chinese consumers. Among them, it is also found that the influence of these factors on Swedish consumers is generally stronger than on Chinese consumers, which is the most obvious difference between the two markets.
are different from those of previous studies. 6.2 Implications
We can provide some practical advice to marketers and companies in the cosmetics industry. These opinions are of value to all companies that conduct cosmetics marketing through social media, especially for companies engaged in cross-border or cross-cultural marketing.
We recommend that companies using SMM for cosmetic marketing should shift their focus appropriately when entering the Chinese and Swedish markets. In both markets, companies should focus on EoU and WoM, which are the two factors that have the greatest impact on consumers’ purchasing intentions. Marketers should also consider improving consumer trust to a certain extent, because trust has an impact on consumers’ purchase intentions.
The difference reflected in the two markets lies mainly in whether risk results in changes to purchasing intention. In the Swedish market, companies should pay more attention to reducing the risk perceived by consumers. The lower the risk perceived by consumers, the more likely they are to purchase. In China, there is no need to spend too much on this, because Chinese consumers are not sensitive to risk. We also found that all four factors involved in the study have a stronger influence on purchase intentions in Sweden than in China. Therefore, we infer that this result can rise to the contrast of cultural differences. Swedish society is typically individualist, and Chinese society is collectivist. When a company enters the market of an individualistic society, it should pay more attention to the influence of consumers’ subjective perception factors on PB. The reality requires that marketers place customers in a core and important position, then design a customer-friendly marketing plan and put it into practice according to customer needs.
6.3 Limitations and recommendations for future study
and 104 from Swedish respondents. However, the online platform made it difficult to screen the information of these interviewees further. Even though some control variables were preset in the survey to control the accuracy of the interviewed targets, we cannot confirm them face to face and cannot fully confirm the reliability of the information. In addition, respondents may have shared the questionnaire with family and friends, and these new respondents may not meet the specifications of the preset target group. These problems may cause deviations in the results of this survey. It is recommended that future research should collect more samples for testing and the questionnaire should be converted into a combination of online and offline models to improve the accuracy of the research.
Reference
Adams, J., Khan, H.T., Raeside, R., and White, D.I., 2007. Research methods for
graduate business and social science students. India: SAGE Publications.
Ahuja, M., Gupta, B., and Raman, P., 2003. An empirical investigation of online consumer purchasing behavior. Communications of the ACM, 46(12), pp.145–151. Ajzen, I. and Fishbein, M., 1977. Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84(5), p.888.
Alexandria, 2018. Micro Influencers – Fashion & Beauty Brands Win On Instagram [online]. Socialbakers.com. Available at: <https://www.socialbakers.com/blog/2757-how-influencer-content-has-put-fashion-and-beauty-brands-ahead-on-instagram> [Accessed 2 June 2020].
Annamoradnejad, I., Fazli, M., Habibi, J., and Tavakoli, S., 2019. Cross-cultural studies using social networks data. IEEE Transactions on Computational Social
Systems, 6(4), pp.627–636.
Arndt, J., 1967. Role of product-related conversations in the diffusion of a new product. Journal of Marketing Research, 4(3), pp.291–295.
Barefoot, D. and Szabo, J., 2009. Friends with benefits: A social media marketing
handbook. No Starch Press, San Francisco, USA.
Bauer, R.A., 1960. Consumer behavior as risk taking. Proceedings of the 43rd National Conference of the American Marketing Assocation, June 15, 16, 17, Chicago, Illinois, 1960. pp.384–398.
Beer, C., 2020. Social browsers are brand-engaged [online]. GlobalWebIndex Blog. Available at: <https://blog.globalwebindex.com/chart-of-the-day/social-browsers-brand/> [Accessed 25 March 2020].
Blackwell, R.D., Miniard, P.W., and Engel, J.F., 2006. Consumer behavior. 10th edition. Thomson/South-Western, Mason, USA. pp.234–238.
Buffer, 2020. State of social 2019 [online]. Available at: <https://buffer.com/state-of-social-2019> [Accessed 25 March 2020].
Chen, J. and Dibb, S., 2010. Consumer trust in the online retail context: Exploring the antecedents and consequences. Psychology & Marketing, 27(4), pp.323–346.
Cheung, C.M., Chan, G.W., and Limayem, M., 2005. A critical review of online consumer behavior: Empirical research. Journal of Electronic Commerce in
Organizations (JECO), 3(4), pp.1–19.
Chong, A.Y.L., Ooi, K.B., Lin, B., and Bao, H., 2012. An empirical analysis of the determinants of 3G adoption in China. Computers in Human Behavior, 28(2), pp.360– 369.
Chu, S.C. and Kim, Y., 2011. Determinants of consumer engagement in electronic word-of-mouth (eWOM) in social networking sites. International Journal of
Advertising, 30(1), pp.47–75.
Clement, J., 2019. Instagram: Most-followed beauty brands 2019 [online]. Statista. Available at: <https://www.statista.com/statistics/536991/leading-beauty-brands-instagram-followers/> [Accessed 2 June 2020].
Coakes, S.J. and Steed, L.G., 2003. SPSS : analysis without anguish : version 11.0 for
Windows. Brisbane: Jacaranda Wiley.
Cole, S.J., 1998. Testimony before the subcommittee on telecommunications, trade and consumer protection committee on commerce, US House of Representatives, Washington, DC. Better Business Bureau.
Consulting, D., 2020. Selling cosmetics in China: Analysis of the cosmetics market in
China – Daxue Consulting – Market Research China [online]. Daxueconsulting.com.
Davis, F.D., 1985. A technology acceptance model for empirically testing new
end-user information systems: Theory and results. Doctoral dissertation, Massachusetts
Institute of Technology.
Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, Vol. 13, No. 3, pp.319–340.
Dehghani, M., Kim, K.J., and Dangelico, R.M., 2018. Will smartwatches last? Factors contributing to intention to keep using smart wearable technology. Telematics and
Informatics, 35(2), pp.480–490.
Emarsys, 2020. Top 5 social media predictions 2019 [online]. Available at: <https://emarsys.com/learn/blog/top-5-social-media-predictions-2019/> [Accessed 25 March 2020].
Felix, R., Rauschnabel, P.A., and Hinsch, C., 2017. Elements of strategic social media marketing: A holistic framework. Journal of Business Research, 70, pp.118–126. Fishbein, M. and Ajzen, I., 1977. Belief, attitude, intention, and behavior: An introduction to theory and research. Philosophy & Rhetoric, Vol. 10, No. 2, pp. 130– 132.
Forsythe, S.M. and Shi, B., 2003. Consumer patronage and risk perceptions in Internet shopping. Journal of Business Research, 56(11), pp.867–875.
Freiden, J., Goldsmith, R., Takacs, S., and Hofacker, C., 1998. Information as a product: Not goods, not services. Marketing Intelligence & Planning, Vol. 16, No. 3, pp. 210-220.
Gefen, D. and Straub, D.W., 2004. Consumer trust in B2C e-commerce and the importance of social presence: Experiments in e-products and e-services. Omega,
32(6), pp.407–424.
George, D. and Mallery, M., 2003. Using SPSS for Windows step by step: A simple
Gilsenan, K., 2019. In review: Social media is changing, and it’s not a bad thing
[online]. GlobalWebIndex Blog. Available at:
<https://blog.globalwebindex.com/trends/2019-in-review-social-media/> [Accessed 25 March 2020].
Given, L.M. (ed.), 2008. The Sage encyclopedia of qualitative research methods. Sage Publications, Thousand Oaks, USA.
Godey, B., Manthiou, A., Pederzoli, D., Rokka, J., Aiello, G., Donvito, R., and Singh, R., 2016. Social media marketing efforts of luxury brands: Influence on brand equity and consumer behavior. Journal of Business Research, 69(12), pp.5833–5841.
Goldsmith, R.E., 2000. Bridges E. E-tailing vs. retailing: Using attitudes to predict online buying behavior. Quarterly Journal of Electronic Commerce 1, No. 3: 245–53. Goodrich, K. and De Mooij, M., 2014. How ‘social’ are social media? A cross-cultural comparison of online and offline purchase decision influences. Journal of
Marketing Communications, 20(1–2), pp.103–116.
Grönroos, C., Holmqvist, A., and Gylldorff, L., 2008. Service management och
marknadsföring: kundorienterat ledarskap i servicekonkurrensen. Malmö: Liber.
Hair, J.F., Anderson, R.E., Babin, B.J., and Black, W.C., 2010. Multivariate data
analysis: A global perspective (Vol. 7). Pearson, Upper Saddle River, USA.
Hasslinger, A., Hodzic, S., and Opazo, C., 2008. Consumer behaviour in online shopping. Kristiansand: Kristiansand University, Department of Business Studies. Retrived from http://www.diva-portal. org/smash/get/diva2:231179/FULLTEXT01. Heale, R. and Twycross, A., 2015. Validity and reliability in quantitative studies.
Evidence-based Nursing, 18(3), pp.66–67.
Helm, S. and Schlei, J., 1998. Referral potential–potential referrals. An investigation into customers’ communication in service markets. Track, 1, pp.41–56.
Hoffman, D.L. and Novak, T.P., 2012. Social media strategy. Handbook of marketing
Hofstede, G., 1983. The cultural relativity of organizational practices and theories.
Journal of International Business Studies, 14(2), pp.75–89.
Hofstede, G., Hofstede, G.J., and Minkov, M., 2010. Cultures and organizations:
Software of the mind. Revised and expanded 3rd edition. New York: McGraw-Hill.
Hoonakker, P. and Carayon, P., 2009. Questionnaire survey nonresponse: A comparison of postal mail and internet surveys. International Journal of Human–
Computer Interaction, 25(5), pp.348–373.
Hsiao, K.L. and Chen, C.C., 2018. What drives smartwatch purchase intention? Perspectives from hardware, software, design, and value. Telematics and Informatics,
35(1), pp.103–113.
Hulin, C., Netemeyer, R., and Cudeck, R., 2001. Can a reliability coefficient be too high? Journal of Consumer Psychology, 10(1/2), pp.55–58.
Ibrahim, B., Aljarah, A., and Ababneh, B., 2020. Do social media marketing activities enhance consumer perception of brands? A meta-analytic examination. Journal of
Promotion Management, 26(4), pp.544–568.
Jansson, V., 2016. Online shopping behavior in Sweden and China: What is different
about Chinese consumers? [online]. Retrieved from
http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-275327</div>
Jenn, C., 2020. Social media demographics to inform your brand’s strategy in 2020 [online]. Sprout Social. Available from: https://sproutsocial.com/insights/new-social-media-demographics/ [Accessed 2 April 2020].
Jing, M., Wang, F.H., and Zhou, Y., 2005. Study on dynamic model of perceived risk during consumer’s online shopping. Industrial Engineering and Management, pp.28-32.
Kaplan, A.M. and Haenlein, M., 2010. Users of the world, unite! The challenges and opportunities of social media. Business Horizons, 53(1), pp.59–68.
Kihn, L.A., 2001. Strategies, decentralization, and controls in internationalized Finnish firms.Finnish Journal of Business Economics, 1 (1): 35–57.
Kim, A.J. and Ko, E., 2010. Impacts of luxury fashion brand’s social media marketing on customer relationship and purchase intention. Journal of Global Fashion
Marketing, 1(3), pp.164–171.
Koh, T., 2020. SEO Singapore: A complete guide [online]. MediaOne. Available at: <https://mediaonemarketing.com.sg/seo-singapore-complete-guide/> [Accessed 25 March 2020].
Komiak, S.Y. and Benbasat, I., 2006. The effects of personalization and familiarity on trust and adoption of recommendation agents. MIS Quarterly, pp.941–960.
Kotler, P. and Keller, K.L., 2006. Marketing management. Pearson international edition. Pearson Prentice Hall, Upper Saddle River, USA.
Kuester, S., 2012. MKT 301: Strategic marketing and marketing in specific industry contexts. University of Mannheim, 110, pp.393–404.
Lazer, W. and Kelley, E.J., 1973. Social marketing: Perspectives and viewpoints. McGraw-Hill/Irwin, New York, USA.
Lee, W.N. and Choi, S.M., 2005. The role of horizontal and vertical individualism and collectivism in online consumers’ responses toward persuasive communication on the web. Journal of Computer-Mediated Communication, 11(1), pp.317–336.
Lonner, W.J., Berry, J.W., and Hofstede, G.H., 1980. Culture’s consequences: International differences in work-related values. University of Illinois at
Urbana-Champaign’s Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship.
Perez-Sindin Lopez: <https://xaperezsindin.com/2013/12/11/advantages-and-disadvantages-of-secondarydata-collection/>
McKnight, D.H., Choudhury, V., and Kacmar, C., 2002. The impact of initial consumer trust on intentions to transact with a web site: A trust building model. The
Journal of Strategic Information Systems, 11(3–4), pp.297–323.
Milwood, P., Marchiori, E., and Zach, F., 2013. A comparison of social media adoption and use in different countries: The case of the United States and Switzerland.
Journal of Travel & Tourism Marketing, 30(1–2), pp.165–168.
Morrison, D.G., 1979. Purchase intentions and purchase behavior. Journal of
Marketing, 43(2), pp.65–74.
Muntinga, D.G., Moorman, M., and Smit, E.G., 2011. Introducing COBRAs: Exploring motivations for brand-related social media use. International Journal of
Advertising, 30(1), pp.13–46.
Pallant, J.F., 2000. Development and validation of a scale to measure perceived control of internal states. Journal of Personality Assessment, 75(2), pp.308–337. Park, C. and Lee, T.M., 2009. Antecedents of online reviews’ usage and purchase influence: An empirical comparison of US and Korean consumers. Journal of
Interactive Marketing, 23(4), pp.332–340.
Peter, J.P., Olson, J.C., and Grunert, K.G., 1999. Consumer behaviour and marketing
strategy. London: McGraw-Hill, pp.329–348.
Pettersson, C. and Söderström, B., 2015. Factors affecting consumers purchase intentions for digital news: A quantitative study of Resumé[online]. Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-105706
Pires, G., Stanton, J., and Eckford, A., 2004. Influences on the perceived risk of purchasing online. Journal of Consumer Behaviour: An International Research
