To what extent do beauty bloggers influence the purchase behavior of their audience : wxploring the links between personality and blog preferences

(1)

TO WHAT EXTENT DO BEAUTY BLOGGERS INFLUENCE

THE PURCHASE BEHAVIOR OF THEIR AUDIENCE:

EXPLORING THE LINKS BETWEEN PERSONALITY AND

BLOG PREFERENCES.

Author: Mai Nguyen Student number: 10919104 Date of submission: 29 Jan 2016 (Final version) Program: MSc. in Business Administration – Marketing Track

(2)

ii STATEMENT OF ORIGINALITY

This document is written by Mai Nguyen, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

ACKNOWLEDGEMENTS

I am grateful for the guidance of my supervisor, who always responses quickly and encourages me to look at the problems from different angles. He has showed me the value of a master thesis: it is not only about meeting the requirements, but also about perusing personal interests and searching for the most comprehensive answers possible.

My appreciation also goes to all respondents who participated in the online survey. Thanks to their input, I was able to grasp more information about the purchase of beauty products.

(4)

iv ABSTRACT

This quantitative research sets the scene in Vietnam, an emerging beauty market with high potential. It studies the extent to which beauty bloggers have an influence on the purchase behavior of their audience. The dataset consists of 200 respondents, among which half has never and half has purchased beauty products from blog recommendations. The independent variables are Personality (Neuroticism/Extraversion) and Frequency of blog visits. It was revealed that there is a negative relationship between neuroticism and the likelihood of making purchase from blog recommendation; however, this relationship becomes positive if the blog is written scientifically. There is also a positive relationship between extraversion and the likelihood if purchase under the moderating effect of emotional writing style. A positive relationship is found between frequency of visits and the likelihood of making purchase, to a certain degree. These findings have a number of practical implications for bloggers and companies who collaborate with them, such that their blog posts can be more influential.

(5)

Previous studies acknowledge the power consumers gain from Web 2.0 (Labrecque et al., 2013), which allows them to voice their opinions about certain brands and/or products to others, even to a mass audience via social media (Boyd & Ellison, 2007; Smith et al., 2012). A study in the USA finds that 70% of consumers does not consider advertising claims to be truthful; advertisement is believed to have the purpose of persuading people to buy things they do not want (Calfee and Ringold, 1994). This declining of trust in advertisement, together with the development of Web 2.0, encourages the growth of online third-party opinions. The results of a study by Dean (1999) suggest that third-party endorsement can also be perceived as a signal of product quality, together with other established signals such as brand, price, physical appearance and store name (Dawar and Parker 1994). In an empirical study regarding individuals’ skepticism to blog review and advertisement on a cosmetic product, Wang & Chien (2012) reveal that people are convinced much more by blog product review than by advertisement.

Herring et al. (2005) define a blog as a frequently edited webpage, the main utility of which is the chronological recording and display of individual articles. Understanding the value of blog review, public relations practitioners often work with bloggers to provide support for content and to increase product reviews. The brand coverage and product discussion on blogs is considered by many marketers to be more valuable in terms of credibility than other forms of more traditional promotion efforts (Smith, 2011). There are two main explanations for the credibility of independent blogs. Firstly, blog readers believe that independent blogs display honest and reasonable opinion instead of biased information, as compared to sponsored or corporate sources (Flanagin and Metzger, 2007; Johnson and Kaye, 2004; Senecal and Nantel, 2004). Secondly, the blogger who specialized in writing

(8)

2

about certain theme is seen as someone who is closer to the source and can offer more up-to-date information (Kaye and Johnson, 2011; Zhu & Tan, 2007). Undoubtedly, blogs are seen as a credible source and can have an effects on the readers. However, it is unclear how large the influence of beauty bloggers on their audience is, or under which circumstances this influence is presence. The research question is formulated as below:

TO WHAT EXTENT DO BEAUTY BLOGGERS INFLUENCE THE PURCHASE BEHAVIOR OF THEIR AUDIENCE?

The goal of this research is to investigate the role of audience personality, blog writing style, and frequency of blog visits on the purchase of beauty products recommended in the blog. This study enriches the academic literature by examining the extent to which beauty bloggers can influence their audience, and providing a better understanding of how personality and writing style affect blog reception among audience. It also offers ideas for blog writers to identify proper target groups and to make their posts more influential. Cosmetics companies who partner up with beauty bloggers to promote their products can use these suggestions as guidelines to enhance the positive effect of blog posts on customers’ buying behavior.

(9)

1.2 The Vietnamese beauty market

This study sets the scene in the Vietnamese beauty market, an emerging market with distinctive characteristics from the well-established ones. In larger cosmetics markets, customers can easily visit a store to test a product, e.g. feel the texture, see the color in real life, or smell the fragrance. In an interview with GCI Magazine in 2014, Le, former President of the ASEAN Cosmetic Scientific Body, argues that the Vietnamese beauty market is very fragmented, and cannot be seen as one single market. Consumers who live in Vietnamese rural areas have less opportunity of visiting a physically store; whereas consumers who live in big cities have more opportunity, but also limited availability of international brands compared to customers in large foreign markets. Due to this lower accessibility level of products, Vietnamese consumers usually place online orders, and receive the products via a shipping agent or acquaintances after a long time haul. Prior to buying, customers have almost no way to come to physical contact with a tester. Even for products whose colors play a primary role, such as lipsticks or eye shadows, customers can only see the colors via their screens, which may not be the true representative of the products in real life. This is where the role of beauty bloggers becomes dominant: consumers actively seek for online reviews to have a better idea about the products. Although being fragmented, the beauty market in Vietnam is expected to have huge potential and strong performance in the future, with the current growth rate 30% per annum (Euromonitor International, 2015; GCI Magazine, 2014). Thus, the insights discovered through this thesis are also beneficial to develop a vision in shaping such emerging market.

This paper continues with a Literature Review chapter, where relevant theories and findings are discussed and developed into hypotheses. Subsequently, chapter three Data and Method provides detailed information about the data collection procedure and research method. Data collected from the survey is processed in SPSS and presented in chapter four

(10)

4

Results. Next, implications of the results are considered in chapter five Discussions. Lastly, chapter six Conclusion provides an overview of the thesis, together with outlining some limitations and recommendations for future research.

(11)

CHAPTER 2. LITERATURE REVIEW 2.1 Personality

Extraversion (as opposed to introversion) and Neuroticism (as opposed to stability) have long been used as the two main dimensions to describe human personality (Eysenck, 2013). In social sciences, personality has been linked to a wide range of human attitudes and behaviors. A study by Soest et al. (2009) about the relationship between personality and cosmetics surgery shows that applicants for cosmetic surgeries score significantly higher on extraversion compared to the representative sample. In a study by Robertson et al. (2008) positive relationship is found between cosmetic usage and anxiety, which is a trait of neuroticism (Jeronimus et al., 2014). These studies show that both neurotics and extraverts more or less care about their appearance and thus may purchase beauty products. The focus of this thesis is narrowed down to the purchase of beauty products that are recommended by beauty bloggers.

This is when the concept of “trust” comes in. According to Rousseau et al. (1998), trust can be defined as “a psychological state comprising the intention to accept vulnerability based upon the positive expectations of the intentions or behavior of another”. In order to follow a recommendation, the audience should trust the beauty blogger: they expect the product to perform as positive as the blog review says. Evans and Revelle (2008) find trust to be positively associated with extraversion, and negatively correlated with neuroticism. As neurotics find it more difficult to trust other people, it can be hypothesized that they are less likely to purchase an item based on an online review. On the other hand, extraverts are more open to taking online recommendations because they are more trusting. In addition, according to Eysenck's model of personality (1981), risk-taking is also a trait of extraversion, which further enhance the possibility that extraverts are more likely to purchase an item based on online review.

(12)

6

Taking the above-mention literature into consideration, the following hypotheses are made about the relationship between personality and the likelihood of buying products from beauty blog recommendations:

• H1: There is a negative relationship between neuroticism and the likelihood of

purchasing products from beauty blog recommendations.

• H2: There is a positive relationship between extraversion and the likelihood of

(13)

2.2 Writing style

The writing style of a beauty blogger can be classified into “scientific” or “emotional”. Bloggers who write more on the scientific side focus on analyzing chemical ingredients and how they penetrate the skin, evaluating whether the packing is sufficient to keep the chemicals stable, and providing clear evidence about the performance of the products. On the other hand, bloggers who write more on the emotional side mostly express their subjective feeling and their personal experience with the products. An example of emotional writing style could be: “This cream feels very nice on my face”, whereas a highly scientific review would say: “This cream contains 5% Alpha Hydroxy Acid, which gently exfoliates the skin, resulting in even skin tone and smooth skin surface”.

To a limited degree, word use of blog authors can reflect their personalities; for example, extravert writers use more positive emotion words (Li & Chignell, 2010). In the same paper, it is found that blog readers might be more attracted to the writing style which signals that the author has similar personalities to their own. In other words, it can be hypothesized that extravert readers prefer the writing style of extravert authors, i.e. the writing style that contains more positive emotion words.

Previous research shows that there are links between personality and media preferences. Weaver (1991) finds that people who score high on neuroticism expressed a strong preference for information/news television programs. Applying this finding to the beauty blogs in an intuitive way, it can be hypothesized that neurotic audience prefer a highly informative and scientific writing style.

(14)

8

The following hypotheses are formed to examine whether the correspondence between writing style and audience personality has a relationship with the likelihood of making a purchase from the blogger’s recommendation.

• H3: Scientific writing style is more appealing to audience with neuroticism

personality type, turning the negative relationship in H1 into positive.

• H4: Emotional writing style is more appealing to audience with extraversion

(15)

2.3 Frequency of blog visits

Agarwal et al. (2008) deem that bloggers can influence the buying decisions of their audience. As cited in this paper, different sources acknowledge the blogosphere for its ability to “reshape business models, inspire viral marketing, provide trend analysis and sales prediction, […]”. A study by Qader and Zainuddin (2011) confirms that media exposure had a significant positive influence on consumers’ purchase intention. Thus exposure to blog contents is hypothesized to have positive effect on the purchase behavior of blog audience.

• H5: There is a positive relationship between high frequency of visiting a beauty blog

and the likelihood of purchasing products from that blog recommendations.

2.4 Income of blog audience

Although not depicted in the research model, income is a demographic factor that is used as a control variable in the analysis. Consumers with higher income are likely to have higher discretionary income. This is the amount of income remain after paying all taxes and essential costs to maintain certain standard of living, such as mortgage, food, medical, insurance, or transportation. It is likely that people with high discretionary income can afford to spend more money on beauty products.

According to a survey in 2015 by Q&Me, a marketing research service in Vietnam, there is a correlation between income and spending on cosmetics. Those who earn less than five million VND monthly spend averagely 110,000 VND, while people who earn more than five million VND monthly spend averagely 190,000 VND on cosmetics per month. Similarly, a positive relationship between income and the likelihood of purchasing products from beauty blog recommendations is expected to be found in this research.

(16)

10

The above-mentioned hypotheses are graphically presented in the following model: Figure 1. Research model

(17)

CHAPTER 3. DATA AND METHOD 3.1 Method

This quantitative research was carried out using a cross-sectional survey. It is useful for statistical analyses; the findings can be generalized on an aggregate level, or provide insights about the performance of individual bloggers.

The population for this study was Vietnamese beauty blog audience. The survey was posted online and the population was encouraged to participate, i.e. volunteer sampling by self-selection. Online survey is especially appropriate in this case because the research population is certainly Internet users. Thus, the risk of having a biased sample because not everybody has access to WWW (Couper, 2000) is not relevant.

There were two methods used to increase the participation rate. Firstly, the author of this study personally contacted several popular Vietnamese beauty bloggers about this research. A number of bloggers, including those with more than forty thousands (40k) followers, agreed to support the research by sharing and encouraging their audience to participate in the online survey. Secondly, respondents were informed that valuable cosmetic items could be won by finishing all the questions.

The questionnaire distributed to blog audience consisted of four main parts. Part I asked questions related to the engagement of the respondents with a beauty blogger of their choice. Part II asked questions about the purchase(s) that the respondents have made in the past based on this blogger’s reviews. Part III was the personality test to determine whether a respondent is neuroticism or extraversion. This 12-question personality test was the “Improved short questionnaire for the measurement of extraversion and neuroticism” by Eysenck & Eysenck (1964). The short version was used to ensure a good response rate, as various researches (cited in Fan & Yan, 2009) show that the length of the survey has a

(18)

12

negative linear relation with response rates. Lastly, in part IV, respondents could fill in their demographic information.

The questionnaire was written in Vietnamese because the first language of all respondents was Vietnamese. The original personality test by Eysenck & Eysenck was first translated into Vietnamese by the author of this thesis. In order to ensure that the content of the questions remains unaffected, the translated test was back translated into English by a professional translator. The author worked together with the professional translator to correct the discrepancies and create the final Vietnamese version of the personality test. An English version of the full survey is included in the appendix of this thesis.

(19)

3.2. Measures

The survey was run for one month in December 2015. A frequency check was performed to identify and delete incomplete responses. At the beginning of the survey, respondents could name one blogger that they wanted to discuss throughout all questions. Forty beauty bloggers were named among 200 valid responses; some were more popular than the others. The most frequently cited blogger accounted for 23.5% responses, and the least cited blogger accounted for 0.5% responses. The dataset consisted of 100 respondents who had never and 100 respondents who had purchased beauty products from a blogger’s recommendation.

The main variables of interest are as follows:

Purchased:

Respondents were asked if they had ever purchased an item after reading the review written by the above-mentioned beauty blogger. The answers could be Yes (SPSS value = 1), or No (SPSS value = 0)

Personality

The personality test by Eysenck & Eysenck (1964) consisted of twelve questions; each corresponded to a “key” – Extraversion or Neuroticism. The result was calculated by giving one point for agreement with the key, and no points for disagreement. Based on the answer to the 12-question personality test, each respondent were classified as Extraversion (SPSS value = 1), Neuroticism (SPSS value = 2). Furthermore, dummy variables Ext and Neu were created.

Writing style

The author of this thesis and Happy Skin team evaluated the writing style of forty mentioned bloggers. Happy Skin is a Vietnamese online magazine that specializes in skin care and makeup. Their team consisted of beauty editors, makeup artists, estheticians, and

(20)

14

dermatologists. During the survey, respondents were also asked to evaluate whether the blogger of their choice wrote scientifically or emotionally. There is very little difference between experts’ and respondents’ evaluation, showing that most blog audience are able to identify the writing style of beauty bloggers. Each blogger’s writing style was either Scientific (SPSS value = 1) or Emotional (SPSS value = 2). Dummy variables Sci and Emo were also created.

Correspondence between personality and writing style

Dummy variable NeuSci was assigned SPSS value = 1 for cases where there was a match: the respondent was a neurotic, and the blogger that she mentioned had scientific writing style. Dummy variable ExtEmo was created by similar mechanism: the respondent was an extravert, and the blogger that she mentioned had emotional writing style.

Frequency of blog visit

Respondents were asked to explain how often they visited the cited blog in their own words. The author of this thesis then converted the given answers into “Frequency per month”. For example, if the respondent claimed to visit the blog once every two months, her frequency per month was 0.5; if the respondent claimed to visit the blog everyday, her frequency per month was 30. The four levels of frequency per months were:

• 0.5 – 9 times per month (SPSS value = 1, representing Low Frequency) • 10 – 19 times per month (SPSS value = 2, representing Medium Frequency) • 20 – 29 times per month (SPSS value = 3, representing Medium High Frequency) • More than 30 times per month (SPSS value = 4, representing High Frequency)

(21)

Based on the frequency of blog visits, three dummy variables were created for Medium, Medium High, and High Frequency. Their names were FreqMed, FreqMedHigh, and FreqHigh, respectively.

Income level

Income might be a sensitive topic for respondents to give details about. Thus, four income brackets were given, and respondents could select the bracket that was most relevant to them. As the survey was distributed among Vietnamese respondents, the currency is Vietnam Dong (VND). The four income brackets are:

• 0 – 4,499,999 VND (SPSS value = 1, representing Low Income)

• 4,500,00 – 14,999,999 VND (SPSS value = 2, representing Medium Income)

• 15,000,000 – 29,999,999 VND (SPSS value = 3, representing Medium High Income) • 30,000,000 VND and above (SPSS value = 4, representing High Income)

Based on the income level, three dummy variables were created for Medium, Medium High, and High Income. Their names were IncMed, IncMedHigh, and IncHigh, respectively.

Gender

The respondents were asked to identify themselves as Male (SPSS value = 1), or Female (SPSS value = 2).

Age:

The respondents were asked to give their year of birth in four digit numbers. The Birth Year variable was recoded into Age by subtracting the Birth Year from 2015.

(22)

16 CHAPTER 4. RESULTS

4.1 Descriptive Statistics

Besides having an equal amount of respondents who have never and who have purchased beauty products from a blogger’s recommendation, the dataset is a good representative because there are also roughly equal shares of personality and writing style: 49% of the respondent is neurotic, and 51% is extravert; 51% of the writing style is scientific, and 49% is emotional. There are 59 cases (29.5%) where there is a match between neuroticism and scientific writing style; and 63 cases (31.5%) where there is a match between extraversion and emotional writing style.

Regarding frequency of blog visits, there is a large difference among the readers (MFreqPerMonth = 13.27, SDFreqPerMonth = 20.14, frequency rage: 0.5-240 times per month); the most common frequency is four times per month (39%). There is a serious skewness in gender and age: 99% of respondents are female, and most of them are at young age (Mage = 21.24, SDage = 3.84, age-rage: 13-31 years old). Due to this skewness, gender and age are not used as control variable in the testing.

Table 1. Means, Standard Deviations, Correlations (N = 200, rounded up to two decimal places)

* Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).

Variables M SD 1 2 3 4 5 6 7 1. Purchased 0.50 0.50 - 2. Personality 1.49 0.50 -.14* - 3. Writing style 1.51 0.50 -.06 -.22** - 4. Frequency of blog visits 13.27 20.14 .22** .07 -.16* - 5. Income level 1.37 0.64 0.32** -.17* .01 -.11 - 6. Gender 1.99 0.10 -.10 -.00 .00 -.02 .06 - 7. Age 21.24 3.84 .26** -.03 -.14* -.03 .52** .12 -

(23)

4.2 Hypotheses Testing

4.2.1 Neuroticism and scientific writing style

A hierarchical multiple regression was performed to investigate the effect of neuroticism on the purchase behavior and the moderating effect of scientific writing style on this relationship. In the first step of the hierarchical multiple regression, three income dummy variables (IncMed, IncMedHigh, and IncHigh) were enter. This model was statistically significant F (3,196) = 9.49; p = .00 and explained 12.69% of variance in the likelihood of making a purchase from beauty blog recommendation. In the second step, three independent variables Neuroticim (Neu), Scientific writing style (Sci), and the match (NeuSci) were entered. The total variance explained by the model as a whole was 15.82% with F (6,193) = 6.05; p = .00. The added variables explained 3.14% more variance in the purchase behavior, after controlling for income. (R2_{change = .03, F change (3,193) = 2.40, p = .07). In the final} model, four out of six predictor variables were statistically significant: IncMed (β = .36, p = .00), IncMedHigh (β = .36, p = .02), Neu (β = -.24, p = .02), and NeuSci (β = .29, p = .03).

(24)

18

Table 2. Hierarchical regression model: The relationship between neuroticism, scientific writing style, and the likelihood of purchase

(Rounded up to two decimal places)

R R2 _R2 change B SE β t Step 1 .36 .18 IncMed .39 .08 .33 4.84 IncMedHigh .34 .15 .16 2.30 IncHigh .61 .34 .12 1.82 Step 2 .40 .16 .03 IncMed .36 .08 .31 4.46 IncMedHigh .36 .15 .17 2.46 IncHigh .55 .33 .11 1.65 Neu -.24 .10 -.24 -2.43 Sci -.06 .10 -.06 -.65 NeuSci .29 .14 .27 2.15

(25)

4.2.2 Extraversion and emotional writing style

A similar hierarchical multiple regression was performed to investigate the effect of extraversion on the purchase behavior and the moderating effect of emotional writing style on this relationship. As explained in part 4.2.1 above, in the first step of the hierarchical multiple regression, three income dummy variables (IncMed, IncMedHigh, and IncHigh) were enter. This model was statistically significant F (3,196) = 9.49; p = .00 and explained 12.69% of variance in the likelihood of making a purchase from beauty blog recommendation. In the second step, three independent variables Extraversion (Ext), Emotional writing style (Emo), and the match (ExtEmo) were entered. The total variance explained by the model as a whole was 15.82% with F (6,193) = 6.05; p = .00. The added variables explained 3.14% more variance in the purchase behavior, after controlling for income. (R2 change = .03, F change (3,193) = 2.40, p = .07). In the final model, four out of six predictor variables were statistically significant: IncMed (β = .36, p = .00), IncMedHigh (β = .36, p = .02), Emo (β = -.23, p = .02), and ExtEmo (β = .29, p = .03).

(26)

20

Table 3. Hierarchical regression model: The relationship between extraversion, emotional writing style, and the likelihood of purchase

R R2 _R2 change B SE β t Step 1 .36 .18 IncMed .39 .08 .33 4.84 IncMedHigh .34 .15 .16 2.30 IncHigh .61 .34 .12 1.82 Step 2 .40 .16 .03 IncMed .36 .08 .31 4.46 IncMedHigh .36 .15 .17 2.46 IncHigh .55 .33 .11 1.65 Ext -.06 .10 -.06 -.60 Emo -.23 .10 -.23 -2.39 ExtEmo .29 .14 .27 2.15

(27)

4.2.3 Frequency of blog visits

A similar hierarchical multiple regression was performed to investigate the relationship between frequency of visiting a blog and the likelihood of making a purchase from that blog recommendation. Three dummy variables of income levels were also used as control variables. In the second steps, three dummy variables of frequencies were entered: FreqMed, FreqMedHigh, and FreqHigh. The total variance explained by the model as a whole was 25.18% with F (6,193) = 10.83; p = .00. The added variables explained 12.49% more variance in the purchase behavior, after controlling for income. (R2 change = .12, F change (3,193) = 10.74, p = .00). In the final model, five out of six predictor variables were statistically significant: IncMed (β = .44, p = .00), IncMedHigh (β = .47, p = .00), FreqMed (β = .29, p = .01), FreqMedHigh (β = .77, p = .02), and FreqHigh (β = .36, p = .00).

(28)

22

Table 4. Hierarchical regression model: The relationship between frequency of blog visits and the likelihood of purchase

R R2 _R2 change B SE β t Step 1 .36 .18 IncMed .39 .08 .33 4.84 IncMedHigh .34 .15 .16 2.30 IncHigh .61 .34 .12 1.82 Step 2 .50 .25 .12 IncMed .44 .08 .37 5.74 IncMedHigh .46 .14 .21 3.31 IncHigh .622 .32 .12 1.97 FreqMed .30 .11 .18 2.71 FreqMedHigh .77 .32 .15 2.44 FreqHigh .36 .07 .33 5.06

(29)

CHAPTER 5. DISCUSSION

5.1 Theoretical and practical implications

The result from the 4.2.1 confirms that there is a negative relationship between neuroticism and the likelihood of purchasing products from beauty blog recommendations (β = -.24); moreover, this relationship is significant (p =.02). Interestingly, in cases where the respondent is a neurotic, and the blogger that she mentioned has scientific writing style, the likelihood of making a purchase is positive and statistically significant (β = .29, p = .03). Thus, the first and the third hypothesis are supported.

The result from 4.2.2 does not show a significant impact of extraversion on the likelihood of making a purchase, meaning the second hypothesis is not supported. In cases where the respondent is an extravert, and the blogger that she mentioned has an emotional writing style, the likelihood of making a purchase is positive and statistically significant (β = .29, p = .03). Thus, the forth hypothesis is partially supported: extraverts are likely to make purchase based on emotionally written recommendations; but emotional writing style could not be proven to enhance the unestablished relationship in H2.

For beauty bloggers and companies who collaborate with them for promotion, these outcomes suggest that each writing style is appealing to a different type of audience; no writing style is necessarily better than the other. However, when checking the relationship between the writing style and the amount of money spent on purchasing the recommended product, the author of this thesis found that there is a positive relationship between scientific writing style and the purchase of more expensive products. This finding can be explained in the sense that when buying more expensive products, consumers tend to look for more solid evidence of the high performance, rather than just taking someone’s objective feeling about it. Therefore, when promoting expensive products, beauty bloggers and companies are encouraged to write the review in a scientific way.

(30)

24

The result from 4.2.3 partially supported the fifth hypothesis: to a certain degree, there is a positive relationship between high frequency of visiting a beauty blog and the likelihood of purchasing products from that blog recommendation. When there is an increase from medium frequency (10-19 times per month) to medium high frequency (20-29 times per month), there is a statistically significant increase in the likelihood of making purchase. (βFreqMed = .29, pFreqMed = .00; (βFreqMedHigh = .77, pFreqMedHigh = .02). A possible explanation why there is no significant increase of purchase by the group with highest blog visit frequency (more than 30 times per month) is that if a person could afford to spend so much time on reading beauty blogs, she is probably not working or only works part time, and hence does not have much income to spend on beauty products. Indeed, there is a negative correlation between frequency of blog visit and income level, as displayed in Table 1.

The proposed models could explain only a small percentage of variance in the likelihood of making a purchase, suggesting that there were other reasons that encouraged a reader to buy a beauty product. As part of the online survey, the open-ended question:

“Besides the recommendation of this blogger, were there other factors that lead to your purchase?” allowed a respondent to mention the reasons behind every purchase. Among 193

purchases made by 100 respondents (one respondent could list more than one product that she had purchased from the recommendation of a beauty blogger), only 6.73% was solely accountable to the recommendation of a single blogger. 29.53% was made after looking at various reviews written by different beauty bloggers. 30.57% was made because the respondents just happened to come across the recommended product when they urgently needed a product with similar function. Other reasons for purchasing that were listed by respondents included: limited-time discounted price, beautiful packaging, or not having such color of lipstick in their collection. Note that these factors were not mutually exclusive because a respondent could mention more than one reason. These insights suggest that a

(31)

many customers look for more than one review before buying a product. This may has an implication for companies who use beauty bloggers in their promotion campaigns: companies should collaborate with several beauty bloggers to create different sources of information about one product. In addition, a beauty blogger should not be discouraged in reviewing a product that has been discussed by other bloggers; blog audience would like to see more than one opinion.

(32)

26 5.2 Strengths and Limitations

A noticeable strength of this research is that it examined the actually purchase behavior in the past instead of purchasing intention. Since purchase intention does not necessarily lead to actual purchase behavior (Chandon, Morwitz & Reinartz, 2005), a study of past purchase is more accurate in reporting the influence of beauty bloggers. The ability for respondents to disclose their purchase based on an individual blogger’s review can be very useful to also analyze the result at the individual blogger level. Each blogger will receive a customized report about her audience in return for supporting this thesis.

However, there are also some limitations. Firstly, a cross-sectional research can study a particular phenomenon only at a certain point of time, and the results do not allow interpretations about causal effects. Secondly, the survey is dependent on the ability of the respondent to recall their previous purchases. Moreover, as respondent filled in the survey themselves at their convenient time and location, the surrounding environment while answering the questions could not be controlled. Last but certainly not least, this research only looked at two writing styles and two personality types that are mutually exclusive, making the result of the first two regressions very similar to each other. As explained earlier, the short test to distinguish neuroticism and extraversion was used to keep the survey shorter, since the length of surveys has a negative linear relation with response rates. Future research about the links between personality and the likelihood of making purchase should entail a more extensive test to identify more personality types, such as openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism.

(33)

CHAPTER 6. CONCLUSIONS 6.1 Conclusions

To answer the question “To what extent do beauty bloggers influence the purchase behavior

of their audience?”, a quantitative research was carried out using a cross-sectional survey

distributed among Vietnamese blog audience. The findings reveal that there are links between personality, blog writing style, frequency of visits, and the likelihood of making a purchase from blog recommendation in Vietnam. There is a negative relationship between neuroticism and the likelihood of making a purchase, but this relationship becomes positive under the moderating effect of scientific writing style. There is also a positive relationship between extraversion and the likelihood of making a purchase under the moderating effect of emotional writing style. To a certain degree, there is a positive relationship between the high frequency of visiting a beauty blog and the likelihood of purchasing products from that blog recommendation.

6.2 Recommendations for future research

Firstly, to overcome the limitation mentioned in 5.2, future research about the links between personality and the likelihood of making purchase should look at a variety of personality types, such as openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism. Besides the likelihood of making a purchase from blog recommendation, research can also look at the amount of money spent on such purchase. Lastly, this research was carried out among Vietnamese blog audience; similar studies should be replicated in different countries, such that the results can be highly generalized.

(34)

28 REFERENCES

Agarwal, N., Liu, H., Tang, L., & Yu, P. S. (2008, February). Identifying the influential bloggers in a community. In Proceedings of the 2008 international conference on web search and data mining (pp. 207-218). ACM.

Beauty and Personal Care in Vietnam (Rep.). (2015, September). Retrieved January 10, 2016, from Euromonitor International website: http://www.euromonitor.com/beauty-and-personal-care-in-vietnam/report

Calfee, J. E., & Ringold, D. J. (1994). The 70% majority: Enduring consumer beliefs about advertising. Journal of public policy & marketing, 228-238.

Chandon, P., Morwitz, V. G., & Reinartz, W. J. (2005). Do intentions really predict behavior? Self-generated validity effects in survey research. Journal of Marketing, 69(2), 1-14.

Couper, M. P. (2000). Review: Web surveys: A review of issues and approaches. Public opinion quarterly, 464-494.

Dawar, N., & Parker, P. (1994). Marketing Universals: Consumers' Use of Brand Name, Price, Physical Appearance, and Retailer Reputation as Signals of Product Quality. Journal of Marketing, 58(2), 81-95.

Dean, D. H. (1999). Brand Endorsement, Popularity, and Event Sponsorship as Advertising Cues Affecting Consumer Pre-Purchase Attitudes. Journal of Advertising, 28(3), 1-12.

(35)

Ellison, N. B., & Boyd, D. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230.

Evans, A. M., & Revelle, W. (2008). Survey and behavioral measurements of interpersonal trust. Journal of Research in Personality, 42(6), 1585-1593.

Eysenck, H. J. (Ed.). (1981). A model for personality. New York: Springer.

Eysenck, H. J. (2013). The Structure of Human Personality (Psychology Revivals). Routledge.

Eysenck, S. B., & Eysenck, H. J. (1964). An improved short questionnaire for the measurement of extraversion and neuroticism. Life sciences, 3(10), 1103-1109.

Fan, W., & Yan, Z. (2010). Factors affecting response rates of the web survey: A systematic review. Computers in human behavior, 26(2), 132-139.

Flanagin, A. J., & Metzger, M. J. (2007). The role of site features, user attributes, and information verification behaviors on the perceived credibility of web-based information. New Media & Society, 9(2), 319-342.

Herring, S. C., Scheidt, L. A., Wright, E., & Bonus, S. (2005). Weblogs as a bridging genre. Information Technology & People, 18(2), 142-171.

(36)

30

Jeronimus, B. F., Riese, H., Sanderman, R., & Ormel, J. (2014). Mutual reinforcement between neuroticism and life experiences: A five-wave, 16-year study to test reciprocal causation. Journal of Personality and Social Psychology, 107(4), 751-764.

Johnson, T. J., & Kaye, B. K. (2004). Wag the Blog: How reliance on traditional media and the Internet influence perceptions of credibility of Weblogs among blog users. Journalism & Mass Communication Quarterly (81) 622–642.

Kaye, B., & Johnson, T. (2011). Hot Diggity Blog: A Cluster Analysis Examining Motivations and Other Factors for Why People Judge Different Types of Blogs as Credible. Mass Communication and Society, 236-263.

Labrecque, L. I., vor dem Esche, J., Mathwick, C., Novak, T. P., & Hofacker, C. F. (2013). Consumer power: Evolution in the digital age. Journal of Interactive Marketing, 27(4), 257-269.

Li, J., & Chignell, M. (2010). Birds of a feather: How personality influences blog writing and reading. International Journal of Human-Computer Studies, 68(9), 589-602.

Makeup survey among Vietnamese (Rep.). (2015, February 12). Retrieved January 10, 2016, from Q&Me/Asia Plus Inc. website: http://qandme.net/vietnam-report/Makeup-survey-among-Vietnamese.html#links

Opportunities in the Vietnamese Cosmetics Market: Q&A with Le Chau Giang. (2015, October 13). CGI Magazine. Retrieved January 10, 2016, from http://www.gcimagazine.com/marketstrends/regions/asiapacific/279013281.html

(37)

Qader, I. K., & Zainuddin, Y. B. (2011). The Impact of Media Exposure on Intention to Purchase Green Electronic Products amongst Lecturers. IJBM International Journal of Business and Management, 6(3), 240-248.

Robertson, J., Fieldman, G., & Hussey, T. (2008). “Who wears Cosmetics?” Individual Differences and their Relationship with Cosmetic Usage. Individual Differences Research, 6(1), 38-56.

Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not So Different After All: A Cross-Discipline View Of Trust. Academy of Management Review, 23(3), 393-404.

Senecal, S. & Nantel, J. 2004. The influence of online product recommendations on consumers' online choices. Journal of Retailing (80), 159–169.

Smith, A. N., Fischer, E., & Yongjian, C. (2012). How does brand-related user-generated content differ across YouTube, Facebook, and Twitter?. Journal of Interactive Marketing, 26(2), 102-113.

Smith, B. G. (2011). Becoming―quirkyǁ towards an understanding of practitioner and blogger relations in public relations. Public Relations Journal,5(4), 1-17.

Soest, T. V., Kvalem, I. L., Skolleborg, K. C., & Roald, H. E. (2009). Cosmetic surgery and the relationship between appearance satisfaction and extraversion: Testing a transactional model of personality. Journal of Research in Personality, 43(6), 1017-1025.

(38)

32

Wang, C. C., & Chien, H. Y. (2012). Believe or Skepticism? An Empirical Study on Individuals’ Attitude to Blog Product Review. International Journal of Innovation, Management and Technology, 3(4), 343-348.

Weaver, J. B. (1991). Exploring the links between personality and media preferences. Personality and Individual Differences, 12(12), 1293-1299.

Zhu, J. and Tan, B. (2007). Effectiveness of Blog Advertising: Impact of Communicator Expertise, Advertising Intent, and Product Involvement. ICIS 2007 Proceedings. Paper 121. Retrieved from http://aisel.aisnet.org/icis2007/121

(39)

Appendix A

IMPROVED SHORT QUESTIONNAIRE FOR THE MEASUREMENT OF

EXTRAVERSION AND NEUROTICISM (Eysenck & Eysenck, 1964) 1. Do you like plenty of excitement and bustle around you? 2. Does your mood often go up and down?

3. Are you rather lively?

4. Do you ever feel "just miserable" for no good reason? 5. Do you like mixing with people?

6. When you get annoyed do you need someone friendly to talk to about it? 7. Would you call yourself happy-go-lucky?

8. Are you often troubled about feelings of guilt?

9. Can you usually let yourself go and enjoy yourself a lot at a gay party? 10. Would you call yourself tense or "highly strung”?

11. Do you like practical jokes? 12. Do you suffer from sleeplessness?

(40)

34 Appendix B

SURVEY QUESTIONNAIRE (Translation into English)

Welcome to the survey. Upon completion of this survey, you will be automatically registered for a lottery from which you can win valuables cosmetics items. The winner will be contacted by the email address entered at the end of this survey.

Part 1 - Questions about engagement of the respondents with the bloggers:

1.1. Please name the beauty blogger you would like discuss throughout this survey. This blogger must be a person who writes her blog in Vietnamese language.

1.2. Are you connected with this beauty blogger on any social channels? (like on Facebook page, follow on Instagram, or subscribe on Youtube)

1.3. How often do you read/watch this blogger’s reviews?

1.4. Do you consider this blogger’s writing style to be more on the scientific side or more on the emotional side?

1.5. Have you ever purchased a product after seeing this blogger’s review? o Yes à Proceed to Part 2, Part 3, and Part 4.

o No à Proceed to Part 3 and Part 4.

Part 2 - Questions about purchased product(s):

2.1. What is the full name of the product including brand, name, and function? (For example: Estee Lauder multi-action facial cleanser)

2.2. How much did you pay for this product?

2.3 Besides the reviews by the blogger, were there other factors that lead to the purchase? 2.4. How important was the blogger’s review to your purchasing decision? (Scale 1-5, not important at all – extremely important)

Each respondent can list more than one product; for each product, respondents must answer the same question as 2.1-2.4.

(41)

Part 3 - Questions about respondent’s personality

As in Appendix A - Improved short questionnaire for the measurement of extraversion and neuroticism (Eysenck & Eysenck, 1964)

Part 4 - Questions about respondent’s personal information: 4.1. What is your gender?

4.2. What is your year of birth? 4.3. What is your income level?

o 0 - 4,499,999 VND

o 4,500,000 - 14,999,999 VND o 15,000,000 - 29,999,999 VND o 30,000,000 VND trở lên

4.4. What is your email address? This address will be used to contact you in case you are the lucky winner.

(42)

36 Appendix C

LETTER TO BEAUTY BLOGGERS TO ASK FOR SUPPORT (Translation into English) Dear [name of beauty blogger],

My name is Mai Nguyen. I am preparing for my Master thesis at University of Amsterdam. I would like to examine the influence of beauty bloggers on cosmetics purchasing decisions. I am contacting you and other beauty bloggers to ask for support.

An online survey will be published with questions regarding the engagement with the blog, the purchasing behavior in the past, the personality type test, and personal information. I would greatly appreciate if you could share my survey on your channels to encourage your audience to participate. For your benefit, after analyzing the answers, I will provide you with a customized report about your audience and your influence.

The information in this email is my preliminary intentions. Minor changes can be made later. However, I promise the followings:

1. I will always provide you with information and honest analysis of your audience.

2. This project is solely to support my Master thesis; I will not commercialize or earn profit from it in any way, shape, or form.

I am looking forward to your response. Wish you all the best! Best regards,

To what extent do beauty bloggers influence the purchase behavior of their audience : wxploring the links between personality and blog preferences