1. Introduction

Female consumption decision-making is a complex process that can be influenced by a combination of psychological and social factors, such as personal motivation, peer pressure, and fashion trends. One element that stands out for its symbolic power in the luxury industry is the logo. A logo is a distinctive visual symbol or design adopted by a company to identify its products, differentiate its products from competitors, and convey brand values.

From a social signaling perspective, logos operate as a quick recognizable marker that shows one’s economic resources and social status to others [1]. On social media, this signaling power can be amplified, as an image featuring a luxury product with a logo can be shared and evaluated rapidly, gaining public visibility and approval through likes and comments, which in turn can affect female consumers’ desire to purchase.

Our study builds upon the work of Chen and Zhuang from Zhejiang Sci-Tech University and Zhejiang University. They investigated the motivations and underlying mechanisms of Chinese consumers’ decision-making regarding luxury fashion products in the age of social media. They identified two key drivers of trend conformity behavior: trend perception and reference group pressure [2]. These external stimuli influence whether consumers want to fit in socially, therefore leading to both rational and impulsive purchasing behavior, such as demand amplification and urge to buy impulsively (UBI). Their findings show that external stimuli can operate as both informational cues and social norms.

Our study extends their work by narrowing the focus on two groups of Chinese female consumers, students (aged between 15 to 24) and young working professionals (aged between 25 to 40). We aim to examine how specific social media factors influence their evaluation of luxury products. Here, social media refers to “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of user-generated content.” [3]. This highlights the role of interactive social platforms where user-generated content is regularly shared and posted. We investigate the effects of brand logo and visible number of likes on social media posts as independent variables (IVs). The dependent variables (DVs) in our study are consumers’ likability of luxury products and their willingness to purchase.

According to Nguyen, Melewar and Chen (2013), likability is defined as the degree of perceived appeal a consumer has for a brand or product, capturing how attractive or appealing the consumer finds the product [4]. Willingness to purchase refers to” the willingness or unwillingness of a person to buy or procure a good or service” [5].

2. Methodology

This study employed quantitative methods, particularly questionnaires, to investigate the motivations of female purchasing behavior toward luxury handbags. Four distinct questionnaires were designed, each featuring unique visual stimuli to identify the impact of brand logos and social media likes on consumer preferences. Responses were collected using WenJuanXing and social media platforms, such as WeChat.

2.1. Handbag image rating under different conditions

For this section, the theme is to study the influence of logos and likes on females’ purchasing desires. The questionnaire is set as a single comprehensive one. After entering the questionnaire page, four options- Questionnaire 1, Questionnaire 2, Questionnaire 3, and Questionnaire 4 – will appear. Clicking on the corresponding option will redirect you to the respective questionnaire for answering. Participants were randomly assigned to one of four experimental groups, each exposed to handbag images with varying attributes: Images with no logos and no social likes (questionnaire A – Neither Likes or Logos ) . Images displaying likes only (questionnaire B – Only Likes). Images featuring logos only. (questionnaire C – Only Logos). Images with both logos and likes (questionnaire D – Both Logos and Likes).

Participants rated each handbag image using a 7-point Likert scale (Scale: 1 = least; 7 = most) across two dimensions: Likability, meaning the level of attractiveness of the handbag ("How much do you like this handbag?"), and Willingness to Purchase, meaning the desirability of the handbag ("How likely would you be to buy this handbag?").

2.2. Section two – females' consumption notions

For this section, the study of females’ consumption notions as the main research area. Subjects evaluated nine statements reflecting three types of purchasing motivations. Responses were recorded on a 5-point Likert scale. (Scale: 1 = strongly disagree; 5 = strongly agree).

Construct Statements: Follow the Trend, which means that the 'Following Trends' dimension measured female consumers' susceptibility to bandwagon consumption. Seek Information, which means that drop formal words like "seek" – use action verbs like check, find, or get for clarity. Personal Motivation, which means that reasons consumers genuinely want to buy include needs, personal liking, or habits.

The research targeted two distinct female demographic segments: students and young working professionals, enabling comparative analysis of purchasing values across social roles.

3. Data analysis

3.1. T-test between Likability and Willingness to Purchase rating

The analysis started by conducting a T-test between the variables Likability and Willingness to Purchase to explore any potential relationship. The results show that, for all four questionnaire conditions: Having both Logos and Likes, Only Likes, Only Logos, and Neither Likes nor Logos, the p-value is below .001, indicating a significant difference between Likability and Willingness to Purchase for all conditions.

3.2. Likability ANOVA analysis

One-way ANOVA was further employed to investigate if any of the variables had statistical significance. For both variables, the p-values are below the significant threshold p=.05, with Likability showing a p-value under .001 and Willingness to Purchase at .002. These finding states that, for each of the variables, at least one of the four questionnaire conditions significantly influenced the variable rating. The Post Hoc test method in ANOVA further identified the specific conditions that are influential.

The four questionnaire conditions were organized into three comparison groups, where condition Only Likes, Only Logos, and Both Likes and Logos were compared with condition Neither Likes or Logos, respectively. Variables Likability and Willingness to Purchase were both analyzed from these three perspectives.

In the case of Likability, group Only Likes compared to Neither Likes or Logos and Only Logos compared to Neither Likes or Logos both show statistical significance, with p-values below the threshold .05. Such a result aligns with the hypothesis. However, contradictory to the hypothesis, the group Both Likes and Logos, compared to Neither Likes nor Logos, shows no statistical significance, with a p-value of .057 above the threshold.

3.3. Willingness to Purchase ANOVA analysis

As for variable Willingness to Purchase, only the group Only Likes compared to Neither Likes or Logos demonstrates a significant statistical difference, with a p-value of .001. However, the p-values are greater than the significance threshold p=.05 for the other two groups: Only Logos compared to Neither Likes or Logos, and Both Likes and Logos compared to Neither Likes or Logos. This indicates that no statistically significant difference lies between the two comparisons, which contradicts the hypothesis.

3.4. Statement rating T-test 

As for the second section of the questionnaires --- the statement rating section, the nine statements were organized into three categories: Personal Motivation, Seek For Information, and Following Trends.

To identify statistical differences, a T-test was conducted between the two investigated social groups (students and young working professionals) for each statement category. As demonstrated in the collected data, only the Personal motivation category yielded a significant p-value of .033, which is below the significant threshold.

4. Discussion

4.1. Product Likability and Willingness to Purchase rating results

After breaking down all test results for Likability and Willingness to Purchase, the basic conclusion can be illustrated. The statistically significant data in the ANOVA for Likability demonstrates that adding elements Likes (questionnaire Only Likes) and Logos (questionnaire Only Logos) separately highly influenced the variable rating, compared to when no additional factors were present (questionnaire Neither Likes nor Logos). Such a result aligns with our hypothesis. However, contradictory to our hypothesis, the two elements didn’t have a significant effect on the Likability rating when presented simultaneously.

Further, the data for the two statistically significant groups also demonstrates that incorporating Likes or Logos increased the average Likability rating. This suggests that Likes/Logos can elevate the average level of attractiveness of given products. However, further investigations are needed to consolidate this conclusion.

Similar to the findings for Likability, the data for Willingness to Purchase indicate that the element Likes significantly influenced the Willingness to Purchase of respondents. This again suggests that Likes increased the mean rating for Willingness to Purchase, implying that it acts as an encouragement for purchasing decisions. Further investigations are also necessary for a solid conclusion.

As for the other two comparison groups, the absence of significant data indicates that neither Logos nor the combination of Likes and Logos has a significant influence on encouraging the consumption of presented products. This again contradicts the hypothesis.

Overall, to answer our research question, female consumers are influenced by conformity pressure when assessing the level of attractiveness of a product, particularly by social likes and brand status. However, according to our research, only social likes contribute as motivation for actual consumption, as they increase the desirability of the product. This highlights that female consumers are driven by social approval in their purchasing decisions.

4.2. Purchasing value statement rating results

Since the significant data only appeared in the category 'Personal Motivation’, a conclusion can be made that the statistical difference between the ratings of the two social groups ---- students and young working professionals- lies only in the context of this category. From this result, further conclusions can be obtained, suggesting that the value differences between the two groups also lie solely in the category of 'Personal Motivation’.

4.3. Limitations and future insights

The key limitation of the research is connected to the insignificant results obtained from our investigation: the insignificant influence Logos has on motivating consumption. Reflecting on the study, this is likely due to the different layers of influence within the single factor -- Logos. Elements such as price acceptability, brand reputation, brand design, and influence of celebrity endorsement might have already left stereotypes for the respondents in their lives. Such could’ve caused extreme brand loyalty or specific hatred, either one leading to rating decisions that were not made under the mere context of conformity level. Therefore, the results from this part of the study should be re-experimented with better methods.

Looking ahead, the future of this study will build on the limitations and further investigate the influence of Logos. Additional factors listed above can stem from the element Logos and form multiple independent variables, which can be investigated separately to form more considerate conclusions. Incorporating qualitative methods, such as open-ended questions and interviews, could also deepen our understanding of why Logos are insignificant in pressuring consumption. Ideally, these improvements will enhance the exploration of female consumer motivations and enrich the scientific contribution of the study.

Acknowledgement 

Mingyi Du, Zhu Gong, and Zhiyi Zhang contributed equally to this work and should be considered co-first authors.