Download
Introduction
This paper investigates the monetary value of privacy in second-hand clothing markets. It aims to analyse how privacy considerations shape consumer behaviour and influence market transactions, with a specific focus on the role of image sharing in driving sales and fostering trust.
Research Objective
The primary goal of this research is to explore how the use of personal images, professional models, or catalogue images compares to simple product photos in influencing sales outcomes. Specifically, the study aims to answer:
- What drives a user to choose one type of picture over another?
- How does the effort invested in pictures affect sales outcomes?
Methodological Approach
To achieve this, the research utilizes a robust dataset from a popular second-hand clothing platform, Tise. This dataset includes detailed information on the items on sale, user characteristics, and reviews. A key aspect of the study involves analysing the behaviour of consumers when they decide whether to include personal images (such as photos of themselves wearing the clothes) in their listings.
Additional variables were created to enhance the analysis:
- Profile picture variables: Presence of a person in the picture and assessment of perceived attractiveness.
- Ad image categorization: Categorizing images into Homemade Picture, Real Person, Catalogue Image, or Professional Model.
- Ad picture quality: Evaluating the quality of the images used, using No Reference Image Quality Metrics (IQA).
Data Analysis
- Logit Regression: This statistical method was applied to examine the drivers influencing the choice of picture category and the subsequent sales outcomes. The analysis focuses on understanding the economic impact of listing image choices, particularly in relation to privacy concerns.
Privacy Valuation:
- Measuring the implicit value users place on privacy regressing sales on image category, controlling for other as and users’ characteristics.
Consumer Decision-Making:
- Exploring the decision-making processes behind why sellers include personal photos in their listings.
Summary of Findings
Catalogue and professional model images enhance perceived quality, with catalogue images boosting sales for tops (0.365***) and jackets (0.234***), and professional model images excelling for dresses (0.585***). Real-person images show mixed results, positively impacting dress sales (0.352***) but negatively affecting sweaters (-0.124***) and pants (-0.137***).
Additionally, the number of images consistently improves sales, with each extra image increasing the likelihood of sales for sweaters (0.070***) and pants (0.052***). These findings highlight the importance of strategic image use, helping sellers and platforms optimize listings to drive sales.
Dependent Variable: Sold Outcome (0/1)
| Category | Full Sample | Sweaters | Tops | Pants | Dresses | Jackets |
|---|---|---|---|---|---|---|
| Real Person | 0.039 | -0.124*** | -0.023 | -0.137*** | 0.352*** | 0.006 |
| (0.111) | (0.013) | (0.017) | (0.013) | (0.011) | (0.015) | |
| Catalogue Image | 0.212*** | 0.122*** | 0.365*** | -0.043** | 0.352*** | 0.234*** |
| (0.076) | (0.020) | (0.026) | (0.022) | (0.019) | (0.020) | |
| Professional Model | 0.230 | 0.050*** | 0.079*** | 0.002 | 0.585*** | 0.185*** |
| (0.142) | (0.018) | (0.020) | (0.013) | (0.012) | (0.019) | |
| Number of Images | 0.049*** | 0.070*** | 0.042*** | 0.052*** | 0.039*** | 0.055*** |
| (0.006) | (0.005) | (0.007) | (0.005) | (0.004) | (0.005) |
Controls: ✓
Years: 2023
Fixed Effects: Product Category
Detailed Results
The analysis reveals key insights into how different image types influence sales outcomes across various categories of second-hand clothing. The regression results, summarized in the table above, provide a nuanced understanding of these dynamics:
Impact of Real Person Images:
- Listings featuring images of real people wearing the items yielded mixed results. While these images positively influenced sales for dresses (0.352***), they had a significant negative effect on categories like sweaters (-0.124***) and pants (-0.137***). This suggests that the perceived authenticity of real-person images resonates more with certain product types, such as dresses, where fit and presentation may be more visually impactful.
Catalogue Images:
- Catalogue-style images consistently delivered positive outcomes across most categories. For instance, the use of catalogue images significantly boosted sales of tops (0.365***), jackets (0.234***), and dresses (0.352***). These results indicate that professional, standardized imagery aligns well with consumer expectations for quality and reliability in these categories.
Professional Model Images:
- Professional model images were particularly effective for dresses (0.585***) and jackets (0.185***), highlighting their value in presenting clothing in an aspirational context. However, their impact was less pronounced or neutral for other categories, such as pants and sweaters, where fit and texture may be less reliant on high-end presentation.
Number of Images:
- The total number of images in a listing had a universally positive impact across all categories. For example, each additional image increased the likelihood of a sale, with coefficients such as 0.070*** for sweaters and 0.052*** for pants. This finding underscores the importance of comprehensive visual representation in online marketplaces.
Category-Specific Trends:
- The effectiveness of image types varies significantly between categories. Dresses and jackets appear to benefit the most from professional and catalogue images, while pants and sweaters see limited or negative effects from real-person photos.