Comparing cultural trends in skincare marketing with data analysis.
Hype vs. Hyperpigmentation
Python
Project Overview
Timeline | 2 Months (2025) |
Role | Wrote initial research question, formed hypothesis, collected data, programmed interactive graphs, and trained and tested random forest model. |
Tools Used | Pandas, Numpy, NTLK, Scikit-Learn, Plotly, iPyWidgets |
Link |
Background
This project uses various data science methods to uncover the differences in marketing strategies and consumer perceptions of hyperpigmentation treatments in Asian and Western skincare markets. Using exploratory data analysis (EDA), machine learning classification, and hypothesis testing, my group analyzed ingredient emphasis, pricing structures, and linguistic framing in product descriptions from both regions.
Our findings revealed that Asian skincare brands prioritize gentle, long-term skincare benefits, often highlighting ingredients like Niacinamide and Tranexamic Acid, whereas Western brands focus on clinical efficacy and faster results, emphasizing Vitamin C and Alpha Arbutin. Price analysis showed that Western skincare products tend to be more expensive, likely due to branding and positioning, while Asian brands offer more accessible pricing.
To reinforce our findings, we trained a Random Forest model that achieved 95.31% accuracy in distinguishing between Asian and Western skincare products, showing the significant difference in marketing language, pricing, and ingredients.
Additionally, hypothesis testing confirmed a statistically significant difference (p < 0.05) in the words used to market products, with Western brands focusing on “dark spots” and “visible results” while Asian brands emphasized “moisture” and “radiance.”