Data Analytics Case Studies

Adidas Sales Performance Dashboard (2020–2021)

Project Overview
This case study explores over 1 million rows of Adidas sales data transformed into an interactive dashboard designed to uncover trends in profitability, product performance, sales channels, and geographic distribution. The goal was to translate complex, high-volume data into clear, actionable insights for business and marketing stakeholders.

    • Identify top-performing product categories across men’s and women’s lines

    • Analyze quarterly profit trends year-over-year (2020–2021)

    • Compare performance across sales methods (online, outlet, in-store)

    • Surface regional and state-level sales patterns to inform market strategy

    • Present insights in an intuitive, executive-ready visual format

    • Data Preparation & Cleaning

      • Consolidated raw sales data exceeding 1M records

      • Cleaned inconsistencies, standardized fields, and validated metrics in Excel

      • Structured data for efficient querying and visualization

    • Exploratory Data Analysis (EDA)

      • Analyzed revenue, profit, and units sold by product category and gender

      • Evaluated seasonal and quarterly performance trends

      • Assessed geographic concentration of sales by region and state

    • Dashboard Design & Visualization

      • Designed an interactive Tableau dashboard with multiple analytical views

      • Used bar charts, maps, and bubble charts to support quick insight discovery

      • Prioritized clarity, hierarchy, and usability for non-technical stakeholders

    • Identified Men’s Street Footwear as the top-selling product category

    • Revealed strong profit growth in 2021, with Q3 as the peak quarter

    • Confirmed online sales as the leading channel by volume

    • Highlighted the West region as the highest-grossing market

    • Enabled data-driven recommendations for inventory, marketing, and regional investment strategies

    • Tableau (Data Visualization)

    • Python

    • Microsoft Excel

Teen Smartphone Behavior & Addiction: Exploratory Data Analysis

Project Overview
This exploratory data analysis examines smartphone usage patterns among 3,000 teenagers (grades 7–12) to better understand how digital behaviors intersect with sleep, mental health, and addiction levels. Drawing from my background in UX/UI design, this project was motivated by growing concerns around intentional product design, persuasive technology, and their psychological impact on youth—concepts popularized in The Social Dilemma.

The analysis aims to move beyond assumptions about screen time and uncover whether commonly cited factors like social media usage actually correlate with anxiety, depression, sleep loss, or addiction.

    • Investigate relationships between screen time before bed and sleep duration

    • Explore whether social media usage correlates with anxiety and depression

    • Identify potential gender differences in smartphone addiction levels

    • Examine whether phone usage purpose is associated with addiction severity

    • Challenge common narratives using evidence-based analysis

  • 1.Data Understanding & Codebook Development

    • Analyzed a dataset containing 25 behavioral, psychological, academic, and demographic variables

    • Classified each variable by conceptual data type (behavioral, psychological, ordinal, continuous, etc.)

    • Established a clear analytical framework to guide transformations and comparisons

    2. Data Cleaning & Transformation

    • Converted numeric strings to appropriate quantitative formats

    • Standardized and categorized ordinal mental health measures into Low / Medium / High

    • Cleaned inconsistent gender labels and removed ambiguous entries to ensure analytical clarity

    • Reduced noise by rounding and normalizing behavioral metrics (e.g., screen time, sleep hours)

    • Resulted in a cleaned, analysis-ready dataset of ~2,000 participants

    3. Feature Engineering & Subsetting

    • Created focused variable subsets aligned to each research question

    • Split and recombined demographic and behavioral data to support flexible analysis

    • Saved cleaned data to optimized formats for reproducibility and scalability

    4. Exploratory Analysis & Visualization

    • Used histograms and bar charts to examine:

      • Sleep duration distributions

      • Screen time before bed

      • Social media usage patterns

      • Addiction level distributions

      • Mental health self-reports

    • Performed grouped aggregations to compare average behaviors across mental health categories

    • Iteratively refined research questions based on observed trends rather than assumptions

    • Fully cleaned and documented dataset

    • Reusable Python scripts for data preparation and transformation

    • Exploratory visualizations supporting behavioral and psychological analysis

    • Clearly framed analytical conclusions grounded in evidence, not correlation myths

    This analysis reframes how smartphone addiction and mental health are discussed by:

    • Demonstrating the limits of screen time–only explanations

    • Highlighting the need for holistic, systems-level thinking

    • Offering UX professionals and educators a more nuanced foundation for ethical design and intervention strategies

    • Python (Pandas, Matplotlib, NumPy) — Data cleaning, transformation, analysis, and visualization

    • Jupyter Notebook — Reproducible EDA workflow

    • CSV & Feather formats — Optimized data storage and reuse

Previous

Netflix Japan: Anime Expo
Next

Emmys Awards 2022: Presentation GFX