Skip to content

zidene306/first_project

 
 

Repository files navigation

Project name: Amazon Marketplace Decoding

Project Overview

This project analyzes an Amazon product dataset (source: Kaggle) to understand how pricing, discounting, and market saturation influence customer satisfaction. The goal is to validate or refute three business‑driven hypotheses using exploratory data analysis, visualization, and statistical reasoning.

Business Hypotheses

Hypothesis 1 — Premium Pricing Signals Premium Quality Higher product prices naturally signal higher quality, leading to better user sentiment.

This analysis explores:

Hypothesis 1 —

  • How reviewer sentiment varies across price segments

  • Whether premium products consistently receive higher rating

  • How price influences perceived value and user experience?

Hypothesis 2 —

  • Discounts Act as Emotional Anchors Large discounts create a psychological “deal effect,” increasing customer satisfaction and ratings.

This analysis investigates:

  • Whether higher discount percentages correlate with higher ratings

  • Whether discount‑driven purchases show inflated sentiment

  • How discount depth affects rating distribution

Hypothesis 3 —

  • Underserved Niches Show High Review Volume but Low Satisfaction
  • Categories with few products but many reviews and low ratings indicate unmet customer needs.

This analysis identifies:

Categories with low product availability but high engagement

Segments where customers express dissatisfaction despite high demand

Potential opportunities for new product development

Repository Structure

├── README.md <- Project overview and business roadmap ├── data/ │ └── amazon.csv <- Raw source dataset from Kaggle ├── notebooks/ └── eda_claire.ipynb <- Exploratory data analysis & custom plotting pipeline └── explore_clean_data_zidene.ipynb <- Exploratory data analysis & custom plotting pipeline └── functions.py <- all functions used in the project ├── slides/ └── project_slides url <- Project presentation slides

└── figures/ ├── fig1_price_vs_rating.png <- Price segmentation bar charts ├── fig2_discount_trap.png <- Dual-axis discount volume/sentiment cliff └── fig3_niche_quality_price.png <- Opportunity scoring & niche candidate analysis ├── fig4_categories_rating <- User ratings by prodcut category └── fig5_rating_distribution_by_price.png <- Evaluating rating with regards to product pricing

project_slides url: https://docs.google.com/presentation/d/1or2I0dxIeo96H9ww6bakcm8ZBFywKj_Gzh_ZptvTnnY/edit?slide=id.g3e4aa31e4e8_0_1#slide=id.g3e4aa31e4e8_0_1

Tools & Libraries

pandas,matplotlib.pyplot, seaborn, numpy

Jupyter Notebook

Kaggle dataset (Amazon product metadata + ratings)

Git for version control

Methodology

  • Data ingestion from Kaggle

  • Cleaning & preprocessing

  • Handling missing values

  • Creating price and rating segments

  • Normalizing discount rates

  • Exploratory Data Analysis

  • Sentiment distribution

  • Price vs rating relationships

  • Discount vs rating correlations

  • Hypothesis testing

  • Visual evidence

  • Statistical indicators

Insights & recommendations

Key Findings (Summary)

Hypothesis 1 —

Premium products show higher proportions of excellent ratings and fewer poor reviews, indicating that price positively influences perceived quality.

Hypothesis 2 —

Large discounts correlate with slightly higher ratings, but the effect is not uniform across categories. Emotional anchoring exists but is not the dominant driver.

Hypothesis 3 —

Several categories show low product counts, high review volume, and low satisfaction, signaling underserved niches with strong market potential.

How to Run the Project

Place raw data in data/raw/

Run cleaning scripts or notebooks to generate data/cleaned/

Open notebooks in /notebooks to reproduce analysis

Import helper functions from src/functions.py

Example:

python from src.functions import segment_price, plot_rating_distribution

Data Source

Dataset obtained from Kaggle.com
https://www.kaggle.com/datasets/talalhakem/amazon

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Jupyter Notebook 98.8%
  • Python 1.2%