Instacart EDA with SQL

Description

This case study is part of the Data Lemur SQL course and focuses on exploring a complex relational database through SQL.
The analysis combines SQL queries with spreadsheet-based inspection to uncover product trends, user behavior, and pseudo-temporal dynamics within the Instacart dataset.
Tools used:

• PostgreSQL – full database hosting
• Visual Studio Code + SQLTools – query execution & CSV extraction
• Excel Power Query – data shaping & visualization

Important note on the data The original dataset contains no real timestamps, so the case study uses pseudo-timelines:

• order_products_prior as Q2
• order_products_train + order_products_test as Q3

Objective

Perform an Exploratory Data Analysis (EDA) using SQL first, and Excel/Power Query for second-layer insights and data visualization.
The goal is not business forecasting but hands-on analytical practice with a large dataset.

Disclaimers:

• The dataset is outdated (2017), partially incomplete and especially sparse in pesudo-Q3.
  The results should not be used for actual strategic decision.
• The complete dataset is too large for this repository.
  You can download it directly from Kaggle.

Repository content

Work in progress - feel free to browse folders meanwhile

• /img - images used in report, this will contain graphs too
• /output - future CSV files produced from Excel Power Query and XLSX file
• /queries - SQL queries in the analysis
• /report - ongoing report writing

How to view report progress

1. Clone the repository
2. Open the /report/instacart-case-study.docx file