Skip to content

3A Superstore Analysis project banner

3A Superstore Analytics Project

3A Superstore Analytics is a team-based data analytics project on retail transaction data. The project uses BigQuery, dbt, Python notebooks, Power BI to analyze sales, inflation-adjusted revenue performance, customer behavior, category trends, regional concentration, and retention opportunities.

A note on the project

Because the dataset covers Turkish retail sales during a high-inflation period, nominal sales can be misleading. The project combines transaction data with TCMB EVDS CPI data to compare nominal and real revenue, then checks the result against order, customer, unit, and product-price signals.

Analysis Portfolio

Each analysis page focuses on a different business question and is owned by a team member.

  • Revenue Performance & Inflation Analysis

    Nominal vs. real revenue, CPI adjustment, product price validation, and inflation-aware KPIs.

    Author: Doruk Alkan

  • Sales & Revenue Insights

    Sales trends, order volume, geographic revenue patterns, and revenue forecasting.

    Author: Ebubekir Tilbaç

  • Customer Growth Opportunities

    Cross-sell opportunities, churn signals, basket diversity, and high-value customers.

    Author: Ebubekir Tilbaç

  • Customer Health

    Customer value segmentation, health stages, geographic value concentration, and retention priorities.

    Author: Yasemen Dündar

  • Customer Retention & RFM Analysis

    RFM segmentation, active customer rate, revenue at risk, and retention strategy.

    Author: Yasemen Dündar

  • Region & Category Performance

    Regional revenue concentration and category contribution by geography.

    Author: Eda Bilgin

  • Category Trends

    Category revenue, sales quantity, order activity, and category mix stability over time.

    Author: Eda Bilgin

Technical Implementation

The dbt project is organized into a layered model structure:

  • Staging models clean raw BigQuery tables for orders, order details, customers, branches, categories, and CPI.
  • Intermediate models create reusable analytical logic such as order revenue, branch dimensions, monthly CPI metrics, and item-month pricing.
  • Mart models produce dashboard-ready tables for revenue trends, KPI cards, product price trends, category price movement, branch revenue, customer 360, and RFM analysis.
  • Custom dbt tests validate grain, reconciliation, CPI math, endpoint windows, and dashboard KPI calculations.

The current dbt graph includes 26 models, 1 seed, and 189 tests after parsing.

flowchart LR
  raw[Raw BigQuery tables] --> stg[dbt staging]
  stg --> int[Intermediate models]
  cpi[TCMB EVDS CPI seed] --> int
  int --> marts[Analytics marts]
  marts --> powerbi[Power BI dashboards]
  marts --> notebooks[Modeling notebooks]

Tools Used

Tool Role
BigQuery Warehouse, SQL exploration, raw table storage, and analytics outputs.
dbt Transformation modeling, documentation, and automated data tests.
Python, Jupyter, Google Colab Exploratory analysis, validation, and forecasting experiments.
Power BI Final dashboards and business-facing visual analysis.
Zensical Public project website and written portfolio documentation.

  • Dataset

    Source dataset, raw table overview, Kaggle citation, and CPI supplement notes.

  • Team

    Team members, ownership, and project focus areas.

  • GitHub Repository

    Source code, dbt models, notebooks, query archive, and Zensical site files.