Data-Warehouse-Ecosystem
A data warehouse ecosystem for retail stores
This repository includes details & conveys the journey of creating a Data Warehouse solution for sample retail stores 10 & 21 to store pertinent data and use the analytical capabilities of a data warehouse to analyze current store performance, resulting in insights and trends that can be used to make recommendations for enhancing store productivity and assisst in organizational business expansion.
This project is developed using Snowflake and Tableau with SQL being the core programming language.
The JOURNEY
The first step was profiling the data based on initial understandings followed by designing an analysis framework and segregating data into Facts & Dimensions and creating a denormalized ERD for the data warehouse which set the base for the project.
This was followed by creating the data warehouse using the Snowflake platform extensively. This involved designing and loading staging, dimension, and fact tables, ensuring correct relationship between the tables and lastly designing views for future visualizations based on the problem statement using SQL
Lastly, Tableau was used to develop visualizations using the SQL views from Snowflake and user stories were created. The visualizations were built on the folowing:
- An overall assessment of Stores number 10 and 21’s sales.
- How are they performing compared to target? Will they meet their 2014 target?
- Should either store be closed? Why or why not?
- What should be done in the next year to maximize store profits?
- Recommend 2013 bonus amounts for each store if the total bonus pool is $2,000,000 using a comparison of 2013 actual sales vs. 2013 sales targets as the basis for the recommendation.
- Assess product sales by day of the week at stores 10 and 21. What can we learn about sales trends?
- Should any new stores be opened? Include all stores in your analysis if necessary. If so, where? Why or why not?
Based on the insights and trends identified via the tableau visualizations, recommendations were made to improve performkance and make business decisions.