View project

Our project placed 1st out of 34 teams for Best Data Analysis in the 2023 UW Dubstech Datathon. The project was created by a team of three over the course of two days.

Problem

As our team members were just dipping our toes into data analysis, we chose the grocery store data track to work with. Prior to the datathon, we had attended workshops to learn how to use Power BI and Tableau. The event organizers provided us with a dataset and a list of potential questions to explore:

• How does the year 2017 compare to the year 2018? Were there any significant changes in purchasing habits of the customers?
• Which items should the store increase/decrease the price of? By how much? Why?
• Which items should the store stop selling? Why?
• Which hour of the day should the store offer discounts? Why?
• Which items should potentially be sold together or kept closer together in the store to increase sales (Basket Analysis)?
• What trends do we notice in the basket size (total items in one receipt)?
• What trends do you notice for the store with respect to transactions?
• What trends do you notice for the store with respect to time?
• What trends do you notice with respect to the categories & items listed?
• What trends do you notice in item categories & subcategories?
• Are there any categories or items the store should immediately focus on?
• What are the most efficient ways that the store can reduce losses & increase profits?
• Give a sales analysis report answering fundamental questions that the store owner wants to know:
  • The highest-selling products by month and category
  • The least selling products by month and category
  • The most profitable month by sales
  • What are the most efficient ways that the store can reduce losses
  • Any other question you can think of!

Solution

We started answering these questions by examining and playing with the data we had before us. One of my teammates is exceptional at logical data analysis, so we learned a lot just by hearing him think out loud. As a designer, I felt most comfortable providing input on how to best present the data to maximally benefit our target stakeholders, the business managers. My team complimented each other well with our data analysis, data visualization, design, and communication skills. You can view the interactive results of our analysis through the website I designed and coded for our presentation. Our team collaborated on the report and conclusions, and the data visualizations were created using Tableau.

Attributions

Project created with Iris Hamilton and Sean Lim for the UW Dubstech 2023 Datathon.