Dillard's Department Stores

Behind the Data

The Dillard's Department Store Database contains retail sales information gathered from store sales transactions. The sale process begins when a customer brings items intended for purchase (clothing, jewelry, home décor, etc) to any store register. A Dillard's sales associate scans the individual items to be purchased with a barcode reader. This populates the transaction table (TRNSACT), which will later be used to generate a sales receipt listing the item, department, and cost information (related price, sale price, etc.) for the customer. When the customer provides payment for the items, payment details are recorded in the transaction table, the receipt is printed, and the transaction is complete. Other tables are used to store information about stores, products, and departments.

The Dataset

Dillard's has had an ongoing collaboration with Walton College Enterprise Systems. This collaboration has resulted in several gifted datasets, updates and finally a weekly updated gift for advancing the skillsets of both our students and faculty. The initial retail sales information, UA_DILLARDS_2016, was provided to the Walton College of Business by Dillard's stores, Inc. Consisting of five tables with more than 128 million rows already populated and ready for use. Dillard's updated this dataset in 2018 (WCOB_DILLARDS). This dataset included the same five tables and consisted of 75 million customers, 107 weekly updates to WCOB_DILLARDS in a new dataset WCOB_DILLARDS_LIVE. This latest collaboration allows faculty and students to create inferences on prior data in order to test model efficacy with the next week's update.

Contents

This is a gifted dataset that is based on real operational data. Like any real database, integrity problems may be noted. This can provide a unique opportunity not only to expose students to real data but also to illustrate the effects of data integrity problems.