Objective
A report on retail shop such as Walmart recruiting is used and analyzed for the prediction of sales. The link shared below depicts the dataset’s source:
https://www.kaggle.com/c/walmart-recruiting-store-sales-forecasting/
The objective of this report is to conduct data visualization to provide a broad understanding of their business and present the data insights to help them take right business decisions. The retailers are the targeted audiences. A dashboard will be developed for saving time required for scheduling the staff of Walmart, based on the region for the Christmas holidays.
The Walmart data represents the historical sales data of totally 45 retail stores, present in different locations. Every single store comprises several departments, and the sales of all the departments need to be projected separately, along with the challenges of holiday markdown events in a year, which impacts the sales. Moreover, it increases the difficulty of predicting the actual departments that are impacted and its severeness.
The main holiday markdown contains the following events, which are weighted 5 times more during the evaluation in contrast to the weeks without any holiday:
- Thanksgiving
- Super Bowl
- Christmas
- Labor Day
- csv: Itpresents anonymized information related to the 45 stores such as its size and type.
- csv: Itcontains historical training data from February 5th 2010 to November 1st of 2012.
- csv: Itcontains other linked information like number of the store, department, regional activity of the provided dates and so on.
The sales prediction of each store in the upcoming year is carried out to improve the customer experience in all the stores, and predicts employment rate in all the stores for staff allocation at the busy hours in specific region.
- Predicting weekly sales in the stores subjected to holiday weeks.
- Predicting weekly sales in the stores with features and sales data.
- Predicting yearly department-based sales
- Predicting unemployment rate in the stores during holiday weeks.
- Predicting holiday weeks’ promotional markdown events
This report aims to analyze the selected Walmart data with Business Intelligence tool to understand the business better, and enhance its business, its decisions, sales, and customer experience. Further, to decrease the business challenges during the holidays.
Let us start with the essential step i.e., data preprocessing to transform the selected dataset for Tableau. Before this, lets us know about Tableau. Tableau helps to design a dashboard-based solution, with its data visualization platform that specifically focuses on the BI (business intelligence). The created dashboard provides the benefit of viewing the business situation in one glance and it even represents the Key Performance Indicators (KPIs).
Now, to accomplish the actual step of data preprocessing and transforming the data, start importing the 3 files in Tableau. To do this use the union function present in Tableau. The following figure depicts this step and it is now absolutely ready for data visualization.
Tableau data visualization is Walmart recruiting data’s proposed tool via dashboard. The sales data of Walmart stores is taken and analyzed. In each store, customer behaviour is considered for data visualization. The busy hours in the specific areas during the holidays will be analyzed 1.
The aim of BI includes data capturing data manipulation and data analysis. Here, dashboard provides the required BI to represent the Walmart stores’ solution that will help them to be competitive in the business market. The planned question will be used for representing the data visualizations.
The project outcomes are taken with the help of Tableau tool’s data visualization. The process of data cleaning is performed and the missing values are removed from the dataset. Now, the implementation of planned questions is represented as follows:
A bar visualization is created to predict weekly sales in the stores during the holiday weeks. The measures selected to meet this visualization are “Is Holiday” and “Weekly Sales”
Dataset
The below depicted data visualization shows stores’ weekly sales during the holiday weeks. The store number 45 is shown to comparatively have the largest sale that is, 210975533.5.
Firstly, import the dataset, next combine it with the stores data, features data, and sales data. Later, got the worksheet and create visualization. A bar visualization is created to predict weekly sales in the stores, with the help of sales data and features. The measures selected to meet this visualization are “Store” and “Weekly Sales.” The below depicted data visualization shows stores’ weekly sales during the holiday weeks. The store number 20 is shown to comparatively have the largest sale.
A visualization is created to predict yearly sales of the departments, with the help of sales data and features. The measures selected to meet this visualization are “Date”, “Department number”, and “Weekly Sales”
The below depicted data visualization shows department based annual sales. The department number 92 is shown to comparatively have the largest sale that is, 140380890.9.
A visualization is created to predict unemployment rate in the stores during the holiday weeks, with the help of sales data and features. The main purpose here is to allocate staff for the management of busy shops during the holidays. The measures selected to meet this visualization are “Unemployment”, “Is Holiday”, and “Store”
The below depicted data visualization shows rate of unemployment during the holiday weeks. The store number 12 is shown to comparatively have the largest unemployment rate that is, 69.002. Thus, on the sales it reflects greater amount of negative impact. Therefore, Walmart must have sufficient number of staffs, especially during holiday weeks to increase the sales.
In the end, a visualization is created to predict the promotional markdown events during the holiday weeks, with the help of sales data and features. The measures selected to meet this visualization are “Markdown 1”, “Markdown 2”, “Markdown 3”, “Markdown 4”, “Markdown 5”, and “Holiday Weeks.”
The below depicted data visualization shows promotional markdowns during the holiday weeks. The Markdown 5 is shown to comparatively have the largest promotional value to increase the sales of the stores.
Here, the same is represented with the values in a table form.
- Weekly sales in the stores subjected to holiday weeks:The store number 45 is shown to have the largest sale that is, 5.
- Weekly sales in the stores: The store number 20 is shown to comparatively have the largest sale.
- Yearly department-based sales:The department number 92 is shown to have the largest sale that is, 9.
- Unemployment rate in the stores during holiday weeks:The store number 12 is shown to have the largest unemployment rate that is, 69.002. Thus, Walmart must have sufficient number of staffs, if not it will have negative impact on its sales.
- Promotional markdown events during holiday weeks: During the holiday week, Markdown 5 is shown to have the largest promotional value to increase the sales of the stores.
In the world of big data, it is essential to use data visualization for data analysis, as is an effective way for identifying the insights and helps to overcome the challenges of extremely complexity and dimensionality. But it won’t be implemented for all the analysis. If a large number of dimensions are visualized accurately, then there is a higher possibility of identifying the possible interesting correlations, outliers, or patterns.
In big data visualization, the other challenges include perceptual and interactive scalability. Every single data point visualization can result in over-plotting and might crush the perceptual and cognitive capacities of the users, which decreases the data via filtering or sampling. Querying the large data stores could lead to high latency, and it disrupts the fluent interaction 2.
6. Conclusion and Future Work
The benefits of BI is understood through this data analysis, followed by the significant roles of data cleaning and data preprocessing. The results conclude that Walmart should hire enough number of staffs, mainly to handle the busy hours during the main holidays, if not this can cause negative impact on its sales. The staff must be scheduled based on the busy stores and the regions to effectively manage the staff’s working hours during the holiday weeks.
Walmart is suggested to have enough staff and accurate data, and the strength of any retail business. The challenges of staff managing in the stores can be resolved with the insights represented by the dashboard. Thus, solutions based on the dashboard representation suggests to create a data handling system, according to the insights take right business decision that ensures increase in the sales, and take inputs from the customers to determine the customer experience in the stores, and this will be very helpful in improving the sales.
8. References
(1) What Is Data Visualization? Definition, Examples, And Learning Resources https://www.tableau.com/learn/articles/data-visualization (accessed Mar 15, 2022).
(2) M.; Samihardjo, R.; Nugraha, U. Design Of The Business Intelligence Dashboard For Sales Decision Making. International Journal of Psychosocial Rehabilitation 2020, 24, 3498-3513.