Data Prep
-
The first step of preparing the data consisted of cleaning it by deleting or replacing blank fields, correcting data types or formats for the different columns and unifying the sources of data. Once that was complete, I imported the data into Power Bi and started exploring options on how to create the best connections as well as how to create a Date Dimension table that could help me easily navigate through the data, using a single data source. I ended up creating a connection within the four tables by using the variables Date and Store ID. Here I include a picture of my sketch and the actual created connections on PowerBI.
Design
-
The next step was to design and sketch the different graphs I wanted to use. The easiest way to get to the answer was by asking myself research questions to help brainstorm different visuals. The questions included:
-
What states have the most sales? Details for each state’s sales
-
What are the previous year and current sales by quarter?
-
What is the margin growth, sales, rent and wages?
-
What is the kitchen department sales composed of? Other subcategories?
-
What are the year over year sales ($) by department? Has it grown or shrunk in the past year?
-
What are the sales composed of in terms of departments and subcategories of products?
-
What is the actual sale dollar amount in comparison to the target dollar amount?
-
Below is an image of the final sketch I utilized to pick my graphs for the project:
Visualizations
-
I was really excited to start prototyping in Power BI and I'm so glad that I got to experiment with different graphs, colors, sizes, etc. My favorite charts in this dashboard were the waterfall graph, the stacked area graph and the donut chart. These are three graphs that I don't normally go for yet and they were surprisingly effective at showing the running total as values are added or subtracted and the evolution of the value of several groups on the same graphic.
-
Below is an image of the four page dashboard I created:
Reflection
- A myth I had always heard while working in analytics is that pie, donut, and map graphs shouldn’t be used because they can be less effective to understand the data. However, in this project I came to find that when used in the right context and time, they can actually be very useful and easy to follow. In this project I wanted to explore how we could use these types of graphs in a financial dashboard.
- The main reason why they are not as popular is because they require more effort from the user to interpret them which makes it harder for people to understand what they are looking at. However, after doing some research on how other companies were using them and how these graphs can be useful for financial dashboards I realized that there are infinite options to easily showcase information
Store
Insights Dashboard
OBJECTIVE
This project was a personal project I completed to learn more the use of different visualizations in financial dashboards. The data is based on store sales from around the country and it shows details on the different types of products they sell.
Tools
-
Microsoft Excel
-
Microsoft Power BI
Data Source
-
The data for this dashboard was obtained from the Corporate Finance Institute. It consists of three different data documents: Store details, Sales Facts and Costs and Targets.
