Fictional Sales dataset with unusual transactions
Concept & Approach
The provided dataset consists of 250 orders, each with a status (pending, cancelled, completed), customer name, customer location, and order details (category, product, unit price, quantity, and total revenue). At first glance, this may seem promising, but a closer look reveals that the “customers” appear to be greedy crew members of the Starship Enterprise (Star Trek). For example: someone ordering five washing machines from Boston, cancelling the order, and then almost simultaneously purchasing two smartwatches from Denver.
The initial idea was to analyze the customer journey (what was ordered, when, and with what status), but this was not pursued.
The heatmaps were created as an alternative way to explore the data and to try something not yet attempted by others. Submissions for FigureFriday come in throughout the week; I decided to plot payment method against sales volume and order count.
Results
The creative element in the results is limited, but experimenting with the app does spark ideas about what might be worth investigating if the dataset were larger and more realistic—while still producing a similar type of output. It is not necessarily a drawback to build a prototype app that is mainly intended for exploratory data analysis.
Some questions that emerged during exploration:
- Are payments made with gift cards easier to reverse than other types of payments? Should we be concerned that gift card orders are cancelled so frequently?
- Why are so many orders (or such a large share of total order value) still “pending”?
Finally, I switched the color scheme from grayscale to something brighter and more cheerful—something many users tend to appreciate.
Example
- Two heatmaps &
- a select filter for order status.
Figure Friday is an initiative by the Dash/Plotly community, where each Friday a new dataset is shared and participants create a visualization or small app to extract insights from it. The following Friday at 18:00, there’s a Zoom session where some participants explain the thought process behind their work. In the community thread, people also share their code — and when possible, a live demo — to learn from each other.
Don’t make me think 🙂
It doesn’t get much worse than this—or perhaps it’s not so bad after all. Whether this visualization is “good” or “bad” really depends on its purpose and the audience.
The app below was designed to explore whether there were any interesting trends worth visualizing. The heatmaps break down orders by payment method and by day of the week.
