This project is for getting in-depth analysis on how certian KPI'S (key performance indictators) contribute to the growth of an online sales market. These KPI's include:
- Purchase time
- Average daily transactions
- Gender type
- Geographical purchase traffic
- Status of transactions
- Total revenue
This project for the public with much recommendations for people aiming to get into the field of data analysis, and also business managers that want to see how an interactive dashboard can give insights into the activities of their businesses.
The data was sourced from kaggle, an open source platform for getting open source public datasets. The link to the dataset is: https://www.kaggle.com/datasets/roopeshbharatwajkr/ecommerce-dataset
The dataset was gotten in spreadsheet(xlsx) format and imported into PowerBI.
Before the dataset was loaded into PowerBI, it was first transformed.
- The columns that were not needed for the analysis were removed [Transaction Id, Customer_ Id, Year_Month, Time].
- The columns were converted to the right datatypes especially the date column which was converted to the date datatype.
- The value of Transaction_result was changed as 0 = Transaction failed and 1 = Transaction Successful.
- The value of Transaction Start was changed to Transaction start.
- The City['Los Angles'] was changed to City['Los Angeles'] for effective representation of the city on a map chart.
- A date table was created using PowerBI date query code. You can find the code here: PowerBI Date query code
- A new table was created by duplicating the main table and the Transaction result was filtered to have just Transaction Successful.
- The transformed data was then apllied to the visualization panel.
New relationships were created between the tables, and date was used as the relationship between the three (3) tables.
- A measure table was created and total revenue, number of Successful transactions, number products sold, average daily transactions were all calculated using DAX.
- Charts were used to check for:
- Expected Revenue by Transaction result
- Revenue by City
- Revenue by Device Type
- Revenue by Date
- Revenue by Category
- Revenue by Gender
- Date and Customer Login type filters were created as well.
Revenue trended down between Tuesday, December 17, 2013 and Monday, January 13, 2014 with a drop of $875,719.
- Analytical Explanations:
- Male customers contributed immensely to the downward trend with a significant -64.32 % Revenue change (from $653,712 to $329, 361) as compared to Female customers whose revenue contribution fell to -46.49 % ($606, 216 to $324, 360).
- The analysis also shows that the downward trend in revenue significantly came from the members, a category that accounted for a major revenue block of $1,472,887 out of the total $1,529,440 in revenue. This analysis shows that revenue from members fell to $635,409 (-58.86 %) compared to guests which contributed an initial $56,443 in revenue, and then dropped to $18, 312 (-67.62 %).
- Fashion, and Clothing also accounted for a major decrease in revenue by category of products, with Fashion taking a pro-gravity jump to -60.39 % with $376.767 loss and Clothing following suit with -51.85 % and $283, 628 loss in revenue.
.png)
- The revenue gotten customers in Los Angeles which contributed a large chunk in revenue saw a downward trend 0f -59.42 % of total revenue change ($1,353,564 to $549,323) compared to New york with revenue change of $175,876 to $104,394.
- The longest and consistent upward trend occured between Saturday, October 12, 2013 and Saturday, October 26, 2013 with a rise of $405,462.
- This upward trend saw a 65.82 % in revenue change from female customers with fashion products having a 37.29 % increase which amounts to $952,027 from $692,708, and wearables also having an amazing 106.81 % increase in revenue ($106,396 to $220,038).
- A careful look at this analysis shows that within this time fame of upward trends in revenue, the revenue gotten from the male customers fell from $644, 043 to $563,094 indicating a -12.57% revenue growth.
- Also, the revenue gotten from customers in Seattle, Washington, fell by -7% i.e $1,172,619 to %1,090,541 compared to Los Angeles and New York which had revenue growth of 221.28 % and 368.29 % respectively.
.png)
- The web-based platform was the most used medium of purchase by customers contributing 81.09% of total revenue.
+
- Fashion, and Clothing products were the major source of revenue accumulating a total of $197,319,259 in revenue.
+
- Seattle was the most revenue-generating city raking in $140,456,574.
The dashboard was published from the PowerBI environment and a link was generated for others to view and interact with the dashboard. Link to my dashboard: E-Commerce Dashboard
If you have any feedback, please reach out to me at: akdaniel0009@gmail.com
I'm a Data Scientist with experience in building end -to- end data analytics projects and machine learning models.
Introduction to Data Science : cognitveclass.ai
Data Science Tools : cognitveclass.ai
Top 50% AI and ML with Python Regression Hackathon : datasciencenigeria.org (ID for Authorization: DSNAI0009981)
Python (OOP), Data Analytics, Machine Learning, Features engineering.
Here are some related projects