Since this dataset was pulled from Kaggle to much cleaning was not necessary. I trimmed few unnecessary data & added a few columns by using the existing data for analysis. I've also converted string data to boolean for the column called Reached on time. This column recorded if on time delivery was a success or not.
Data Cleaning:
Exploratory Data Analysis (EDA):
EDA was performed in two separate tables to figure out various stuff regarding shipments, effects of shipments & modes used. Both the tables are reformed by selecting sets of columns from the cleaned table depending on the analysis. The tables created for EDA are:
1) Weight range & modes of shipment: This table mostly focuses on the relationship between weight and the modes selected for shipping that weight. This was done to see if any pattern was visible on that context and because of that reason weight ranges (in kgs) was added as a new column. This also came in handy while filtering data.
- Here Cost of products are colorized in a descending fashion, the range is darker shades of blue (meaning larger values) to darker shades of red (meaning smaller values).
- Weight in grams have bar charts showing visual difference also a range column is calculated from it.
- Click on the
icon to filter data live.
2) Modes of shipment related to customer satisfaction: This table mostly focuses on the relationship between the modes selected for shipping and customer satisfaction after getting shipment. This was done to see if any pattern was visible on that context.
- Here Customer ratings are colorized in a descending fashion, the range is red < orange < yellow < light green < green.
- Click on the icon to filter data live.
3) Customer satisfaction to repeating customers ratio: This table mostly focuses on the ratio between the number of returning customers and satisfaction level of that customer. This is also related to mode of transport which effects the customer satisfaction that determines the return rate.
- Here Customer ratings are colorized in a descending fashion, the range is red < yellow < darker shade of yellow < light green < green.
- Click on the icon to filter data live.
Modes:
There are 3 types of modes transporting the loads. A visual representation is shown based on the popularity of each individual mode.
It is clearly shown that mostly ships are used while transporting. Even the combined number of shipments by flights and roads are less than half while comparing with ship transports.
- Because of this large number, the filtered (filtered customer rating - ascending to descending) version of second EDA table shows that even if ships are mostly used they generate the larger portion of low ratings.
- For a better understanding visualization are represented as dashboards.
Dashboard:
1) This interactive Visualization/Dashboard explains the relationship between product weight & shipment modes.
Filtering is possible based on level of importance received from customers.
- High importance shipment;
- Medium importance shipment;
- Low importance shipment.
- The bar chart is divide on the weight class.There are a total of 7 weight classes, also color quotation are specified for mode of transport.
- Filtering them will provide interesting results.
- The percentage of distribution will vary according to selected mode.
- Filtering them will provide interesting results.
- The bar chart is divide on the ratings received. Color notations are provided for better understanding.
- The smaller green line chart is the same thing but shows the percentage rather then the whole number.
- The percentage of distribution will vary according to selected mode.
- Color notations are provided for better understanding. they represent the number of times a customer has visited.
Further filtering is possible by choosing a particular shipment method.
a) The bar chart is showing the weight to mode of transport ration in percentage. The table below shows the number of products shipped via different modes of transport & is divide by the weight range of the products. This gives a more descriptive analysis.
b) The pie chart is a distribution range that shows the ratio of each weight class.
2) This interactive Visualization/Dashboard Talks about various things such as Customer rating analysis which is based on shipment methods. Returning customer analysis & on time delivery ratio.
Filtering is possible based on Customer gender, amount of calls received from an individual customer regarding an order & modes of transportation.
a) Both the green bar charts are showing the ratio of product sales between an individual mode of transport and the number of ratings received.
b) The yellow bar chart is a analysis for returning customers. It is based on the ratings provide by the customer after every visit.
c) The other line chart represents on time delivery ratio. This is related to the customer satisfaction, returning customer analysis & the transportation mode selected. Filtering each chart will provide interesting results.