Skip to content
0

Supply Chain Analytics in Tableau or Power BI

📖 Background

Test your BI skills on a real-world dataset focusing on supply chain analytics. As the main data analyst for Just In Time, you will help solve key shipment and inventory management challenges, analyze supply chain inefficiencies, and create insightful dashboards to inform business stakeholders about potential problems and propose structural business improvements.

Be creative and make use of your full skillset! Use this Workspace to prepare your data, import the tables into your local Tableau or Power BI instance, and share your insights below.

The end goal will be a (set of) interactive dashboards that demonstrate clear insights for Just In Time.

💾 The data

GroupColumn nameDatasetDefinition
CustomerCustomer IDorders_and_shipments.csvUnique customer identification
CustomerCustomer Marketorders_and_shipments.csvGeographic grouping of customer countries, with values such as Europe, LATAM, Pacific Asia, etc.
CustomerCustomer Regionorders_and_shipments.csvGeographic grouping of customer countries, with values such as Northern Europe, Western Europe, etc.
CustomerCustomer Countryorders_and_shipments.csvCustomer's country
Order infoOrder IDorders_and_shipments.csvUnique Order identification. Order groups one or multiple Order Items
Order infoOrder Item IDorders_and_shipments.csvUnique Order Item identification. Order Item always belong to just one Order
Order infoOrder Yearorders_and_shipments.csvYear of the order
Order informationOrder Monthorders_and_shipments.csvMonth of the order
Order informationOrder Dayorders_and_shipments.csvDay of the order
Order informationOrder Timeorders_and_shipments.csvTimestamp of the order in UTC
Order informationOrder Quantityorders_and_shipments.csvThe amount of items that were ordered within a given Order Item (1 record of the data)
ProductProduct Departmentorders_and_shipments.csvProduct grouping into categories such as Fitness, Golf, Pet Shop, etc.
ProductProduct Categoryorders_and_shipments.csvProduct grouping into categories such as Sporting Goods, Women's Apparel, etc.
ProductProduct Nameorders_and_shipments.csvThe name of the purchased product
SalesGross Salesorders_and_shipments.csvRevenue before discounts generated by the sales of the Order Item (1 record of the data)
SalesDiscount %orders_and_shipments.csvDiscount % applied on the catalog price
SalesProfitorders_and_shipments.csvProfit generated by the sales of the Order Item (1 record of data)
Shipment informationShipment Yearorders_and_shipments.csvYear of the shipment
Shipment informationShipment Monthorders_and_shipments.csvMonth of the shipment
Shipment informationShipment Dayorders_and_shipments.csvDay of the shipment
Shipment informationShipment Modeorders_and_shipments.csvInformation on how the shipment has been dispatched, with values as First Class, Same Day, Second Class, etc.
Shipment informationShipment Days - Scheduledorders_and_shipments.csvInformation on typical amount of days needed to dispatch the goods from the moment the order has been placed
WarehouseWarehouse Countryorders_and_shipments.csvCountry of the warehouse that has fulfilled this order, the only two values being Puerto Rico and USA
Inventory & FulfillmentWarehouse Inventoryinventory.csvThe monthly level of inventory of a product, e.g. 930 units
Inventory & FulfillmentInventory cost per unitinventory.csvThe monthly storage cost per unit of inventory, e.g. $2.07
Inventory & FulfillmentWarehouse Order fulfillment (days)fulfillment.csvThe average amount of days it takes to refill stock if inventory drops below zero

The data can be downloaded from the sidebar on the left (under Files).

💪 Challenge

Using either Tableau or Power BI, create an interactive dashboard to summarize your research. Things to consider:

  1. Use this Workspace to prepare your data (optional).
  2. Some ideas to get you started: visualize how shipments are delayed, by country, product, and over time. Analyze products by their supply versus demand ratio. Rank products by over or understock. Don't feel limited by these, you're encouraged to use your skills to consolidate as much information as possible.
  3. Create a screenshot of your (main) Tableau or Power BI dashboard, and paste that into the designated field.
  4. Summarize your findings in an executive summary.

Reading the Data

import pandas as pd
data = pd.read_csv("data/orders_and_shipments.csv")
data.head()

Data Cleaning

Determining if is there any null values and the data types.

def data_info():
    temp = pd.DataFrame(index = data.columns)
    temp["Datatype"] = data.dtypes
    temp["Non-null Values"] = data.count()
    temp["Null Values"] = data.isnull().sum()
    temp["% of Null Values"] = (data.isnull().mean())*100
    temp["Unique Count"] = data.nunique()
    return temp
data_info()

Checking if there is any duplicates

dup = data[data.duplicated()].shape[0]
if dup == 0:
    print("There are not duplicated rows")
else:  
    print(f"There are {dup} rows")

data.columns

Removing the spaces from the columns names

data.columns = data.columns.str.strip()
data.columns

Changing some data types

#Replacing some "-" for zeros and convert it to float the column Discount%
data['Discount %'] = data['Discount %'].replace('  -  ', 0)
data['Discount %'] = data['Discount %'].astype('float')

#Creating a Order Date which is going to be more useful in Tableau. 

data['Order Date'] = pd.to_datetime(
    data["Order Year"].astype(str)+
    data["Order Month"].astype(str).str.zfill(2)+
    data["Order Day"].astype(str).str.zfill(2) +
    ' ' +
    data["Order Time"], format='%Y%m%d %H:%M'
)

#Creating a shipment Date which is going to be more useful in Tableau. 

data['Shipment Date'] = pd.to_datetime(
    data["Shipment Year"].astype(str)+
    data["Shipment Month"].astype(str).str.zfill(2)+
    data["Shipment Day"].astype(str).str.zfill(2) 
)

#lets drop the columns we won't use in tableau

columnstodrop = ['Order Year', 'Order Month', 'Order Day', 'Order Time', 'Order YearMonth', 'Shipment Year', 'Shipment Month', 'Shipment Day']

data = data.drop(columnstodrop,axis=1)

#Checking the new data types of our new Data Frame
data_info()

**Checking each of the variables possible values to see if there is anything to correct on the column names. **

‌
‌
‌