E-commerce Data Analysis for Sales Strategy Using Python
Check out this informative and concise case study applying data analysis using Python to a well-defined e-commerce scenario.
By Juhi Sharma, Product Analyst
Source — https://www.wvgazettemail.com/
Kmart is a leading online retailer in the US and as part of their annual sales review meeting, they need to decide on their sales strategy for the year 2020 based on the insights from the sales data in 2019.
Data is related to sales for each month of 2019 and the task is to generate key insights which will help the sales team of Kmart to take some key business decisions towards Fine-tuning their sales strategy.
- Data Belongs to Kmart -a leading online retailer in the US.
- Time Period — January 2019 — December 2019
- Unique Product — 19
- Total Orders — 178437
- Cities — 9
- KPI’s — Total Sales, Total Products Sold
Source — By Author
Business Problem Statements
- What was the best month for sales? How much was earned that month?
- Which city had the highest number of sales?
- Recommend the most appropriate time to display advertising to maximize the likelihood of customers buying the products?
- What products sold the most? Why do you think it sold the most?
Data Analysis Using Python
- Loaded Data of each month and made data frame using pandas
- Concatenated Dataset to make one Dataset for 2019 sales.
- Treating Null Values and Junk Data.
- Made a Filtered Dataset after preprocessing data
- Analysis and answers to business problems. (visualizations using matplot and seaborn library)
1. Importing Libraries
import pandas as pd
2. Loading Dataset and making Dataframes
df1=pd.read_csv("Sales_January_2019.csv") df1["month"]="Jan" df2=pd.read_csv("Sales_February_2019.csv") df2["month"]="feb" df3=pd.read_csv("Sales_March_2019.csv") df3["month"]="mar" df4=pd.read_csv("Sales_April_2019.csv") df4["month"]="apr" df5=pd.read_csv("Sales_May_2019.csv") df5["month"]="may" df6=pd.read_csv("Sales_June_2019.csv") df6["month"]="june" df7=pd.read_csv("Sales_July_2019.csv") df7["month"]="july" df8=pd.read_csv("Sales_August_2019.csv") df8["month"]="aug" df9=pd.read_csv("Sales_September_2019.csv") df9["month"]="sep" df10=pd.read_csv("Sales_October_2019.csv") df10["month"]="oct" df11=pd.read_csv("Sales_November_2019.csv") df11["month"]="nov" df12=pd.read_csv("Sales_December_2019.csv") df12["month"]="dec"list=[df1,df2,df3,df4,df5,df6,df7,df8,df9,df10,df11,df12]
3. The shape of each month’s dataset
for i in list: print(i.shape)
Source- By Author
4. Concatenating dataset
Source- By Author
5. Columns of Final Dataset
6. Dataframe Information
7. Null values in the dataset
frame.isnull().sum() # there are 545 null values in each column except month
(frame.isnull().sum().sum())/len(frame)*100 # we have 1.75 percent null values , so we can drop them
8. Dropping Null Values
9. Removing Junk Data
we observed that there are 355 columns in which values in rows are the same as the header. so making a new data frame where these values will be excluded.
frame[frame['Quantity Ordered'] == "Quantity Ordered"]
df_filtered = frame[frame['Quantity Ordered'] != "Quantity Ordered"] df_filtered.head(15) df_filtered.shape
Solutions to Business Problems
Q 1. What was the best month for sales? How much was earned that month?
df_filtered["Quantity Ordered"]=df_filtered["Quantity Ordered"].astype("float") df_filtered["Price Each"]=df_filtered["Price Each"].astype("float")# Creating Sales Column By multiplying Quantity Ordered and Price of Each Productdf_filtered["sales"]=df_filtered["Quantity Ordered"]*df_filtered["Price Each"]
month=["dec","oct","apr","nov","may","mar","july","june","aug",'feb',"sep","jan"] df["month"]=monthfrom matplotlib import pyplot as plt a4_dims = (11.7, 8.27) fig, ax = pyplot.subplots(figsize=a4_dims) import seaborn as sns sns.barplot(x = "sales", y = "month", data = df) plt.title("Month wise Sale") plt.show()
The best Month for sales was DECEMBER.
Total sales in December is $ 4619297.
Q 2. Which city had the highest number of sales?
dftemp = df_filtered list_city =  for i in dftemp['Purchase Address']: list_city.append(i.split(",")) dftemp['City'] = list_city dftemp.head()
df_city=df_filtered.groupby(["City"])['sales'].sum().sort_values(ascending=False) df_city=df_city.to_frame() df_city
city=["San Francisco","Los Angeles","New York City","Boston","Atlanta","Dallas","Seattle","Portland","Austin"] df_city["city"]=cityfrom matplotlib import pyplot a4_dims = (11.7, 8.27) fig, ax = pyplot.subplots(figsize=a4_dims) sns.barplot(x = "sales", y = "city", data = df_city) plt.title("City wise Sales") plt.show()
San Francisco has the highest sales f around $8262204.
Q 3 What products sold the most?
source- By Author
df_p=df_filtered.groupby(['Product'])['Quantity Ordered'].sum().sort_values(ascending=False).head() df_p=df_p.to_frame() df_p
product=["AAA Batteries (4-pack)","AA Batteries (4-pack)","USB-C Charging Cable","Lightning Charging Cable","Wired Headphones"] df_p["Product"]=productfrom matplotlib import pyplot a4_dims = (11.7, 8.27) fig, ax = pyplot.subplots(figsize=a4_dims) sns.barplot(x = "Quantity Ordered", y = "Product", data = df_p) plt.title("Prouct and Quantity Ordered") plt.show()
31017.0 quantity of AAA Batteries (4-pack) is sold in a year. It is sold maximum because it is the cheapest product.
Q 4 Recommend the most appropriate time to display advertising to maximize the likelihood of customers buying the products?
dftime = df_filtered list_time =  for i in dftime['Order Date']: list_time.append(i.split(" ")) dftime['Time'] = list_time dftime.head()
df_t=df_filtered.groupby(['Time'])['sales'].sum().sort_values(ascending=False).head() df_t=df_t.to_frame() df_t
Source -By Author
Before You Go
Thanks for reading! If you want to get in touch with me, feel free to reach me at email@example.com or my LinkedIn Profile. Also, you can find the code for this article and some really useful data science projects on my GitHub account.
Bio: Juhi Sharma (Medium, GitHub) has 2+ years of work experience as an Analyst with the role of Project Management, Business Analysis, and Client Handling. Currently, Juhi is working as a Data Analyst for a Product Company. Juhi has hands-on experience in analyzing datasets, creating machine learning and deep learning models. Juhi is passionate about solving business problems with data-driven approaches.
Original. Reposted with permission.
- Pandas Profiling: One-Line Magical Code for EDA
- The question that makes your data project more valuable
- How to frame the right questions to be answered using data