1 of 21

Seoul Team 1

Group C

​

FSDA May 2022

​

Analysis on Pakistan’s Largest E-Commerce

Final Project:

2 of 21

Meet The Teams

Debby Syefira

Visualization & Presentation Team

Hario Ardi Nugroho

Project Lead

Allessandro Yudo A

Data Cleaning & Analysis Team

3 of 21

Table of Contents

01

02

03

04

Dataset Overview

Scope and goals, methodology, and executive summary

Business Problem

Business statement

Data Exploration and Findings

Dataset Analysis and results

Summary Highlight

highlight and recommendation

4 of 21

1. Dataset Overview

Total GMV

(Jul 2016 - Aug 2018)

>Rs 1.6B

Total Number of customer as of Aug 2018

69.6k

Total product solds

​

386,350

Product categories,

15

Including mobiles & tablets,

fashion, appliances, etc

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

5 of 21

Scope and Goals

Scope of Problems

The sale increased annually but the loss of sales was still high

Increasing Sales reported throughout every November. However, the loss of sales was still high

Achieve higher growth in upcoming promotion by increase cust loyalty

Decreasing in loss sales caused by cancelled order in upcoming year

Goals

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

A

B

6 of 21

Methodology

Planning

​

​

  • Data cleaning using Excel and Phyton
  • Visualization using Tableau

Conducting

​

Analyzing

Synthetizing

​

​

  • Collect all findings and results
  • Summarizing the result
  • Choosing Dataset
  • Defining Problems
  • Defining Purpose
  • Defining Hypothesis
  • Analyzing data sales & transaction
  • Interpreting the result

​

​

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

7 of 21

Executive Summary

Top 5 Contributors

Top Payment Method

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

8 of 21

Total Sales and Quantity Grouped by Product Category

  • Around 82% of revenue was generated from mobile & tablets, appliances, entertainment, women's fashion, men's fashion, and superstore.

​

  • Mobile & tablets contributes 37% revenue to company also it’s the second highest in quantity with 16% of orders which is high value product

To stimulate more growth, pakistan e-commerce have to maintaining their top 3 most active categories, specially mobile & tablets which is give ⅓ revenue for company

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

9 of 21

The Total Transaction

VS Transaction Amount

Payment using COD method are decreasing over the year, and customers tend to switch to digital payments such as e-wallets and credit cards.

Even though COD is most popular payment method, E-Wallet has highest transaction amount.

​

Why cod is the most popular payment? many ecommerce trying to tap to rural area with COD since buyers are not suited to mbanking and other modern payment.

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

10 of 21

2. Business Statement

How to Increase sales by identifying which causes canceled orders from 2016 - 2018 and evaluate the effectiveness of the white Friday campaign to customers' loyalty in Pakistan's Largest E-commerce?

?

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

11 of 21

For the last three years, the average revenue received increased annually with highest revenue and total transaction happened in November 2017

3. Data Exploration and Finding

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

12 of 21

White / Blessed Friday campaign every November each year successfully acquired a high number of transactions and also have succeeded in increasing sales

White Friday Campaign Gives Us 8x Increase In Revenue

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

13 of 21

Sadly, White Friday campaign doesn’t attract loyal customers

Customers churn rate from White / Blessed Friday is very high (>90% churn rate in first month after customer registration)

Mostly, customers from 2016 were always repurchased next month (or year), they can be flagged as a loyal customer

​

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

14 of 21

Around 3.3B Potential GMV could acquire from canceled order

Their completion rate never achieve more than 60% which mean cancelation rate is always higher than 40% in every quarter, with 24.64% of the canceled reason from the total canceled order being a refund.

Cancel means that the customer has not processed the payment yet so that there is no delivery process, while the refund means that the customer has made the payment process and the product may have been delivered.

But MORE sadly, cancelation rate is always above 40%

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

15 of 21

E-wallet is the most frequently canceled payment method, which is unfortunate because E-Wallet has the highest amount of transactions

E-wallet payment is the most significant source of canceled orders . . .

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

16 of 21

Specifically, what are the most canceled E-Wallet Payments?

Total transaction with E-Wallet Payments

Total transaction cancelled on E-Wallet Payments

Payaxis and Easypay are the most frequently used E-Wallet but also most canceled payments method

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

17 of 21

For 3 years

Refunded Items

Top 4 items that most canceled are also in the top 5 contributors to revenue with two of them are high value item (mobile & tablets and Appliance) in refund, customer have to cancelled the order forcefully or not

From Q3 2016 - Q3 2018

November always gets more than 10% cancel orders each year, which is 4x the average number of refund each month

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

18 of 21

Canceled Items

The top 3 items that are canceled are also in the top 5 contributors to revenue with mobile & tablets is most frequent cancelled

November always gets more than 13% cancel orders each year, which is 3x the average number of cancel each month

From Q3 2016 - Q3 2018

Business problem

Summary & Recommendation

Dataset Overview

Data Analysis

19 of 21

Data Analysis

High Transaction

Always happen during a promo, especially during the White / Blessed Friday event in November.

4. Highlight Summary

Top Product Category

Mobiles & Tablets and Fashion category show a significant continuous uptrend since 2016. However their cancelation & refund rate were also high

Favorable Payment Method

COD has been the favorite payment method for since 2016 but started to be replaced by E-wallet on 2017 onwards. The customer behavior begin to change to digital payments onward over the year

Canceled Order

are always more than 40% in every quarter and even growing more after Q3 2017 - Q3 2018

Dataset Overview

Business Problem

Summary & Recommendation

20 of 21

Recommendation

High Cancelation Rate

24% canceled reason was refund

80% GMV comes from 6 product

> 40 % cancelation rate within Q3 2016 - Q3 2018.

​

80% GMV is comes only from 6 of 15 categories which 3 of them were low in value item.

What we know

Solutions

Create an upselling events for the underperform 20% categories. Having bulk buying discount for low value item such as fashion to double basket size.

Refunds occur when the stock of goods runs out even though the payment has been processed, or the goods sent do not meet customer expectations (wrong product or poor quality).

Leveraging positive insight

Promotion (such as Blessed Friday) can acquire a lot of new customer which is good for growth of company.

​

Customer who registered from 2016 are most loyal customer.

​

Have more promotion to acquire new customer and more transaction for increasing total customer and GMV.

​

Create loyalty program for loyal customer (such as member point).

​

Performing customer journey funnel analysis for more detail to evaluate what and when happened in cancelation.

Create a good scheduling system to stock items, and use previous data for stock projections for every month especially for the next high season to minimize refunds.

Dataset Overview

Business Problem

Data Analysis

Summary & Recommendation

Goals A -related

Goals B -related

21 of 21

Thank You!